30.01.2015 Views

Medspiration – System Requirements Document - Data User Element

Medspiration – System Requirements Document - Data User Element

Medspiration – System Requirements Document - Data User Element

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Customer :<br />

Contract No :<br />

WP No :<br />

ESA / ESRIN<br />

AO/1-4362/03/I-LG<br />

2100<br />

<strong>Document</strong> Ref :<br />

Issue Date :<br />

Issue :<br />

MED-SOC-RS-001_1<br />

MED-SOC-RS-001_1<br />

22 February 2004<br />

F<br />

Issue F<br />

Title :<br />

<strong>Medspiration</strong> – <strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Abstract<br />

: This document defines the <strong>System</strong> <strong>Requirements</strong> for the <strong>Medspiration</strong> SST data<br />

processing system. It translates the <strong>User</strong> <strong>Requirements</strong> and the GHRSST-PP <strong>Data</strong><br />

Processing Specification into a Reference Baseline for designing the <strong>Medspiration</strong><br />

system.<br />

Author : Approval :<br />

Ian Robinson<br />

SOC-LSO<br />

Peter Challenor<br />

SOC-LSO Deputy Head<br />

Accepted :<br />

Distribution<br />

: Olivier Arino, ESRIN<br />

Dino Di Cola, ESRIN<br />

Craig Donlon, GHRSST-PP International Office Director<br />

Ian Robinson, SOC, Project Manager<br />

James Rickards, VEGA, DPM<br />

<strong>Medspiration</strong> Project Team<br />

Filename:<br />

med-soc-rs-001_1_f.doc<br />

Copyright © 2004 Southampton Oceanography Centre<br />

All rights reserved.<br />

SOC Laboratory for Satellite Oceanography<br />

European Way, Southampton, SO14 3ZH, UK<br />

Tel: +44 (0)23 8059 3438 Fax: +44 (0)23 6059 3059<br />

www.soc.soton.ac.uk<br />

© 2004 SOC Page 1 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

© 2004 SOC Page 2 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

This Page Is Intentionally Blank<br />

© 2004 SOC Page 3 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

TABLE OF CONTENTS<br />

1. INTRODUCTION ....................................................................................................................... 9<br />

1.1 Purpose and Scope .............................................................................................................. 9<br />

1.2 Structure of the <strong>Document</strong>.................................................................................................... 9<br />

1.3 Applicable and Referenced <strong>Document</strong>s ............................................................................. 10<br />

1.4 Definitions of Terms............................................................................................................11<br />

1.5 <strong>System</strong> Requirement Format.............................................................................................. 11<br />

2. SYSTEM OVERVIEW ............................................................................................................. 14<br />

2.1 <strong>System</strong> Context ..................................................................................................................14<br />

2.1.1 The importance of SST definitions used in <strong>Medspiration</strong> ............................................ 14<br />

2.1.2 <strong>Medspiration</strong> SST Products ......................................................................................... 16<br />

2.2 Overall structure of the <strong>Medspiration</strong> Processing <strong>System</strong>.................................................. 18<br />

2.3 Outline of the individual processes at each facility............................................................. 21<br />

2.3.1 L2P processing at CMS ............................................................................................... 21<br />

Ingest 21<br />

Quality Check ....................................................................................................................... 21<br />

Merging with ancillary data................................................................................................... 22<br />

Generate confidence data.................................................................................................... 22<br />

Generate L2P products ........................................................................................................ 22<br />

2.3.2 Processing at IFREMER .............................................................................................. 22<br />

Archive 23<br />

L4 UHR data production at IFREMER ................................................................................. 23<br />

Match-up data records production at IFREMER .................................................................. 24<br />

Dissemination of data from IFREMER ................................................................................. 24<br />

Control of <strong>Medspiration</strong> Processing at IFREMER................................................................ 24<br />

2.3.3 Processing at SOC....................................................................................................... 25<br />

AATSR Ingestion.................................................................................................................. 25<br />

Diagnostic data set generation at SOC................................................................................ 25<br />

3. FUNCTIONAL REQUIREMENTS ........................................................................................... 26<br />

3.1 <strong>System</strong> <strong>Requirements</strong> ........................................................................................................ 26<br />

3.1.1 General.........................................................................................................................26<br />

3.1.2 Input data ..................................................................................................................... 26<br />

3.1.3 <strong>Data</strong> products............................................................................................................... 26<br />

3.1.4 Archive and dissemination ........................................................................................... 27<br />

3.1.5 Overall system structure .............................................................................................. 28<br />

© 2004 SOC Page 4 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

3.2 CMS Facility ........................................................................................................................30<br />

3.2.1 Scheduler......................................................................................................................31<br />

3.2.2 Ingestion .......................................................................................................................31<br />

3.2.3 Quality Check ...............................................................................................................32<br />

3.2.4 Merging auxiliary data with SST ...................................................................................34<br />

3.2.5 Confidence value derivation .........................................................................................36<br />

3.2.5.1 Microwave sensor case ...........................................................................................37<br />

3.2.5.2 Infrared case............................................................................................................39<br />

3.2.6 SSES Assignation.........................................................................................................42<br />

3.2.7 L2P evaluation ..............................................................................................................42<br />

3.3 Ifremer Facility.....................................................................................................................43<br />

3.3.1 Archive..........................................................................................................................44<br />

3.3.1.1 Reference data ........................................................................................................44<br />

3.3.1.2 Online storage..........................................................................................................44<br />

3.3.1.3 Offline storage..........................................................................................................45<br />

3.3.1.4 Backup .....................................................................................................................45<br />

3.3.2 Generate L4 analysed products ...................................................................................45<br />

3.3.3 Matchup <strong>Data</strong> Base ......................................................................................................48<br />

3.3.3.1 Collecting the in-situ data ........................................................................................49<br />

3.3.3.2 Colocating the in-situ and satellite data...................................................................49<br />

3.3.3.3 Generating the MDB records ...................................................................................51<br />

3.3.4 Dissemination ...............................................................................................................52<br />

3.3.4.1 Ftp push/pull ............................................................................................................52<br />

3.3.4.2 DODS.......................................................................................................................52<br />

3.3.4.3 Live Access Server ..................................................................................................52<br />

3.3.5 <strong>System</strong> controller at Ifremer..........................................................................................53<br />

3.3.5.1 <strong>Data</strong>flow control .......................................................................................................53<br />

3.3.5.2 Processing flow control............................................................................................54<br />

3.3.5.3 Triggering the processing ........................................................................................54<br />

3.3.5.4 Chaining the processes ...........................................................................................55<br />

3.3.5.5 Operation log system...............................................................................................55<br />

3.3.5.6 <strong>System</strong> monitoring ...................................................................................................55<br />

3.3.5.7 Synchronization with GHRSST-PP..........................................................................55<br />

3.3.5.8 Exchange MMR .......................................................................................................56<br />

3.3.5.9 Exchange error messages......................................................................................56<br />

3.3.5.10 Exchange data...................................................................................................57<br />

3.4 SOC Facility ........................................................................................................................57<br />

© 2004 SOC Page 5 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

3.4.1 AATSR Ingestion.......................................................................................................... 57<br />

3.4.2 HR-DDS Generation .................................................................................................... 58<br />

3.4.3 HR-DDS Archiving........................................................................................................ 59<br />

3.4.4 HR-DDS Dissemination................................................................................................ 59<br />

3.4.5 HR-DDS Controller....................................................................................................... 59<br />

4. INTERFACE REQUIREMENTS.............................................................................................. 60<br />

4.1 Internal................................................................................................................................ 60<br />

4.1.1 Interfaces within the CMS processing facility............................................................... 60<br />

4.1.2 Interfaces between L2P production and archiving systems......................................... 60<br />

4.1.3 Interfaces between archiving and dissemination systems........................................... 61<br />

4.1.4 Interfaces between archiving and MDB generation systems....................................... 61<br />

4.1.5 Interfaces between archiving and L4 generation system............................................. 61<br />

4.1.6 Interfaces within the archiving system ......................................................................... 62<br />

4.1.7 Interfaces between HR-DDS production and dissemination systems ......................... 62<br />

4.1.8 Interfaces between IFREMER system controller and other systems........................... 63<br />

4.2 External............................................................................................................................... 64<br />

4.2.1 Satellite data providers................................................................................................. 64<br />

4.2.2 In situ data provider...................................................................................................... 65<br />

4.2.3 GDAC ........................................................................................................................... 65<br />

4.2.4 GDAC MDB.................................................................................................................. 65<br />

4.2.5 Master Metadata Repository ........................................................................................ 66<br />

4.2.6 GDS error log system................................................................................................... 66<br />

4.2.7 <strong>User</strong> interfaces ............................................................................................................. 67<br />

4.2.8 DDS external interfaces ............................................................................................... 67<br />

5. PERFORMANCE REQUIREMENTS ...................................................................................... 69<br />

5.1 Scenario Definitions............................................................................................................69<br />

5.2 Characterisation of L2 input data........................................................................................ 70<br />

5.2.1 Satellite SST................................................................................................................. 70<br />

5.2.2 Auxiliary satellite data .................................................................................................. 71<br />

5.2.3 NWP model outputs ..................................................................................................... 72<br />

5.2.4 Reference data sets ..................................................................................................... 72<br />

5.3 Characterisation of in situ data ........................................................................................... 73<br />

6. EXPANDABILITY REQUIREMENTS...................................................................................... 76<br />

6.1 General ............................................................................................................................... 76<br />

6.2 Generation of L2P...............................................................................................................76<br />

© 2004 SOC Page 6 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

6.3 Diagnostic data set..............................................................................................................77<br />

7. SECURITY REQUIREMENTS.................................................................................................78<br />

8. DESIGN REQUIREMENTS .....................................................................................................79<br />

9. RELIABILITY, AVAILABILITY AND MAINTAINABILITY REQUIREMENTS ........................80<br />

10. TESTING REQUIREMENTS....................................................................................................81<br />

11. SAFETY REQUIREMENTS.....................................................................................................82<br />

12. CUSTOMER FURNISHED ITEMS ..........................................................................................83<br />

13. GLOSSARY .............................................................................................................................84<br />

APPENDIX A TRACE BETWEEN USER REQUIREMENTS AND SYSTEM<br />

REQUIREMENTS...........................................................................................................................88<br />

© 2004 SOC Page 7 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

AMENDMENT POLICY<br />

This document shall be amended by releasing a new edition of the document in its<br />

entirety. The Amendment Record Sheet below records the history and issue status of<br />

this document.<br />

AMENDMENT RECORD SHEET<br />

ISSUE DATE DCI No REASON<br />

A 19 Jan 2004 N/A Initial Issue<br />

B 15 Feb 2004 Extensive addition of detailed requirements from<br />

Ifremer, CMS / Avel Mor, and SOC.<br />

Addition of figures<br />

Note the addition of extra categories of SR.<br />

C 18 Feb 2004 Editing by ISR to harmonise the requirements,<br />

some additions to meet the needs of the GDS,<br />

corrections from CMS /Avel Mor and SOC<br />

D 19 Feb 2004 Addition of material from CLS and figure from<br />

SOC. Additions to section 2.<br />

E 20 Feb 2004 Corrections from CMS and SOC.<br />

F 22 Feb 2004 Final corrections from Ifremer. Added appendix<br />

with requirements trace matrix<br />

© 2004 SOC Page 8 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

1. INTRODUCTION<br />

1.1 Purpose and Scope<br />

The purpose of this document is to define all the <strong>System</strong> <strong>Requirements</strong> (SR) that are<br />

applicable to <strong>Medspiration</strong>. The SRs are arranged in a number of categories that will aid<br />

access to them. An associated method for validation of each SR has been identified.<br />

The SRs have been defined to ensure full verification of the <strong>User</strong> <strong>Requirements</strong> (UR)<br />

expressed in the Statement of Work (SOW), the <strong>Medspiration</strong> <strong>User</strong> <strong>Requirements</strong><br />

document and the GHRSST-PP <strong>Data</strong> Processing Specification (GDS). Throughout this<br />

document the SRs are identified and inter-spaced with explanatory text (italicised) to aid<br />

the readability of the document.<br />

Referencing of the system requirements to the user requirements is documented in the<br />

Appendix.<br />

1.2 Structure of the <strong>Document</strong><br />

After this introduction, the document is divided into a number of major sections that are<br />

briefly described below:<br />

SYSTEM OVERVIEW<br />

Provides an overview of the system.<br />

3 FUNCTIONAL REQUIREMENTS<br />

Defines the functional system requirements, uniquely numbers them and states<br />

the method of verification.<br />

4 INTERFACE REQUIREMENTS<br />

Defines the interface system requirements, uniquely numbers them and states<br />

the method of verification. This includes two sub-sections for internal and<br />

external interfaces. In this context external interfaces are defined as being<br />

external to the <strong>Medspiration</strong> system (i.e. not inter-module interfaces which are<br />

defined as internal).<br />

© 2004 SOC Page 9 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

5 PERFORMANCE REQUIREMENTS<br />

MED-SOC-RS-001_1<br />

Issue F<br />

Defines the performance and capacity system requirements, uniquely numbers<br />

them and states the method of verification.<br />

6 EXPANDABILITY REQUIREMENTS<br />

Defines the expandability system requirements, uniquely numbers them and<br />

states the method of verification.<br />

7 SECURITY REQUIREMENTS<br />

Defines the security system requirements, uniquely numbers them and states the<br />

method of verification.<br />

8 DESIGN REQUIREMENTS<br />

Defines the design system requirements, uniquely numbers them and states the<br />

method of verification.<br />

9 RELIABILITY, AVAILABILITY AND MAINTAINABILITY REQUIREMENTS<br />

Defines the reliability, availability and maintainability system requirements,<br />

uniquely numbers them and states the method of verification.<br />

10 TESTING REQUIREMENTS<br />

Defines the testing system requirements, uniquely numbers them and states the<br />

method of verification.<br />

11 SAFETY REQUIREMENTS<br />

Defines the safety system requirements, uniquely numbers them and states the<br />

method of verification.<br />

GLOSSARY<br />

This defines any acronyms used through this document.<br />

1.3 Applicable and Referenced <strong>Document</strong>s<br />

The following is a list of documents with a direct bearing on the content of this report.<br />

Where referenced in the text, these are identified as RD.n, where 'n' is the number in the<br />

© 2004 SOC Page 10 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

list below:<br />

MED-SOC-RS-001_1<br />

Issue F<br />

RD.1<br />

<strong>Medspiration</strong> Statement of Work.<br />

RD.2<br />

<strong>Medspiration</strong> <strong>User</strong> <strong>Requirements</strong> <strong>Document</strong><br />

RD.3 The Recommended GHRSST-PP Processing Specification GDS Version 1<br />

revision 1.1, 7 th January, 2004<br />

1.4 Definitions of Terms<br />

The following terms have been used in this report with the meanings shown.<br />

Term<br />

Definition<br />

External Interface<br />

An interface external to the <strong>Medspiration</strong> system<br />

Internal Interface<br />

An interfaces contained entirely within the system e.g.<br />

inter-module interfaces.<br />

1.5 <strong>System</strong> Requirement Format<br />

Each SR has been identified with a unique identifier of the form:<br />

SR.[CLA.]MOD.NUM.V<br />

where:<br />

SR = <strong>System</strong> Requirement,<br />

CLA = Class of requirement (see Table 1-2 for definitions). This field is<br />

absent in the label for functional requirements.<br />

MOD = Functional Module or Class (see Table 1-1 for definitions),<br />

NUM = Requirement number and<br />

V = Version. Initially this is not required. However any subsequent<br />

versions are labelled A… Z.<br />

Note: the method of requirement validation is not included in the actual<br />

identifier, but rather separated by “/”. Validation methods include: T –<br />

Test, I – Inspection & A – Analysis.<br />

© 2004 SOC Page 11 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

The functional requirements are differentiated according to the module to which they<br />

refer. Table 1-1 shows the decomposition of the <strong>System</strong> into modules, and defines the<br />

SR reference which documents the module Sub-system <strong>Requirements</strong>. <strong>Requirements</strong><br />

are also differentiated by class, as identified in Table 1-2. In this case the part of the<br />

system constrained by the requirement is also specified<br />

Table 1-1 – Functional requirements distinguished by Module<br />

MOD<br />

SYS<br />

ING<br />

QLC<br />

MRG<br />

CFD<br />

L2P<br />

CTR<br />

ARC<br />

L4A<br />

MDB<br />

DIS<br />

DDS<br />

Name of functional module<br />

<strong>System</strong> Level <strong>Requirements</strong><br />

Ingestion of input SST and auxiliary data<br />

Quality check of ingested data<br />

Merging of auxiliary data<br />

Confidence derivation<br />

Level 2-P Processor<br />

<strong>System</strong> controller (Ifremer)<br />

Archive<br />

Level 4 analysis processor<br />

Match-up database processor<br />

Dissemination processor<br />

High Resolution Diagnostic data set<br />

© 2004 SOC Page 12 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Table 1-2 Other classes of requirements<br />

MED-SOC-RS-001_1<br />

Issue F<br />

CLA<br />

INT<br />

EXT<br />

PER<br />

EXP<br />

SEC<br />

DES<br />

REL<br />

TST<br />

SAF<br />

CFI<br />

Class of requirement<br />

Internal Interface requirement)<br />

External Interface requirement<br />

Performance requirement<br />

Expandability requirement<br />

Security requirement<br />

Design requirement<br />

Reliability, availability and maintainability requirement<br />

Testing requirement<br />

Safety requirement<br />

Customer furnished items<br />

© 2004 SOC Page 13 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

2. SYSTEM OVERVIEW<br />

2.1 <strong>System</strong> Context<br />

The purpose of the <strong>Medspiration</strong> sea surface temperature (SST) processing system is to<br />

fulfil the role of the European regional data assembly centre (RDAC) for the Global<br />

Ocean <strong>Data</strong> Assimilation Experiment (GODAE) High Resolution Sea Surface<br />

Temperature Pilot Project (GHRSST-PP). The primary aim of GHRSST-PP is to develop<br />

and operate an operational demonstration system that will deliver global coverage SST<br />

data products for the diverse needs of GODAE and the wider scientific community, at<br />

high resolution (better than 10 km and resolving the diurnal cycle) and within a short<br />

interval (~24 hours) from real-time acquisition. In parallel with RDACs serving other parts<br />

of the world, <strong>Medspiration</strong> will produce in near real time a new generation of SST data<br />

products for Europe, based on the combination of existing SST products from both<br />

infrared and microwave radiometers (nominally level 2 although some are level 3) already<br />

being generated and served by several different agencies. The ingested SST data are<br />

referred to generically as L2 SST inputs<br />

The function of the <strong>Medspiration</strong> processing system is threefold:<br />

• To generate specific combined SST products (described in 2.1.2) in near-real time<br />

and to serve them to the GDAC and to European operational ocean models.<br />

• To populate an off-line archive in which to curate them.<br />

• To provide a data product dissemination service for all types of users.<br />

2.1.1 The importance of SST definitions used in <strong>Medspiration</strong><br />

In its search for accuracy and rigour in calibration when combining SST detected by<br />

different classes of instrument, the GHRSST-PP processing model [RD.1] has had to<br />

take into account the thermal structure of the near-surface layers of the ocean. Since this<br />

has significant consequences for a number of the data processing stages, it is<br />

appropriate to provide a brief explanation at this stage. Figure 2-1 illustrates<br />

schematically how temperature varies in the upper few metres of the water column. Five<br />

different expressions of SST are shown.<br />

• SSTint is the hypothetical concept of the temperature of the interfacial layer of water<br />

and air molecules at the sea surface. It is not measurable and not used in GHRSST-<br />

PP.<br />

• SSTskin is defined within GHRSST-PP as the radiometric temperature of the surface<br />

© 2004 SOC Page 14 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

measured by an infrared radiometer operating in the 10 – 12 µm waveband.<br />

Physically it represents the temperature of the water at a depth of approximately 10 –<br />

20 µm.<br />

• SSTsubskin represents the temperature at the base of the thermal skin layer. The<br />

thermal skin layer is a region less than 1 mm deep in which the convective exchange<br />

of heat by turbulent mixing is inhibited by the proximity of the sea surface, so that the<br />

net outward flow of heat through the surface creates a steep reduction of temperature<br />

towards the surface. Below the skin layer, turbulent processes ensure that the<br />

temperature is nearly uniform over a depth of at least a few centimetres. By<br />

definition within GHRSST-PP this corresponds to the SSTsubskin temperature. In<br />

practice SSTsubskin is assumed to be approximately equal to the radiometric<br />

temperature measured by a microwave radiometer operating in the 6-11 GHz<br />

frequency band, although the relationship is not exact or fully known.<br />

• SSTdepth is the generic term used to represent the temperature measured by a<br />

contact thermometer within the upper few metres of the water column and generally<br />

referred to as the “bulk” SST. If the water column below the skin layer is uniform<br />

(typically the case at night and when wind mixing is strong) then SSTdepth is the<br />

same as SSTsubskin, irrespective of the actual depth of the measurement (see<br />

Figure 2-1a). However, when daytime solar shortwave radiation penetrates to heat<br />

the water below the skin layer, and if wind mixing is weak, stable stratification<br />

develops in the upper few metres of the water column, in which the temperature<br />

increases towards the sea surface (apart from the cool skin layer which lies right at<br />

the surface - see Figure 2-1b). This phenomenon is called a diurnal thermocline.<br />

When it occurs a measurement of SSTdepth, made from a buoy or ship-mounted<br />

thermometer which does not precisely specify the sampling depth, is of limited value.<br />

A diurnal thermocline almost always collapses to a uniform temperature some time<br />

after sunset when surface cooling removes the excess heat. Under calm conditions,<br />

however, it may take several hours for the diurnal thermocline to decay entirely.<br />

• SSTfnd is defined within GHRSST-PP as the temperature at the base of the diurnal<br />

thermocline. It is so named because it represents the foundation temperature on<br />

which the diurnal thermocline develops during the day. SSTfnd changes only<br />

gradually along with the upper layer of the ocean, and by definition it is independent<br />

of skin SST fluctuations due to wind- and radiation-dependent diurnal stratification or<br />

skin layer response. It is therefore updated at intervals of 24 hrs. SSTfnd<br />

corresponds to the temperature of the upper mixed layer which is the part of the<br />

ocean represented by the top-most layer of grid cells in most numerical ocean<br />

© 2004 SOC Page 15 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

models. It is never observed directly by satellites, but it comes closest to being<br />

detected by a microwave radiometer which penetrates the skin, at dawn when the<br />

previous day’s diurnal stratification can be assumed to have decayed and<br />

SSTsubskin, SSTdepth and SSTfnd are equal.<br />

Figure 2-1 Schematic diagram showing (a) idealised night-time vertical temperature<br />

deviations from SSTfnd and (b) idealised day-time vertical temperature deviations<br />

from SSTfnd in the upper ocean.<br />

Although the distinctions between the different definitions of SST may seem arcane, they<br />

are very significant when uncertainties of temperature measurement less than 0.3ºC are<br />

being sought.<br />

2.1.2 <strong>Medspiration</strong> SST Products<br />

The <strong>Medspiration</strong> processing system (MPS) will adhere closely to the data processing<br />

model and product specifications defined by the GHRSST-PP [RD.1]. It will produce the<br />

following data products:<br />

• L2P SST products for the European RDAC area which spans from 70S to 90N (to the<br />

ice limit) and from 100W to 45E and contains the whole Atlantic and the adjacent<br />

European seas (see Figure 2-2). Parts of the Pacific Ocean within these limits are<br />

expressly excluded. L2P products consist of the L2 SST data in their native sampling<br />

grid as ingested, with the addition of a quantitative confidence value attached to<br />

every data point. The confidence value will be based on the sensor specific error<br />

© 2004 SOC Page 16 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

statistics (SSES) to be supplied by GHRSST-PP and ancillary data such as wind<br />

speed and solar radiation, which can influence the diurnal variability of SST. The<br />

L2P data will be provided in near-real time to the GHRSST-PP global data assembly<br />

centre (GDAC) and to selected operational users in Europe for assimilation into<br />

ocean forecasting models. They will also be archived and served widely to general<br />

European users.<br />

Figure 2-2 The GHRSST-PP EURDAC product grid domain (Note that the Pacific<br />

areas are not included)<br />

• Available in situ measurements of SST will be matched with coincident samples<br />

within the L2P SST data, in order to prepare records for delivery to the GHRSST-PP<br />

match-up data base (MDB). This is required by GHRSST-PP to determine the<br />

sensor specific error statistics (SSES) needed when assigning quality values for the<br />

L2P data. The geographical scope of the MDB records to be delivered is the same<br />

European area as defined for the L2P products.<br />

• A L4 SST product will be produced at ultra-high resolution (UHR) on a 2 × 2 km grid<br />

for the Mediterranean Sea (see Figure 2-3). Optimal interpolation (OI) techniques will<br />

be used to combine coincident L2P measures of SST from different types of sensor<br />

and to fill gaps where no observations are available. Whereas L2P data essentially<br />

represent the skin or sub-skin SST, the L4 SST product is defined to represent<br />

SSTfnd (as explained in 2.1.1) and is updated only once every 24 hr. Using<br />

GHRSST-provided parameterisations of the skin layer and diurnal thermocline<br />

thermal structure, L2P products will be converted to SSTfnd before applying OI. The<br />

© 2004 SOC Page 17 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

L4 product also contains an estimate of the skin temperature and its diurnal variation<br />

derived from the analysed SSTfnd.<br />

Figure 2-3 The <strong>Medspiration</strong> Mediterranean Sea product grid domain (30N to 46N<br />

and 6W to 36.5E at 2 km spatial resolution).<br />

• Entries to the GHRSST-PP diagnostic data set (DDS) will be compiled by extracting,<br />

and resampling onto a 0.01º grid, full resolution sub scenes for a set of predefined<br />

small areas (typically 2º × 2º) from every L2P SST and L4 dataset generated by the<br />

<strong>Medspiration</strong> processor. The DDS sites are defined by GHRSST-PP. The purpose<br />

of the DDS is to allow intercomparison of the characteristics of the SST derived from<br />

different sensor types, in order to facilitate the research needed to test and improve<br />

the data merging methods used in GHRSST-PP.<br />

2.2 Overall structure of the <strong>Medspiration</strong> Processing <strong>System</strong><br />

Figure 2-4 outlines schematically the main processing tasks required to produce the four<br />

types of SST products described in 2.1.2. The figure shows the main classes of inputs<br />

and outputs, and assigns distinct processing stages to different boxes. The whole<br />

system is distributed across the facilities of three contractors, CMS, Ifremer and SOC, as<br />

indicated in the figure.<br />

The underlying data flow through the system is basically simple. Level 2 SST data<br />

products generated from different sensors and produced by the mission ground segment<br />

or agency responsible for each are obtained by Internet and ingested into the first main<br />

processing centre. This is hosted in CMS where use will be made of the fact that CMS is<br />

already an operational provider of processed AVHRR data, which is a key SST input<br />

stream for the L2P product. Consequently one of the major sources of input data will be<br />

provided internally. At the same time the CMS processing facility will ingest auxiliary<br />

satellite data, such as wind speed, needed for the quality assessment. The SSES,<br />

reference data fields and any other necessary information will be supplied by the<br />

© 2004 SOC Page 18 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

GHRSST-PP International project office (IPO) so that quality values can be assigned to<br />

each ingested data point, applying the rules set out by the GHRSST data processing<br />

plan, and the new L2P dataset can be generated with the quality flags attached. This<br />

completes the involvement of CMS.<br />

As soon as each L2P data set is generated in near real time the products are pushed to<br />

the core processing centre. This is at IFREMER, where use will also be made of existing<br />

operational capability in processing, archiving and distributing satellite data. The<br />

IFREMER computing facility will support a number of tasks. First is the onward<br />

dissemination of the L2P data to GHRSST and to any operational users requiring them.<br />

Second is the archiving of the L2P data. Third is the use of the L2P data in the<br />

production of the level four interpolated products, using an OI scheme configured to<br />

match the local variability statistics for the region in question. The resulting L4 products<br />

for the Mediterranean Sea will then be disseminated to users and also placed in the<br />

<strong>Medspiration</strong> archive.<br />

The L4 processor is required to accomplish two processes whose detailed specification<br />

by the user (GHRSST-PP) is still under review. These are the conversion of the L2P<br />

data from skin and subskin to foundation temperature, and the subsequent optimal<br />

interpolation that merges SST data from all sources to generate the L4 SST product. The<br />

system requirements for the L4 processor must be defined in such a way that there is no<br />

uncertainty attached to the design of the fundamental functional architecture, and that<br />

any remaining uncertainties in the user specification are confined to the content of<br />

configuration files.<br />

The fourth module in the IFREMER facility is the use of the L2P data to generate entries<br />

for the GHRSST-PP match-up data base (MDB). IFREMER is well placed to handle the<br />

extraction of satellite data from the L2P inputs that are collocated with in situ<br />

observations of SST to form the MDB records. That is because the focal point for the<br />

collection, processing, dissemination and archiving of many of Europe’s operational<br />

measurements of SST from buoys and other platforms is already based at IFREMER in<br />

the CORIOLIS project.<br />

© 2004 SOC Page 19 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

L2 SST data<br />

from several<br />

sensors<br />

Auxiliary<br />

data (wind,<br />

waves etc.)<br />

In situ<br />

SST data<br />

CMS<br />

Ingest input<br />

and apply<br />

quality control<br />

Generate L2P<br />

data products<br />

Sensor<br />

specific error<br />

statistics<br />

Reference<br />

data fields<br />

Generate<br />

MDB<br />

records<br />

L2P data<br />

Generate<br />

analysed SST<br />

L4 products<br />

MDB<br />

records<br />

IFREMER<br />

<strong>Medspiration</strong><br />

Final Archive<br />

L4 SSTfnd<br />

UHR<br />

Disseminate<br />

L2P data to<br />

GHRSST-PP<br />

GDAC & users<br />

Disseminate<br />

L4 UHR data<br />

to EU users<br />

GHRSST<br />

MDB<br />

Archive<br />

Extract DDS<br />

L2P granules<br />

SOC<br />

Extract DDS<br />

L4 granules<br />

L2P DDS<br />

granules<br />

<strong>Medspiration</strong><br />

DDS Archive<br />

L4 DDS<br />

UHR<br />

granules<br />

Figure 2-4 Flow diagram for data processing, showing the distribution between<br />

facilities. Note that the flow of AATSR data through SOC is omitted for simplicity.<br />

© 2004 SOC Page 20 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

The remaining task is the extraction of the diagnostic data sets (DDS) granules from<br />

every L2P and L4 dataset located in the EURDAC area of responsibility. This is the<br />

responsibility of SOC which will serve as a node for the GHRSST-PP DDS. <strong>Data</strong> will be<br />

regridded to a standard 0.01º grid base defined by the GDS. It will be archived and made<br />

accessible through a distributed oceanographic data system (DODS) server. The DDS is<br />

an important resource for the GHRSST-PP task of reviewing and improving its<br />

processing model.<br />

SOC has one other detailed operational role which is not marked on the high level Figure<br />

2-4. <strong>Medspiration</strong> has the task of obtaining the full resolution AATSR data for the DDS<br />

sites globally. SOC will download the globally complete near-real time AATSR SST<br />

product from Envisat as it is made available on the ESA server. The DDS sites will be<br />

stripped out for archive and the non-EUDAC sites will be provided for GHRSST-PP. At<br />

the same time the Mediterranean Sea and Atlantic Ocean data will also be extracted and<br />

gridded, for all orbits which overlap the EURDAC region. These will then be provided to<br />

CMS for ingestion into the L2P processor.<br />

2.3 Outline of the individual processes at each facility<br />

2.3.1 L2P processing at CMS<br />

Figure 3-1 shows the processing tasks that must be performed to generate the L2P<br />

products as defined by GHRSST-PP.<br />

Ingest<br />

The ingestion module must communicate with the diverse providers of the L2 SST data,<br />

ice, wind speed, aerosol optical depth (AOD) and surface solar irradiance (SSI) data. It<br />

must be scheduled regularly to download new data as they become available. A<br />

successful download results in the generation of a data set description (MMR-DSD)<br />

which is sent to the GHRSST-PP master metadata repository (MMR). A failed download<br />

triggers an error message to the data provider.<br />

Quality Check<br />

The data are given a gross quality check to confirm the integrity of the files, and pixel-bypixel<br />

quality tests to ensure that only SST data with quality above a minimum standard<br />

are entered into the system. The quality checks require regularly updated information<br />

about the status of particular SST sensors systems to be provided by GHRSST-PP and /<br />

or the data providers. The data sets are reformatted and renamed at this stage. They<br />

© 2004 SOC Page 21 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

will be extracted over the <strong>Medspiration</strong> EURDAC and UHR areas as appropriate. <strong>Data</strong><br />

are not resampled and remain in their native grid throughout the L2P processing. Failure<br />

of the quality checks triggers a message to the provider and to GHRSST-PP..<br />

Merging with ancillary data<br />

The SST datasets which successfully pass the quality checks are merged with<br />

corresponding ancillary data required for the subsequent L2P processing. This includes<br />

an estimate of the sea ice fraction, wind speed, aerosol optical depth and the SSI at the<br />

time and location of each SST sample. Wind speed is expected to come from NWP<br />

models if not produced simultaneously with the SST value. At the same time information<br />

on location, viewing angle and sun illumination angle are attached to each pixel, if not<br />

already in the data.<br />

Generate confidence data<br />

Confidence values are then generated for each pixel, based on the ancillary data now<br />

attached to the SST for each pixel. This requires a number of tests following a set of<br />

rules defined by GHRSST-PP. Two distinct pathways are followed, one for SST data<br />

obtained from infra red sensors and one for data from microwave radiometers. Within<br />

these paths all data follow the same set of tests, but these may be evaluated differently<br />

for different sensors, following criteria contained in configuration files to be supplied and<br />

updated by GHRSST-PP. The most complex of these tests are concerned with<br />

generating the microwave or infrared proximity confidence values (MPCV or IPCV) which<br />

depend on the spatial structure of the data, for example how far a pixel is away from the<br />

nearest ice or cloud (for infrared) or rain (for microwave).<br />

Generate L2P products<br />

An estimate of bias and standard deviation must be assigned to each pixel. These may<br />

be either derived from the data provider or based on GHRSST-PP sensor specific error<br />

statistics (SSES) using the IPCV or MPCV. In the last stage of processing at CMS all the<br />

additional data generated for each pixel by the foregoing processes are written into the<br />

final L2P data product, in netCDF format. The finished file is sent to the <strong>Medspiration</strong><br />

archive in IFREMER, records are sent to the MMR and the successful completion is<br />

logged.<br />

2.3.2 Processing at IFREMER<br />

The overview of the whole <strong>Medspiration</strong> processing structure at IFREMER is outlined in<br />

© 2004 SOC Page 22 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

Figure 3-4. The Archive is at the core, receiving data input from CMS and providing the<br />

source of DDS data for SOC. The archive supplies data to the dissemination function.<br />

Periodically the L4 processor uses L2P data from the archive to generate L4 products<br />

which are entered into the archive. Periodically the match-up dataset processor<br />

operates, drawing L2P data from the archive and in situ data from the CORIOLIS facility<br />

at IFREMER to generate MDB records for GHRSST-PP.<br />

Archive<br />

The archive receives L2P data as soon as they are produced at CMS. It receives L4 data<br />

products generated at IFREMER by the L4 processor. It also receives ancillary datasets<br />

of wind speed fields and surface solar irradiance required for the diurnal variability<br />

module of the L4 processor. The archive may receive data that are too old to meet the<br />

operational timeliness requirements of GHRSST-PP. <strong>Data</strong> are registered when they<br />

enter the archive, MMR records are sent to GHRSST-PP, and the location and timing of<br />

each dataset is entered into a database to facilitate searching. The archive is partitioned<br />

between on-line and off-line storage, and daily back-ups are made.<br />

L4 UHR data production at IFREMER<br />

The L4 UHR data production is performed on a daily cycle. All the L2P and ancillary data<br />

acquired by sensors between 00:00 to 23:59 on day T, which is available in the archive at<br />

06:00 UTC on day T+1, is used to generate L4 product to be delivered by 11:59 on day<br />

T+1. The L4 processor has four modules as shown in Figure 3-6.<br />

The first identifies the measured SST data available for each location and selects<br />

according to quality criteria, defined in a configuration file. All the measured SST data<br />

correspond to skin or subskin observations.<br />

The second module estimates the diurnal variation (DV) of the difference between<br />

SSTskin and SSTfnd, or SSTsubskin and SSTfnd, making use of ancillary data contained<br />

within the L2P products, additional wind speed or SSI data if available, and any regional<br />

factors that are supplied in a configuration file. By this means every L2P SST<br />

measurement is converted to an estimate of SSTfnd.<br />

The third module applies an optimal interpolation (OI) process to all the selected data,<br />

now converted to estimates of SSTfnd, in order to produce an analysed, gridded field of<br />

SSTfnd. It also attaches error estimates to each pixel in the field. This forms the core of<br />

the L4 SST product.<br />

© 2004 SOC Page 23 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

The fourth module uses an appropriate DV model to estimate the SSTskin associated<br />

with each SSTfnd in the analysed field, evaluated every two hours through the 24 h<br />

period represented by the final L4 product.<br />

While the overall structure of the L4 processor is well defined, the detailed criteria and<br />

models used in each of the four models are not yet optimized by the user (GHRSST-PP).<br />

Indeed it is expected that experience with operating the processor is needed before the<br />

OI and DV models can be finally chosen. For this reason the L4 processor must be<br />

designed in a highly modular way, with the specifications of the internal functions of each<br />

module contained in configuration files.<br />

Match-up data records production at IFREMER<br />

The functions within the MDR production are shown in Figure 3-7. In situ measurements<br />

of SST are obtained form the CORIOLIS system in IFREMER. This provides a<br />

convenient single point access to data from a number of different networks of buos and<br />

other measurement platforms. In situ data can be expected to be up to several days old<br />

when they are finally made available by CORIOLIS. Therefore the MDR production cycle<br />

is operated periodically to find matches for all the in situ data that have come available<br />

since the previous run, irrespective of their actual age.<br />

The in situ data are matched with any available satellite data within prescribed spacetime<br />

overlap windows. A number of conditions are attached to the complexity of<br />

matching discrete point samples to spatially distributed satellite data, and these are<br />

contained in several system requirements. The MDR are then sent to the GHRSST-PP<br />

MDB.<br />

Dissemination of data from IFREMER<br />

IFREMER provides access to the archive of L2P and L4 products through the<br />

dissemination module. This provides three different services, an ftp push-pull server, a<br />

DODS server and a live access server (LAS).<br />

Control of <strong>Medspiration</strong> Processing at IFREMER<br />

Management of the various different <strong>Medspiration</strong> processing functions being carried out<br />

at IFREMER requires a controller to ensure that they operate smoothly. This gives rise to<br />

the requirements presented in 3.3.5. While many of these are not directly traceable to<br />

the user requirements documents, they are inherently implied by the need for an<br />

autonomously operating system. This system controller also acts as a central log for all<br />

© 2004 SOC Page 24 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

relevant error and operation messages provided by the various <strong>Medspiration</strong> systems<br />

wherever they are located. A selection of these log messages is then redirected to the<br />

GHRSST-PP central error log system.<br />

2.3.3 Processing at SOC<br />

Figure 3-10 shows the part of the <strong>Medspiration</strong> processing system located at SOC.<br />

There are two main tasks to be performed, handling the ingest of the AATSR L2 global<br />

data, and extracting the diagnostic data set (DDS) granules of data from the main<br />

<strong>Medspiration</strong> archive at IFREMER.<br />

AATSR Ingestion<br />

<strong>Medspiration</strong> has a responsibility to extract the DDS granules from the AATSR for the<br />

whole world. This is in contrast with all the other data in <strong>Medspiration</strong> which cover<br />

European and Atlantic waters only. For this reason, and also to expedite AATSR data<br />

exchange between <strong>Medspiration</strong> and the ESRIN server, SOC will operate a mirroring<br />

function for AATSR data. From the AATSR data downloaded from ESRIN in Envisat<br />

format, the global DDS sites will be extracted and converted to netCDF format. Also the<br />

full data coverage of the EURDAC area will be extracted to provide the L2 input to the<br />

ingestion module at CMS. To avoid duplication of effort, the pixel location, viewing angle<br />

and sun angle information that is provided at a few tie points will be interpolated and<br />

attached to each pixel before transmission to CMS.<br />

Diagnostic data set generation at SOC<br />

The DDS granule generation for all sensors other than AATSR extends only to the<br />

European area DDS sites. As soon as data are entered into the IFREMER server the<br />

DDS granules will be extracted and MMR records sent to GHRSST-PP. Since the<br />

GHRSST-PP DDS is a distributed dataset, relying on the RDACs to house the data, it is<br />

implicit in the provision of DDS granules to GHRSST-PP that <strong>Medspiration</strong> should<br />

establish a DDS archive for the European DDS sites. This will be located at SOC.<br />

Associated with the DDS archive is an obligation of dissemination, leading to a<br />

requirement to serve the DDS data freely to external users.<br />

© 2004 SOC Page 25 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

3. FUNCTIONAL REQUIREMENTS<br />

3.1 <strong>System</strong> <strong>Requirements</strong><br />

3.1.1 General<br />

SR.SYS.0010/I <strong>Medspiration</strong> shall implement a finite demonstration service of GHRSST-<br />

PP products for the European area.<br />

SR.SYS.0015/I <strong>Medspiration</strong> shall demonstrate, and sustain for a 12 month period, an<br />

operational service able to ingest, process and disseminate large volumes of satellite<br />

data in near real time.<br />

SR.SYS.0020/I Within the scope set out by [RD-1], the <strong>Medspiration</strong> <strong>System</strong> shall<br />

conform to the GHRSST data processing specification (GDS) as defined in Version 1<br />

Revision 1. [RD-3]<br />

SR.SYS.0021/I If later versions of the GDM provided by the <strong>User</strong> contain significant<br />

differences from the version specified in SR.SYS.0010, those changes will be adopted by<br />

<strong>Medspiration</strong> only with the agreement of both the Contractor and the Customer.<br />

3.1.2 Input data<br />

SR.SYS.0030/I The system shall be capable of ingesting those level 2 SST, wind speed<br />

and surface solar irradiance (SSI) data products available in real time that are identified<br />

in 5.2.1, 5.2.2 and 5.2.3.<br />

SR.SYS.0040/I The system shall be capable of protecting the conditions of use (if any)<br />

imposed upon input data by the provider.<br />

3.1.3 <strong>Data</strong> products<br />

SR.SYS.0050/I <strong>Medspiration</strong> shall deliver GHRSST-PP L2P products for the Atlantic<br />

Ocean in near-real time (as defined by the GDS) to the GHRSST-PP GDAC and to<br />

selected operational users.<br />

SR.SYS.0051/I The Atlantic product shall encompass the area 70S to 90N (to the ice<br />

limit) and from 100W to 45E at a resolution of 1/12º (~10km)<br />

SR.SYS.0052/I The L2P products shall conform to the specification in [RD.1], [RD.2] and<br />

[RD.3] and detailed in 3.2.7<br />

© 2004 SOC Page 26 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.SYS.0055/I For every L2P file that is generated, a L2P MMR_FR shall be created and<br />

registered under the appropriate MMR_DSD at the GHRSST-PP MMR system.<br />

SR.SYS.0060/I <strong>Medspiration</strong> shall be able to produce records for the GHRSST-PP<br />

match-up data base (MDB) of in situ measurements of SST coincident in space and time<br />

with L2P data for the European area.<br />

SR.SYS.0061/I The MDB products shall conform to the specification in [RD.1], [RD.2]<br />

and [RD.3] and detailed in 3.3.3<br />

SR.SYS.0070/I <strong>Medspiration</strong> shall be able to extract GHRSST-PP Diagnostic <strong>Data</strong> Set<br />

(DDS) granules [RD-4] for the European area.<br />

SR.SYS.0071/I The DDS shall conform to the specification in [RD.1], [RD.2] and [RD.3]<br />

and detailed in 3.4<br />

SR.SYS.0080/I <strong>Medspiration</strong> shall be capable of delivering L4 UHR products for the<br />

Mediterranean Sea in near-real time to selected operational users.<br />

SR.SYS.0081/I The <strong>Medspiration</strong> Mediterranean Sea product grid domain shall cover<br />

from 30N to 46N and from 6W to 36.5E at 2 km spatial resolution.<br />

SR.SYS.0082/I The system for generating L4 products shall be based on an optimal<br />

interpolation (OI) procedure that can be adapted to widely differencing conditions through<br />

the use of configuration files.<br />

SR.SYS.0083/I The L4 products shall conform to the specification in [RD.1], [RD.2] and<br />

[RD.3] and detailed in 3.3.2<br />

3.1.4 Archive and dissemination<br />

SR.SYS.0090/I <strong>Medspiration</strong> shall archive all L2P and L4 products in a static archive so<br />

that they can be available for any future reprocessing, and for the GHRSST-PP<br />

reanalysis project.<br />

SR.SYS.0100/I <strong>Medspiration</strong> shall provide a dissemination service for data products<br />

including the operational delivery of SST products to specific European operational users<br />

and to the GHRSST-PP GDAC.<br />

SR.SYS.0110/I <strong>Medspiration</strong> shall provide access to all its data products as an ftp push<br />

service for registered clients and ftp pull service for all users.<br />

© 2004 SOC Page 27 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.SYS.0120/I <strong>Medspiration</strong> shall conduct an evaluation of data products and service by<br />

solicited user feedback including a proper validation of SST data products.<br />

3.1.5 Overall system structure<br />

SR.SYS.2010/I There shall be a processing facility at CMS that will interface with the L2<br />

providers to ingest the required input SST products and auxiliary files required for<br />

<strong>Medspiration</strong> (detailed in 3.2), that will produce quality values for each SST data point,<br />

and will append these to produce L2P products<br />

SR.SYS.2020/I L2P products generated at CMS shall be served immediately to the<br />

processing facility at Ifremer<br />

SR.SYS.2030/I There shall be a UHR/L4 product generation system at IFREMER that will<br />

produce UHR SSTfnd interpolated grids.<br />

The detailed requirements are set out in 3.3.2.<br />

SR.SYS.2040/I There shall be an archiving system at Ifremer for all L2P, UHR/L4<br />

products.<br />

The requirements for this are detailed in 3.3.1.<br />

SR.SYS.2050/I There shall be a dissemination system at IFREMER that will provide<br />

access to L2P and UHR/L4 products for GHRSST and <strong>Medspiration</strong> users<br />

The detailed requirements are set out in 3.3.4.<br />

SR.SYS.2060/I There shall be a MDB records production system at IFREMER that will<br />

interface with the CORIOLIS in-situ data provider and deliver match-up database records<br />

to the GHRSST-PP MDB.<br />

The requirements for this are detailed in 3.3.3.<br />

SR.SYS.2070/I There shall be a system controller at IFREMER that controls, monitors<br />

and logs the processing flow and the data flow within <strong>Medspiration</strong>.<br />

The requirements for this are detailed in 3.3.5.<br />

SR.SYS.2080 There shall be a facility at SOC to generate the DDS using L2P and L4<br />

data from Ifremer.<br />

© 2004 SOC Page 28 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

The detailed requirements are presented in 3.4.2.<br />

MED-SOC-RS-001_1<br />

Issue F<br />

SR.SYS.2081/I There shall be an archiving system in SOC for all HR-DDS granules<br />

generated by <strong>Medspiration</strong>.<br />

SR.SYS.2082/I There shall be a dissemination service at SOC for the <strong>Medspiration</strong> HR-<br />

DDS granules.<br />

SR.SYS.2090 The facility at SOC shall ingest global high resolution AATSR data from<br />

ESA and serve selected subsets to GHRSST-PP and to the L2P processor at CMS, as<br />

detailed in 3.4.1.<br />

© 2004 SOC Page 29 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

3.2 CMS Facility<br />

<strong>Data</strong> informations<br />

SCHEDULER<br />

query order<br />

PROVIDERS<br />

Query<br />

UPLOAD<br />

INGEST<br />

RAISE ERROR<br />

SEND MMR_DSD<br />

SST + ANC FILE<br />

RAISE ERROR<br />

CHECK<br />

PARAMETERS<br />

QUALITY<br />

CHECK<br />

ANC FILE<br />

CONFIG FILE<br />

IR<br />

CONFIDENCE<br />

SST FILE<br />

ANC FILES<br />

MERGING<br />

MERGED FILE<br />

CONFIG FILE<br />

MW<br />

CONFIDENCE<br />

CONFIDENCED FILE<br />

SSES DATA<br />

SSES<br />

ASSIGNATION<br />

L2P FILE<br />

L2P<br />

EVALUATION<br />

REPORT<br />

L2P FILE<br />

Figure 3-1 Schematic of the L2P processing system at CMS.<br />

SR.L2P.0010/I <strong>Data</strong> processing flow shall be designed to Ingest SST and auxiliary data<br />

sets.<br />

SR.L2P.0020/I <strong>Data</strong> processing flow shall be designed to control quality of input SST<br />

data sets.<br />

SR.L2P.0030/I <strong>Data</strong> processing flow shall be designed to assign auxiliary data (wind<br />

© 2004 SOC Page 30 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

speed, Surface Solar Irradiance (SSI) and, Aerosol Optical Depth (AOD)) to each SST<br />

measurement.<br />

SR.L2P.0040/I <strong>Data</strong> processing flow shall be designed to assign pixel confidence data to<br />

each SST measurement.<br />

SR.L2P.0050/I <strong>Data</strong> processing flow is designed to assign error estimates (called<br />

Sensor Specific Error Statistics, SSES) to each SST measurement.<br />

SR.L2P.0055/I Optional fields shall only be included prior to a cut-off time at which point<br />

the L2P data will be shipped ‘as is’.<br />

Comment: Optional fields are not described in GDS.<br />

SR.L2P.0060/I <strong>Data</strong> processing flow shall be designed to format a netCDF GHRSST-PP<br />

L2P output data file.<br />

SR.L2P.0070/I There shall be L2P products for every L2 satellite SST data file.<br />

Comment: 0 to 2 per L2 data file. 0 if rejected, 2 when ascending and descending orbits<br />

are included in the same data file.<br />

SR.L2P.0080/I L2P products shall have the spatial resolution of input data.<br />

SR.L2P.0090/I In the case of 1km input SST data streams, averaging of these data is<br />

permitted to provide data at a resolution of no greater than 1/12° latitude x longitude.<br />

3.2.1 Scheduler<br />

SR.ING.1010/I There shall be a scheduler that orders ingestion of the data streams.<br />

SR.ING.1020/I The scheduler shall be specifically programmed for each data stream, by<br />

a delay or a fixed time.<br />

3.2.2 Ingestion<br />

SR.ING.0010/I Ingestion of data streams shall establish operational access to data<br />

streams.<br />

SR.ING.0020/I If data are rejected by the GDS then an appropriate error message<br />

should be generated stating why the data have been rejected.<br />

SR.ING.0030/I Ingestion of data streams shall create a Master Metadata Repository<br />

© 2004 SOC Page 31 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

<strong>Data</strong> Set Description (MMR_DSD) record within the GHRSST-PP MMR for each data<br />

stream.<br />

SR.ING.0040/I Ingestion of data streams shall rename the data file by prefixing the<br />

RDAC identification code to the filename, as defined in Appendix 2 of [RD-3].<br />

Comment: A time stamp may be used to avoid ambiguity in original naming.<br />

SR.ING.0050/I The MMR implementation and delivery of MMR_DSD shall be performed<br />

as defined in GDS appendix A6.<br />

SR.ING.0060/I The data file integrity should be evaluated by identifying a data delivery/a<br />

data format problem.<br />

SR.ING.0070/I The ingestion operation shall compute pixel measurement time.<br />

SR.ING.0080/I The time of the earliest SST measurement within this data set should be<br />

coded as a continuous time coordinate specified from 00:00:00 UTC January 1, 1981<br />

(start of the useful AVHRR SST data record).<br />

SR.ING.0090/I The time of SST measurement should be coded as the deviation in<br />

minutes from the time of the earliest SST measurement.<br />

SR.ING.0100/I In case of bad data file integrity, a dialog with the data provider shall be<br />

initiated, by sending error message, to establish a solution as soon as possible.<br />

SR.ING.0110/I Ingestion must evaluate the timeliness of each data file.<br />

SR.ING.0120/I The MMR_DSD records contains the data provider contact details and<br />

point of contact for error reports.<br />

SR.ING.0130/I L2 data will be extracted over the EURDAC and MEDSPIRATION UHR<br />

areas, as appropriate.<br />

3.2.3 Quality Check<br />

SR.QLC.0010/I Generation of L2P shall determine if SST data are suitable for further use<br />

in the GDS. If not inform the data provider and generate an error message for the<br />

GHRSST-PP ERRLOG explaining why these data have not been used in the GDS.<br />

SR.QLC.0020/I A configuration file, that will be revised as required, is used to store all<br />

QC parameter limits and thresholds (see SR.EXT.ING.0040).<br />

© 2004 SOC Page 32 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.QLC.0030/I For each data stream controlled, a rejection rate shall be compute.<br />

SR.QLC.0035/I Rejection rate above a threshold will lead to rejection of the file.<br />

SR.QLC.0040/I A data rejected shall be replaced by a missing value depending on the<br />

data format.<br />

SR.QLC.0045/I Each input data file is assessed pixel by pixel, using gross error checking<br />

rules.<br />

SR.QLC.0050/I If an input SST data stream includes information that indicates that a<br />

pixel value is cosmetic rather than measured, these data should be rejected from the L2P<br />

data file.<br />

SR.QLC.0060/I If an input SST pixel data value or associated confidence data indicate<br />

the pixel to be cloudy, these data should be rejected from the L2P data file.<br />

SR.QLC.0065/I If a (microwave) measurement is identified as rain contaminated in the<br />

native data input file, this measurement should not be included in a L2P data file.<br />

SR.QLC.0070/I A valid SST observation must lie within a temperature range of<br />

LowSSTLimit < SST < HighSSTLimit. If the pixel SST value is out of these limits, the<br />

observation must be rejected from the final L2P data file.<br />

SR.QLC.0080/I The difference DT_min between the pixel SST value (SSTobs) and the<br />

reference SST field (derived as a 10 day coldest climatology from Pathfinder SST data<br />

products) shall be computed and inserted in the L2P confidence data record.<br />

SR.QLC.0090/I A valid wind speed observation must lie within a wind speed range of 0 <<br />

u < HighWindSpeed. Wind speed values that fall outside this range are considered<br />

erroneous and should not be used within the GDS.<br />

SR.QLC.0100/I A valid 6 hourly integrated SSI observation/forecast must lie within a SSI<br />

range of 0 < SSI < HighSSIValue. SSI values that fall outside this range should not be<br />

used within the GDS.<br />

SR.QLC.0110/I A valid AOD observation/estimate must lie within a range of<br />

LowAODLimit < u < HighAODLimit. AOD values that fall outside this range are<br />

considered erroneous and should not be used within the GDS.<br />

SR.QLC.0120/I LowSSTLimit, HighSSTLimit, HighWindSpeed, HighSSIValue,<br />

© 2004 SOC Page 33 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

LowAODLimit, HighAODLimit and the corresponding maximum rejection rates will be set<br />

in a configuration file.<br />

3.2.4 Merging auxiliary data with SST<br />

SR.MRG.0010/I Generation of L2P shall extract auxiliary data (wind speed, surface solar<br />

irradiance and aerosol optical depth) and collocate with SST data.<br />

SR.MRG.0020/I The merging of auxiliary data (L2P-PP) must consist of a common preprocessing<br />

pixel based operation for Infra-red and microwave data.<br />

SR.MRG.0030/I The merging operation shall assign a fractional sea ice value if required<br />

(if no previous fractional sea ice exist in the data file).<br />

SR.MRG.0040/I The merging operation shall assign a wind speed value.<br />

SR.MRG.0050/I The merging operation shall assign a surface solar irradiance (SSI)<br />

value.<br />

SR.MRG.0060/I The merging operation shall assign an Atmospheric Optical Depth (AOD)<br />

value if required (if no previous AOD exist in the data file).<br />

SR.MRG.0070/I If an input data set pixel fractional sea ice estimate exists, this should be<br />

used to set the FractionalSeaIce flag of the L2P confidence data record.<br />

SR.MRG.0080/I If an input data set pixel sea ice flag exists (i.e. indicating sea ice but not<br />

the fractional amount of coverage), this should be used to set the FractionalSeaIce flag of<br />

the L2P confidence data record to 100%. If the flag is a binary flag, a check should be<br />

made against the reference fractional sea Ice data set to determine if the pixel is likely to<br />

be 100% contaminated and the fraction reported by the reference data set should be set<br />

if it is < 100%.<br />

SR.MRG.0090/I If an input data set pixel sea ice flag does not exist, and the pixel is<br />

located in or close to a region that may be ice contaminated, the reference sea ice data<br />

sets (NISE) or auxiliary (AMSRE-ICE) or NWP (NWP-ICE) fractional sea ice field should<br />

be used to determine the value of the L2P confidence flag FractionalSeaIce.<br />

SR.MRG.0110/I The L2P confidence variable FractionalSeaIce_source shall be set to<br />

indicate the data source used to set FractionalSeaIce. The code values are defined in<br />

[RD-3] Table 2.1.3.<br />

© 2004 SOC Page 34 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.MRG.0120/I If the SST input data set is derived from a microwave sensor and a sea<br />

ice flag is present in the data stream, bit 4 of the L2P confidence data variable<br />

MW_SST_flags should be set for this pixel.<br />

SR.MRG.0130/I A 10m surface wind speed value should be assigned to each SST<br />

measurement pixel. As a general rule for the GDS v1.0, only simultaneous wind speed<br />

measurements should be used otherwise NWP analyses should be used.<br />

SR.MRG.0150/I In the case of microwave SST measurements, the use of a simultaneous<br />

microwave 10m wind speed measurements obtained from the same instrument providing<br />

the SST measurement may be used.<br />

SR.MRG.0160/I In the absence of a simultaneous surface wind speed measurement, an<br />

NWP estimated 10m surface wind speed should be used to set the L2P confidence data<br />

variable Wspd_value.<br />

SR.MRG.0170/I The difference in time expressed in hours between the time of SST<br />

measurement and the time of wind speed measurement should be provided.<br />

SR.MRG.0180/I A 6 hourly integrated downwelling SSI measurement (e.g., derived from<br />

satellite measurements) should be assigned to each SST pixel. The SSI measurement<br />

nearest in space and time before the input pixel SST value should be used.<br />

SR.MRG.0190/I If no SSI measurement is available, a 6 hourly integrated SSI value<br />

derived from an NWP system (See Appendix A3) nearest in space and time to the SST<br />

measurement should be used instead.<br />

SR.MRG.0200/I The source of data used to set the L2P confidence data variable should<br />

be provided. Code values are defined in [RD-3] Table 2.1.5.<br />

SR.MRG.0210/I The difference in time expressed in hours between the time of SST<br />

measurement and the mean time of SSI measurement/data validity should be provided.<br />

SR.MRG.0220/I A L2P AOD measurement (e.g., derived from satellite measurements)<br />

should be assigned to each SST pixel value using the AOD_value L2P confidence data<br />

variable. The AOD measurement nearest in space and time to the input pixel SST value<br />

should be used.<br />

SR.MRG.0230/I If no AOD measurement is available, an AOD value derived from an<br />

NWP system or aerosol forecast nearest in space and time to the SST measurement<br />

should be used instead.<br />

© 2004 SOC Page 35 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.MRG.0240/I The source of data used to set the L2P confidence data variable should<br />

be provided. Code values are defined in [RD-3] Table 2.1.6.<br />

SR.MRG.0250/I The difference in time expressed in hours between the time of SST<br />

measurement and the time of AOD observations/analysis should be provided.<br />

SR.MRG.0260/I If no AOD measurement is available, an AOD value derived from NWP<br />

nearest in space and time to the SST measurement should be used.<br />

Comment: There is no AOD estimates in NWP outputs for the time being. When NAAPS<br />

are available, it is suggested they are given priority over the AVHRR-AOD.<br />

3.2.5 Confidence value derivation<br />

SR.CFD.0010/I A “confidence value”, having a value spanning 0 (no data, measurement<br />

is sea ice) to 5 (highest confidence in the SST measurement), is derived for each pixel.<br />

Comment: In GDS, figure 2.2.1.4.1 is not consistent with table 2.2.1.4. A confidence<br />

value should identify a cloudy case. That would enlarge the range of confidence values<br />

from 0 to 6.<br />

SR.CFD.0020/I The confidence value is to be used when assigning appropriate bias and<br />

rms error estimate to each measurement.<br />

SR.CFD.0030/I Confidence data are extracted from native SST data streams if present<br />

and they are derived using auxiliary and reference data streams and sensor specific error<br />

statistics.<br />

Comment: We may adopt native bias and sdtdev, but not confidence value. There is a<br />

risk of inconsistency. It is suggested to admit also native confidence values (flagged as<br />

such, by a L2_native_confidence_value flag).<br />

SR.CFD.0050/I There shall be separate schemes to derive Proximity_Confidence values<br />

for infrared and microwave SST data streams.<br />

© 2004 SOC Page 36 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

3.2.5.1 Microwave sensor case<br />

MED-SOC-RS-001_1<br />

Issue F<br />

MW<br />

SST<br />

(Input)<br />

contaminati<br />

on map<br />

sidelobe<br />

contaminated<br />

y<br />

raise flag<br />

raise flag<br />

y<br />

rain<br />

contaminated<br />

TSanal<br />

TSanal-SST ><br />

threshold <br />

n<br />

TMISST 12ms-1<br />

y<br />

raise flag<br />

LANDMSK<br />

calculate distance to<br />

rain (Drain), ice (Dice)<br />

and land (Dland)<br />

Use wsp , TMI flag<br />

and<br />

min(Drain,Dice,Dland)<br />

to determine MPCV<br />

MW SST<br />

(output)<br />

Figure 3-2 Schematic of procedure to set the microwave proximity confidence<br />

value (MPCV).<br />

As Figure 3-2 shows, several different pieces of information are needed to perform the<br />

tests leading to evaluate the MPCV which is to be added to the data quality information in<br />

the L2P products derived from microwave sensors. This is the basis for the following<br />

system requirements.<br />

© 2004 SOC Page 37 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.CFD.0100/I <strong>Data</strong> in close proximity to rain, land, sea ice and at high wind speeds are<br />

flagged as suspect.<br />

SR.CFD.0110/I If a pixel is located within an area of known side lobe contamination as<br />

defined by sensor specific side lobe contamination maps, bit 0 of the L2P confidence<br />

data variable MW_SST_flags should be set.<br />

SR.CFD.0120/I If a side lobe contamination map is not available, any pixel within<br />

sidelobe_distance_threshold (specified in the L2P configuration file in km) from land<br />

should flagged as contaminated.<br />

Comment: A contamination map will be derived from the LPDAAC/LANDMASK at the<br />

mw SST field resolution till the appropriate map is made available by GHRSST.<br />

SR.CFD.0140/I If a pixel lies within a radius of relaxed_MW_rain_distance specified in<br />

km of a clearly flagged rain contaminated pixel, bit 1 of the L2P confidence data variable<br />

MW_SST_flags should be set.<br />

SR.CFD.0150/I If the difference between the most recent GHRSST-PP analysis SST and<br />

the potentially contaminated pixel is greater than relaxed_rain_threshold (K), the pixel<br />

should not be included in a L2P data file.<br />

Comment 1: A most recent analysis map will be derived from the CMS (Faugère eet al.)<br />

climatology until available from GHRSST.<br />

Comment 2: sidelobe_distance_threshold, relaxed_MW_rain_distance and<br />

relaxed_rain_threshold should be included in the configuration file.<br />

SR.CFD.0160/I If a TRMM TMI SST measurement value is less than 285K, bit 2 of the<br />

L2P confidence data variable MW_SST_flags should be set<br />

SR.CFD.0170/I If a microwave wind speed measurement contemporaneous with a<br />

microwave SST measurement is greater than 12 ms -1 , the wind flag which is the bit 3 of<br />

the L2P confidence data variable MW_SST_flags should be set.<br />

SR.CFD.0180/I If a contemporaneous microwave wind speed measurement is<br />

unavailable, the nearest NWP analysis in space and time to the SST measurement<br />

should be used to evaluate the wind flag defined above.<br />

SR.CFD.0190/I The distance (Dice) to the nearest Sea ice shall be computed.<br />

© 2004 SOC Page 38 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.CFD.0200/I The distance to the nearest problem (Dpb=min(Dcont, Dice, Drain)) shall<br />

be computed<br />

SR.CFD.0210/I Derive MPCV value from Dpb and wspd on the scale shown in Table 3-1,<br />

using thresholds MPCV_wind1, MPCV_wind2 and MPCV_proximity.<br />

Table 3-1 Microwave proximity confidence value (MPCV) scale definitions.<br />

MPCV Value<br />

Description<br />

10 Unprocessed: <strong>Data</strong> that have not been classified (measurement indicates sea<br />

ice)<br />

11 Questionable: <strong>Data</strong> that are may be contaminated by land, rain, sea ice, RF<br />

interference, degraded due to uncertainty in seawater emissivity at higher<br />

wind speeds, and/or are below a low SST threshold.<br />

12 Acceptable: <strong>Data</strong> that are far from land, rain, sea ice, RF interference and are<br />

within a favourable wind speed regime.<br />

13 Diurnal: <strong>Data</strong> are far from any rain or land flags but are within a low wind<br />

speed regime where a cool skin or warm layer may influence the observed<br />

SST.<br />

SR.CFD.0220/I The threshold values MPVC_wind1, MPVC_wind2 and MPVC_proximity<br />

are in a configuration file.<br />

3.2.5.2 Infrared case<br />

SR.CFD.0300/I If an indication of sun-glint is provided with an input SST pixel value, this<br />

should be used to set the SunGlint flag of the corresponding L2P confidence data record.<br />

SR.CFD.0310/I If an indication of sun-glint is not provided with an input pixel value, an<br />

appropriate method should be used to set the SunGlint flag of the L2P confidence data<br />

record that is based on the geometry of sun and satellite position and surface roughness<br />

(e.g., Gardashov and Barla, 2001)<br />

© 2004 SOC Page 39 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

IR SST<br />

(Input)<br />

sunglint<br />

raise flag<br />

calculate distance to<br />

cloud (Dcloud)<br />

Use DT_min and<br />

Dcloud to calculate<br />

IPCV<br />

IPCV=6<br />

n<br />

upwelling<br />

map<br />

y<br />

Upwelling<br />

y<br />

n<br />

assign<br />

IPCV to 1<br />

(cloudy)<br />

IR SST<br />

(output)<br />

Figure 3-3 Procedure for determining infrared proximity confidence value (IPCV)<br />

The proximity confidence value for SST measurements derived from infrared sensors<br />

(IPCV) flags the likelihood of cloud contamination of pixels which have already passed<br />

the cloud screening of the system which generated the ingested L2 products. It is based<br />

(see Figure 3-3) on the proximity to known cloudy pixels, and on comparison with<br />

expected (reference) SST values, but also needs information about the likelihood of<br />

upwelling. The process of deriving it gives rise to the following system requirements.<br />

© 2004 SOC Page 40 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.CFD.0320/I Infrared proximity confidence value (IPCV) must be determined on the<br />

scale given in Table 3-2.<br />

Comment: In GDS, figure 2.2.1.4.1 is not consistent with table 2.2.1.4. A confidence<br />

value should identify a cloudy case.<br />

Table 3-2 IPCV value definitions<br />

IPCV Value<br />

0<br />

Description<br />

Unprocessed: <strong>Data</strong> that have not been classified (measurement indicates sea<br />

ice)<br />

1 Erroneous cloudy.<br />

2 Bad: <strong>Data</strong> that are probably contaminated by cloud.<br />

3 Suspect: <strong>Data</strong> that may be contaminated by cloud.<br />

4<br />

5<br />

6<br />

Acceptable: <strong>Data</strong> that are reasonably distant from cloudy areas and in good<br />

agreement with the expected reference SST threshold.<br />

Excellent: <strong>Data</strong> that are far from any cloudy areas and in good agreement with<br />

the expected reference SST threshold.<br />

Cool skin: <strong>Data</strong> are far from any clouds but significantly cooler than the<br />

expected reference SST threshold. This could be isolated cloud, upwelling or<br />

other dynamical feature or a strong cool skin deviation.<br />

SR.CFD.0330/I The system shall calculate the distance to the nearest cloudy pixel<br />

Dcloud<br />

SR.CFD.0340/I Use Dcloud and DT_min to determine Infrared Proximity Confidence<br />

Values IPCV. Two threshold values are set by the GDS, IPCV_D1 and IPCV_D2, for<br />

each L2P data stream. The distance IPCV_D2 is set on the assumption that cloudy pixels<br />

beyond this distance do not influence the SST measurement. The distance IPCV_D1 is<br />

set assuming that cloudy pixels where IPCV_D1 < nearest_cloudy_pixel < IPCV_D2 have<br />

a reasonable probability of being cloud contaminated. Pixels within a distance less than<br />

IPCV_D1 from the nearest cloud flagged pixels are assumed to have a high probability of<br />

cloud contamination.<br />

SR.CFD.0350/I The second IPCV test uses the deviation from a minimum SST<br />

climatological value defined in the L2P confidence data record variable DT_min. Three<br />

© 2004 SOC Page 41 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

threshold values for DT_min will be set by the GDS L2P configuration file IPCVThresh1 -<br />

IPCVThresh3.<br />

SR.CFD.0360/I A static map of upwelling regions may be used to decide if a pixel is<br />

cloud contaminated in the case of IPCV value 6. Maps will be developed by RDAC as<br />

appropriate for each region of interest.<br />

Comment: No upwelling maps exists. Upwelling maps EUR/ATLUPW and<br />

EUR/MEDUPW): will be derived from the Landmask (LPDAAC/LANDMASK) at 10 km<br />

resolution over the Atlantic and 2 km resolution over the mediterranean. A coastal area<br />

will be defined as upwelling_distance (km) away from coastline, upwelling_distance<br />

should be distinct in the Atlantic and in the Mediterranean .<br />

SR.CFD.0370/I IR only: The values IPCV_D1, IPCV_D2, IPCVThres1, IPCVThres2,<br />

IPCVThres3 used in the above procedure are expected to vary for each sensor. These<br />

values are maintained by GHRSST-PP and stored in the L2P configuration file. (see<br />

SR.EXT.ING.0040)<br />

Comment: This implies the use of tables of values.<br />

3.2.6 SSES Assignation<br />

SR.L2P.6010/I If pixel bias error and standard deviation error statistics are provided by a<br />

data provider, these values may be assigned to the L2P bias and standard deviation<br />

variables respectively. In this case, the L2_native_bias and L2_native_sd flags of the<br />

corresponding L2P confidence data record should also be set as these error estimates<br />

are independent of GHRSST-PP methods.<br />

SR.L2P.6020/I If pixel bias error and standard deviation error statistics are not available<br />

for SST measurements, these values should be assigned the most recent SSES bias and<br />

sd values for each data stream and L2P confidence data record Proximity_Confidence<br />

value. In addition, the L2_native_bias and L2_native_sd. flags of the corresponding L2P<br />

confidence data record should be set to zero.<br />

3.2.7 L2P evaluation<br />

SR.L2P.7000 The L2P data will be assembled from the L2 SST, the ancillary data, the<br />

mean and sd, and the confidence values. Only those datasets with quality above a given<br />

threshold will be accepted. Files will be written in netCDF format.<br />

SR.L2P.7010/I The L2P evaluation shall build a report about the L2P data files<br />

© 2004 SOC Page 42 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.L2P.7020/I The L2P evaluation report shall contain the rejection rates of the data<br />

sources used<br />

SR.L2P.7030/I The L2P evaluation report shall contain status of the confidence values<br />

calculation<br />

SR.L2P.7040/I The L2P evaluation report shall contain the list of the missing parameters<br />

3.3 Ifremer Facility<br />

EU<br />

GHRSST-<br />

L2P, UHR/L4<br />

L2P<br />

dissemination<br />

MDB records,<br />

MMR-FR<br />

L2P producer<br />

(CMS)<br />

L2P<br />

ancillary data<br />

L2P, UHR/L4<br />

<strong>System</strong><br />

controller<br />

HR-DDS<br />

producer(SOC)<br />

L2P<br />

archiving<br />

UHR/L4<br />

L2P<br />

L2P<br />

ancillary data<br />

MDB records<br />

L4<br />

generation<br />

MDB records<br />

generation<br />

In situ<br />

In situ provider<br />

(CORIOLIS)<br />

Figure 3-4 Outline of the <strong>Medspiration</strong> processing structure at IFREMER<br />

© 2004 SOC Page 43 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

3.3.1 Archive<br />

MED-SOC-RS-001_1<br />

Issue F<br />

archiving<br />

Reference data Online storage Offline storage backup<br />

Figure 3-5 The <strong>Medspiration</strong> archiving function at IFREMER<br />

SR.ARC.0010/I The archiving system shall ingest all valid L2P (received from CMS) and<br />

UHR/L4 products (generated at IFREMER). <strong>Data</strong> rejected at some stage of their<br />

processing or of their control shall not be archived.<br />

SR.ARC.0020/I Ancillary surface solar irradiance and wind speed fields (received from<br />

CMS) may also be archived if needed to compute the SSTskin to be included in the<br />

L4/UHR product.<br />

SR.ARC.0030/I L2P or UHR/L4 products received by the archiving system beyond the<br />

delivery time limits defined by the GHRSST-PP shall be registered anyway for users not<br />

constrained by real time needs.<br />

SR.ARC.0040/I L2P or UHR/L4 products shall be ingested in the archiving system as<br />

soon as they are generated by the related production system (or received, in case of a<br />

remote production system).<br />

3.3.1.1 Reference data<br />

SR.ARC.0110/I Each L2P and L4 data file ingested in the archiving system shall be<br />

referenced in a database (inventory).<br />

SR.ARC.0120/I Spatio-temporal characteristics of each <strong>Medspiration</strong> product registered<br />

to the archiving system shall be extracted and stored within a database to facilitate and<br />

make faster product search and restoration.<br />

3.3.1.2 Online storage<br />

SR.ARC.0210/I The archiving system shall provide online storage for the most recent<br />

© 2004 SOC Page 44 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

data. These storage capability shall be fully scalable depending on the available<br />

hardware and the occupation rate of each storage device.<br />

3.3.1.3 Offline storage<br />

SR.ARC.0310/I The archiving system shall provide an offline storage facility for historical<br />

data (data files removed from the online storage space). This offline storage shall rely on<br />

the mass storage facility available at IFREMER.<br />

SR.ARC.0320/T The archiving system shall move data files from the online storage<br />

space to the offline storage space on a selective basis, for instance when the occupation<br />

rate of the online storage space is getting critical. This operation shall be initiated by a<br />

human operator as a maintenance task.<br />

SR.ARC.0330/T The archiving system shall be able to restore any offline data to the<br />

online storage space in reasonable delays (less than a few hours).<br />

3.3.1.4 Backup<br />

SR.ARC.0410/I Changes in the online storage space (new or updated data files) shall be<br />

saved every night using the IFREMER backup facility.<br />

3.3.2 Generate L4 analysed products<br />

Select/collate L2P observations<br />

SSTskin or SSTsubskin<br />

Diurnal cycle estimation (DV)<br />

SSI<br />

Wind speed<br />

SSTfnd<br />

Optimal interpolation (OI)<br />

Analysed SSTfnd<br />

L4 products generation<br />

SSTfnd<br />

SSTskin<br />

Figure 3-6. The four functional sub-modules of the L4 processor<br />

© 2004 SOC Page 45 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

SR.L4A.0010/I The UHR/L4 product generation subsystem shall be operated only over<br />

the Mediterranean sea during the two years of <strong>Medspiration</strong>.<br />

SR.L4A.0015/I The L4 module shall be decomposed into 4 sub-modules: L2P<br />

collation/selection ; diurnal cycle estimation ; optimal interpolation ; products generation<br />

SR.L4A.0020/I The UHR/L4 product generation shall be triggered daily at the cut-off<br />

time (defined as day T + 1 at 6:00 UTC). A latency delay may be added to complete<br />

ongoing L2P data uploads from CMS to IFREMER.<br />

SR.L4A.0025/I The computation time necessary to generate the L4 UHR SST fnd shall be<br />

6 hours maximum.<br />

SR.L4A.0030/I The UHR/L4 product generation system shall take as inputs to the OI<br />

scheme all L2P data file covering day T available at the cut-off time, a L4 data processor<br />

configuration file and reference data files (ancillary data)<br />

SR.L4A.0035/I The configuration file will contain flags to activate the estimation of diurnal<br />

cycle, the parameters of the correlation function, parameters that characterise the<br />

analysis (time, grid, sensors selected, sensors bias and errors coming from SSES )<br />

SR.L4A.0040/I The UHR/L4 product generation system shall take as additional inputs to<br />

the diurnal cycle computation ancillary surface solar irradiance and wind speed fields<br />

covering day T.<br />

SR.L4A.0050/I The UHR/L4 product grid shall be that specified in Table 2.4.7.3 of [RD-<br />

3]<br />

SR.L4A.0051/I The L4 processor shall provide error estimates for each analysed grid<br />

cell,<br />

SR.L4A.0052/I The L4 processor shall account for differences in spatial and temporal<br />

sampling characteristics of each data stream,<br />

SR.L4A.0053/I The L4 processor shall account for gaps in coverage due to the presence<br />

of cloud, rain or lack of data,<br />

SR.L4A.0054/I The L4 processor shall account for SST diurnal variability and retrieve an<br />

estimate of the subsurface temperature field (SSTfnd),<br />

© 2004 SOC Page 46 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.L4A.0055/I The L4 processor shall provide a measure of diurnal variability of the skin<br />

deviation from SSTfnd within the data product time domain to accompany the SSTfnd<br />

estimate<br />

SR.L4A.0060/I The L2P data shall be selected/collated before the OI analysis<br />

The rules proposed for the collation of L2P data have to be more defined because they<br />

are completely correlated to the OI method. In particular, the rules proposed to collate the<br />

L2P data induce an averaging of them that is inconsistent with the formalism of the<br />

interpolation method. Consequently, the only collation possible will be into a 0.02° grid<br />

without any averaging. However, in most optimal analysis systems, the analysis itself will<br />

optimise the re-gridding of data although re-gridding prior to interpolation provides a more<br />

straightforward processing system (p 85 of the GDS, line410 of UR-IFREMER ).<br />

SR.L4A.0070/I The activation of the DV depends on the validation of the DV<br />

parameterisation in the Mediterranean sea that will be undertaken by the CNR team<br />

CNR will perform a study to verify the accuracy of the A Stuart-Menteth DV<br />

parameterisation in the Mediterranean sea. If results are satisfying, we will activate the<br />

DV parameterisation through the configuration file.<br />

SR.L4A.0080/I The L4A interpolation sub-module shall be based on the optimal<br />

interpolation scheme proposed by the GHRSST-PP (Reynolds et al., 2002, Guan and<br />

Kawamura, 2003)<br />

SR.L4A.0090/I OI scheme is expected to evolve as more experience is gained, therefore<br />

the L4 analysis system will be built in a modular manner<br />

SR.L4A.0095/I The regional parameters used by the OI scheme (correlation functions,<br />

spatial and temporal correlation radius) will come from the correlation study led by the<br />

CNR team in the frame of the <strong>Medspiration</strong> project.<br />

SR.L4A.0100/I The sub-module “products generation” will generate L4 SST fnd that<br />

contains SST fnd , 12 values of SST skin and associated errors. The SST skin values will be<br />

filled if the DV parameterisation is activated.<br />

© 2004 SOC Page 47 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

3.3.3 Matchup <strong>Data</strong> Base<br />

MED-SOC-RS-001_1<br />

Issue F<br />

MDB records<br />

generation<br />

Collect in situ<br />

data<br />

Colocate in situ<br />

with satellite<br />

data<br />

Generate MDB<br />

records<br />

Figure 3-7 The functions of the MDB records generation module at IFREMER<br />

SR.MDB.0010/I The MDB records generation shall be triggered on a periodic basis<br />

related to the delivery frequency of in situ data. They should be produced within the 24<br />

hours from when the in situ data are made available by the in-situ data provider.<br />

SR.MDB.0020/I The MDB records generation system shall take both in situ and satellite<br />

data as inputs.<br />

SR.MDB.0030/I The MDB records generation system shall take as input the relevant in<br />

situ data collected by the in situ data provider since their last delivery. It is admitted that<br />

due to delivery and quality control constraints, a minimum delay of 24 hours from<br />

acquisition is to be expected for in-situ data availability. However, all MDB records are<br />

admissible irrespective of their timeliness for use by the GHRSST-PP reanalysis project.<br />

Therefore all incoming in-situ data shall be ingested in the MDB records production<br />

system regardless of their age, in the limit of one month.<br />

This limitation of one month is introduced to avoid using historical L2P data which may<br />

not be available online anymore. Using offline data would add heavy and time consuming<br />

operations with respect of the expected amount of involved in situ data.<br />

SR.MDB.0040/I The MDB records generation system shall take as satellite input all the<br />

latest L2P products archived in the limit of one month of data.<br />

See comment above.<br />

SR.MDB.0050/I The MDB records shall contain in situ SST (and other associated oceanatmosphere<br />

measurements) that are matched to L2P satellite SST measurements<br />

according to specified time and space constraints (referred to as colocation criteria).<br />

© 2004 SOC Page 48 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

No MDB records shall be generated for the UHR/L4 products since they are not referred<br />

to as GHRSST-PP products and thus shall not appear in the MDB.<br />

3.3.3.1 Collecting the in-situ data<br />

SR.MDB.0110/I The main source of in situ observations used in this system will be those<br />

available from the GTS supplied through international in situ data centres (e.g., MEDS,<br />

CORIOLIS, FNMOC) at which QC procedures are used to filter out poor quality data.<br />

In the frame of MEDSPIRATION, the CORIOLIS data center will be considered as the<br />

only source of in-situ data. However, CORIOLIS is a gateway to GTS data and MEDS<br />

data.<br />

SR.MDB.0120/I The MDB shall be open to all other source of in situ observations that are<br />

deemed of sufficient quality by the <strong>Medspiration</strong> team.<br />

Taking advantage that the CORIOLIS data centre is located, funded and operated by<br />

IFREMER, the MEDSPIRATION team will rely on the CORIOLIS team for ingesting new<br />

in-situ data streams within its easily extendable database, before delivering them to the<br />

MDB records production system. Keeping a single interface for in-situ data will make it<br />

easier to maintain and expend the system. In its first month of operation, <strong>Medspiration</strong><br />

may for instance benefit from the in situ data collected by MFSTEP that will be ingested<br />

in CORIOLIS database.<br />

SR.MDB.0130/I All in-situ data shall be quality controlled before being ingested within the<br />

MDB records production system, using local ocean/atmosphere knowledge and<br />

experience. The MDB records production system will use only quality-controlled data<br />

(uncontrolled data will be rejected). The <strong>Medspiration</strong> system will rely on the CORIOLIS<br />

data centre for quality control procedures.<br />

3.3.3.2 Colocating the in-situ and satellite data<br />

SR.MDB.0210/I The satellite pixel having the smallest deviation in both space and time<br />

from an in situ observation within the colocation criteria should be selected as a MDB<br />

record.<br />

It is not clearly defined within the GDS which in-situ observation should be selected for<br />

the match-up record when several observations match the space and time criteria. It is<br />

proposed to select the closest one in space (which means that the space criterion has<br />

priority over the time criterion) or in time if their distance to the satellite cell centre is the<br />

© 2004 SOC Page 49 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

same.<br />

MED-SOC-RS-001_1<br />

Issue F<br />

SR.MDB.0220/I The colocation method will be the same for each satellite product<br />

regardless of their resolution.<br />

It is stated in the GDS that, for coarser resolution (typically 25 km), the mean of a<br />

transect of several measurements obtained within the grid cell will be preferred to a<br />

single point observation. This solution introduces a lot of complexity because of the<br />

ancillary parameters provided with the in-situ measurements (for instance ship speed or<br />

heading, location, sensor identifiers) which can’t be averaged and because in-situ<br />

observations may come from different sensors and platforms. In addition, a mean<br />

observation could be computed from both day-time or night-time measurements. It is<br />

proposed to use the same algorithm as for high resolution measurements.<br />

SR.MDB.0230/I The time and space constraints are stored in a configuration file<br />

(MDB_time_constraint and MDB_space_contraint variables) for each L2P dataset.<br />

These colocation criteria shall be defined and maintained by the GHRSST-PP<br />

International Project Office.<br />

SR.MDB.0240/I The space/time criteria shall be set as the largest as acceptable<br />

(typically 25km in space and 6 hours in time). These criteria may be strengthened when<br />

filtering out the data from the MDB at GDAC, for instance to take into account the<br />

day/night asymmetry, but not when generating the MDB records at RDAC in order to add<br />

more flexibility and extendibility.<br />

It is stated in the GDS that “because of the day-night asymmetry in the growth and decay<br />

of the diurnal thermocline, more stringent conditions for acceptable time intervals may be<br />

required during the day than at night. For these and ambiguous cases, such as close to<br />

the terminator, the smallest spatial separation and shortest possible time interval should<br />

be sought”. It is not clear if this should be taken into account for the MDB records<br />

generation or when retrieving the data records from the database, which allows more<br />

flexibility (easier to generate the most complete MDB dataset and then filter out when<br />

extracting the data). Moreover the definition and the computation of the day/night<br />

threshold need to be provided in order to ensure consistency through RDACs.<br />

SR.MDB.0250/I An in-situ measurement shall be colocated to only one pixel of each L2P<br />

dataset.<br />

This is a refinement from the GDS which states that only one instance of each in situ<br />

observation for each individual satellite sensor should occur within the MDB which is<br />

© 2004 SOC Page 50 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

ambiguous since some sensor measurements may exist in different datasets with various<br />

resolutions (eg AVHRR).<br />

SR.MDB.0260/I A satellite pixel shall be colocated to only one in situ measurement from<br />

a single in situ data file. However, the scope of the colocation process for in situ data is<br />

limited to the data files : a satellite pixel may be colocated again with another in situ<br />

measurement delivered by CORIOLIS a few days or weeks later in another datafile<br />

There is no user requirements addressing this issue. Trying to avoid a satellite pixel<br />

being colocated with more than one in situ measurement would be complex to implement<br />

while addressing very rare cases. In particular it would almost imply building a MDB to<br />

filter out these cases which is not the aim of this system. If this issue is considered as<br />

critical, it should be adressed by the GDAC MDB when ingesting its records.<br />

SR.MDB.0270/I The targeted in situ measurement for the MDB records production are<br />

the SST1m data derived from drifting/moored buoy, ship and float (Argo) observations.<br />

For some sensors, the SST1m may not be accessible : it is thus proposed to catch the<br />

closest to the 1m depth in the limits of [-5m, 0m].<br />

It is stated in the GDS that the targeted in situ measurement for the MDB records<br />

production are the SST1m data derived from drifting/moored buoy and ship observations.<br />

After meeting with the CORIOLIS staff, it appears this may be too restrictive since very<br />

few ship observations are available beyond –5m. Relaxing this constraint allows to select<br />

more data and also to capture the floats data.<br />

3.3.3.3 Generating the MDB records<br />

SR.MDB.0310/I All L2P confidence data each pixel stored in a MDB record must<br />

accompany the extracted data.<br />

SR.MDB.0320/I For each L2P pixel matched to an in-situ measurement, a n x n array of<br />

L2P data shall be extracted centred on this pixel and provided within the related MDB<br />

record.<br />

SR.MDB.0330/I All in situ parameters available in addition to the SST measurement<br />

should be included in the MDB record.<br />

Upgrade of the MDB format (DTD) may be necessary. Additional parameters will be<br />

submitted to GHRSST-PP through ESA in order to keep consistency (variable names,<br />

units,…).<br />

© 2004 SOC Page 51 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

3.3.4 Dissemination<br />

MED-SOC-RS-001_1<br />

Issue F<br />

dissemination<br />

ftp push/pull DODS Live Access<br />

Server<br />

Figure 3-8. The dissemination module at IFREMER<br />

SR.DIS.0010/I The dissemination service shall provide free access to L2P and L4/UHR<br />

products.<br />

SR.DIS.0020/I The dissemination service shall give access to the L2P and L4/UHR<br />

products as soon as they have been ingested by the archiving system.<br />

SR.DIS.0030/I The dissemination service should aim to minimize the volume of<br />

transferred data, using appropriate means such as compression or subsetting.<br />

3.3.4.1 Ftp push/pull<br />

SR.DIS.0110/I The dissemination system shall provide near real time ftp push service for<br />

L2P and L4/UHR products to registered users (including GHRSST-PP for the L2P).<br />

SR.DIS.0120/I The dissemination system shall provide near real time ftp pull service to<br />

public users for L2P and L4/UHR products.<br />

3.3.4.2 DODS<br />

SR.DIS.0210/I The dissemination service shall provide DODS access to L2P and<br />

UHR/L4 data.<br />

3.3.4.3 Live Access Server<br />

SR.DIS.0310/I The dissemination service shall provide LAS (Live Access Server) access<br />

to UHR/L4 data.<br />

© 2004 SOC Page 52 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

3.3.5 <strong>System</strong> controller at Ifremer<br />

MED-SOC-RS-001_1<br />

Issue F<br />

<strong>System</strong><br />

controller<br />

operation log<br />

system<br />

Control<br />

processing flow<br />

Control<br />

data flow<br />

<strong>System</strong><br />

monitoring<br />

Synchronize<br />

with<br />

GHRSST-PP<br />

Trigger<br />

Chain<br />

Exchange<br />

data<br />

Exchange<br />

MMR-FR<br />

Exchange<br />

error messages<br />

Figure 3-9. The controller for the <strong>Medspiration</strong> processor at IFREMER<br />

3.3.5.1 <strong>Data</strong>flow control<br />

SR.CTR.0110/I The system controller shall ensure that all data and messages are<br />

correctly transmitted and received within the overall system.<br />

The <strong>Medspiration</strong> team recommends that the requirements listed below are applied on<br />

each site receiving or transmitting data. We also suggest these requirements to be<br />

extended at the GHRSST-PP level to ensure overall consistency of the global system.<br />

SR.CTR.0120/I The system shall control the integrity of all binary data received from a<br />

remote site using appropriate error control code (md5 or crc checksum) when available.<br />

An error control code shall be associated with all the L2P files sent from CMS to<br />

IFREMER.<br />

SR.CTR.0130/I The system shall control the validity of each XML formatted message<br />

(Operation or error log entry) received from a remote site. The DTD associated with each<br />

XML file will be used to validate the messages.<br />

This requirement may be extended at GHRSST-PP level to MDB and MMR messages.<br />

SR.CTR.0140/I Retransmission shall be requested to the sender for each rejected<br />

datafile or message on integrity or format control.<br />

SR.CTR.0150/I Acknowledgement shall be transmitted by the receiver to the sender for<br />

data files.<br />

This may be very useful for the sender to clean its sending spooler.<br />

© 2004 SOC Page 53 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.CTR.0160/I Corrupted L2P files by the receiver shall be deleted with all their related<br />

information (error control code) before retransmission is requested.<br />

SR.CTR.0170/I When transmitting files by ftp, the error control code (md5 or crc<br />

checksum, having a fixed size) delivered with a datafile shall be used as a flag by the<br />

sender indicating the transfer is completed. The receiver shall assume the transmission is<br />

completed when reading the related error control code. Consequently, the error control<br />

codes shall be sent after the related data file transmission is completed.<br />

SR.CTR.0180/I The selected on-the-shelf ftp tools used by <strong>Medspiration</strong> shall feature<br />

resume facilities in order to optimize the bandwidth and the delivery time in case of<br />

interrupted transmission.<br />

SR.CTR.0190/I On-the-shelf dissemination systems such as DODS or Live Access<br />

Server are out of the scope of these requirements.<br />

3.3.5.2 Processing flow control<br />

SR.CTR.0210/I The system controller shall trigger, schedule and sequence all the<br />

processes runned on its own site and generate appropriate error and operation<br />

messages.<br />

SR.CTR.0220/I The scope of the processing flow control shall be limited to IFREMER site<br />

processing. Other production sites (CMS, SOC) shall rely on their own local practices<br />

and facilities for the control and the sequencing of their processes.<br />

3.3.5.3 Triggering the processing<br />

SR.CTR.0230/I The system controller shall provide a poll mechanism to trigger the<br />

relevant processing – referred to as ‘data-driven tasks’ - on new incoming data.<br />

<strong>Data</strong>-driven tasks include :<br />

• L2P archiving<br />

• L2P delivery (ftp push) to operational users (GHRSST or EU)<br />

• L4 delivery (ftp push) to operational users (EU)<br />

• MMR generation and delivery<br />

SR.CTR.0240/I The system controller shall provide a scheduling mechanism to trigger<br />

processing on a periodic basis, referred to as ‘time driven’ tasks.<br />

© 2004 SOC Page 54 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Time-driven tasks include :<br />

• UHR/L4 product generation<br />

MED-SOC-RS-001_1<br />

Issue F<br />

• MDB records generation<br />

• Backup operations<br />

SR.CTR.0250/I Some system operation shall be manually triggered by the system<br />

operators.<br />

This addresses mainly the offline storage or restoration of the historical data.<br />

3.3.5.4 Chaining the processes<br />

SR.CTR.0260/I The system controller shall appropriately chain the processes when<br />

sequencing is required.<br />

3.3.5.5 Operation log system<br />

SR.CTR.0310/I The system controller shall centralize, in a system log, all relevant<br />

operations and errors occurring within the <strong>Medspiration</strong> system (SOC, IFREMER, CMS).<br />

SR.CTR.0320/I The log information shall be archived in an appropriate way to extract<br />

relevant operation statistics about the <strong>Medspiration</strong> system operations. These statistics<br />

shall then be used in the monthly operation reports delivered to ESA.<br />

3.3.5.6 <strong>System</strong> monitoring<br />

SR.CTR.0410/I The system controller shall report to the staff in charge of the system<br />

operations all relevant events or error occurring within the <strong>Medspiration</strong> system through<br />

a graphical user interface.<br />

SR.CTR.0420/I The log information shall be easily readable by the technicians in charge<br />

of the system monitoring.<br />

SR.CTR.0430/I Failures requiring human intervention shall be highlighted.<br />

3.3.5.7 Synchronization with GHRSST-PP<br />

SR.CTR.0510/I The system controller shall ensure all synchronization procedures with<br />

© 2004 SOC Page 55 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

the external sites within <strong>Medspiration</strong> or GHRSST-PP systems, GHRSST-PP entities,<br />

master metadata repository and error log system.<br />

SR.CTR.0520/I The IFREMER system controller shall also ensure the control and the<br />

sequencing of all the processes run on its own site. Other production sites (CMS, SOC)<br />

shall rely on their own local practices and facilities for the control and the sequencing of<br />

their processes.<br />

3.3.5.8 Exchange MMR<br />

SR.CTR.0530/I A MMR-FR file shall be delivered to the Master Metadata repository for<br />

each L2P and DDS data registered to the archiving system. This will inform the<br />

GHRSST-PP that the related data file is available. No MMR-FR will be generated nor<br />

delivered for the UHR/L4 products.<br />

This restriction comes from the SoW which does not state the GHRSST-PP as a client of<br />

the <strong>Medspiration</strong> UHR/L4 products. Therefore, delivering related MDB records, MMR-FR<br />

or HR-DDS is considered as meaningless in the frame of the GHRSST-PP.<br />

SR.CTR.0540/I Each archiving site (IFREMER for L2P, SOC for HR-DDS) shall be<br />

responsible for generating and sending the MMR-FR files related to the data files it is in<br />

charge of. The MMR-FR shall be delivered to the MMR as soon as they have been<br />

successfully ingested in the archiving system.<br />

3.3.5.9 Exchange error messages<br />

SR.CTR.0550/I The system controller shall format and deliver all specified error<br />

messages to the GHRSST-PP error log (ERRLOG) system.<br />

A tentative list of relevant messages is detailed in GDS Annex 7. This list is likely to<br />

evolve : as <strong>Medspiration</strong> project progresses, error or operation messages of interest will<br />

be submitted to be included or deprecated. The system shall thus be flexible enough to<br />

accommodate new messages.<br />

SR.CTR.0560/I The system controller shall ingest all GHRSST-PP error messages sent<br />

to <strong>Medspiration</strong>. The ingestion of these messages shall be reflected in the operation log<br />

subsystem and in the monitoring subsystem.<br />

No specific action from RDACs is requested by the GDS when receiving an error<br />

message (only one message specified so far : “MDB data rejected from MDB”).<br />

© 2004 SOC Page 56 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

3.3.5.10 Exchange data<br />

MED-SOC-RS-001_1<br />

Issue F<br />

SR.CTR.0570/I The system shall deliver the L2P products to the GDAC as soon as they<br />

have been archived.<br />

SR.CTR.0580/I The system shall deliver the MDB records to the GDAC MDB as soon as<br />

they have been generated.<br />

3.4 SOC Facility<br />

ftpops.pde.envisat.esa.int<br />

ATS_NR_2P<br />

ATS_NR__2P<br />

Mirror<br />

IFREMER<br />

ATS NR 2P<br />

L2P/L4UHR<br />

GHRSST-<br />

PP<br />

ATS_NR_2P<br />

MMR_FR<br />

GDAC<br />

netCDF<br />

netCDF<br />

Generation<br />

& extraction<br />

netCDF<br />

HR-DDS & Metadata<br />

Generation<br />

CMS<br />

HR-DDS<br />

HR-DDS Dissemination<br />

HR-DDS<br />

HR-DDS<br />

HR-DDS<br />

<strong>System</strong><br />

Controller<br />

ERRLOG<br />

GHRSST-<br />

PP<br />

<strong>User</strong><br />

Figure 3-10. Schematic of the <strong>Medspiration</strong> processing at SOC<br />

SR.DDS.1000/I There shall be a system controller at SOC that monitors and logs all<br />

events within the HR-DDS and AATSR systems<br />

3.4.1 AATSR Ingestion<br />

SR.DDS.1010/I All AATSR ATS_NR__2P products made available by ESA on the server<br />

ftp-ops.pde.envisat.esa.int shall be made securely available via a file transfer protocol at<br />

SOC for the GHRSST-PP.<br />

© 2004 SOC Page 57 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.DDS.1030/I All AATSR ATS_NR__2P products made available to <strong>Medspiration</strong> shall<br />

be converted into GHRSST-PP netCDF format.<br />

SR.DDS.1040/I There shall be an extraction process to select data from the AATSR<br />

netCDF format which corresponds to the Atlantic region.<br />

SR.DDS.1050/I An interpolated geographic tie point, satellite elevation and satellite<br />

azimuth shall be provided for each pixel in the netCDF format AATSR data.<br />

SR.DDS.1060/I Only records of Combined View SST will be included in the netCDF files<br />

from the Distributed Product MDS.<br />

3.4.2 HR-DDS Generation<br />

SR.DDS.0010/I The HR-DDS processor shall consist of an AATSR HR-DDS Processor,<br />

and L2P HR-DDS Processor, and L4UHR HR-DDS Processor and an HR-DDS Metadata<br />

Processor<br />

SR.DDS.0020/I All <strong>Medspiration</strong> netCDF AATSR data elements corresponding to<br />

GHRSST-PP HR-DDS sites (global) shall be resampled to HR-DDS data format and<br />

ingested into the HR-DDS system.<br />

SR.DDS.0030/I All L2P data elements corresponding to a <strong>Medspiration</strong> Area HR-DDS<br />

site shall be resampled to HR-DDS data format by the L2P HR-DDS Processor and<br />

ingested into the HR-DDS system.<br />

SR.DDS.0040/I All UHRL4 data elements corresponding to a <strong>Medspiration</strong><br />

Mediterranean Area HR-DDS site shall be resampled to HR-DDS data format by the<br />

L4UHR HR-DDS Processor and ingested into the HR-DDS system.<br />

SR.DDS.0050/I There shall be an HR-DDS Parser and Metadata Processor to ensure<br />

quality of HR-DDS products.<br />

SR.DDS.0060/I A MMR_DSR XML file shall be generated for each data type and sent to<br />

GDAC.<br />

SR.DDS.0070/I A MMR_FR shall be created for each L2P derived HR-DDS file and sent<br />

to GDAC.<br />

SR.DDS.0090/I The HR-DDS data grid shall be defined as in Table 3-3.<br />

SR.DDS.0091/I The HR-DDS system shall be a satellite only archive.<br />

© 2004 SOC Page 58 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Table 3-3. Definition of the DDS grid<br />

Boundary<br />

Projection<br />

Reference Ellipsoid<br />

Grid Cell Size<br />

Grid Cell Reference Point<br />

Multiple to be defined by GHRSST-PP<br />

Cylindrical / Rectangular Equidistant<br />

WGS84<br />

½ UHR L4 grid size<br />

Centre Point<br />

3.4.3 HR-DDS Archiving<br />

SR.DDS.0110/I There shall be an archiving facility at SOC for the storage of<br />

<strong>Medspiration</strong> HR-DDS data.<br />

SR.DDS.0120/I All <strong>Medspiration</strong> HR-DDS data shall be stored in GHRSST-PP HR-DDS<br />

netCDF format.<br />

3.4.4 HR-DDS Dissemination<br />

SR.DDS.0140/I All <strong>Medspiration</strong> HR-DDS data shall be made publicly available via a<br />

DODS server.<br />

3.4.5 HR-DDS Controller<br />

SR.DDS.0170/I There shall be an HR-DDS system controller.<br />

SR.DDS.0180/I The HR-DDS system controller shall monitor and log all activities of the<br />

AATSR mirroring system and the HR-DDS system.<br />

SR.DDS.0190/I The HR-DDS system controller shall report all ERRLOG activity to GDAC<br />

and <strong>Medspiration</strong> central log at IFREMER.<br />

© 2004 SOC Page 59 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

4. INTERFACE REQUIREMENTS<br />

4.1 Internal<br />

4.1.1 Interfaces within the CMS processing facility<br />

SR.INT.L2P.0010/I<br />

each data stream.<br />

<strong>Data</strong> information file shall contain scheduling information for<br />

SR.INT.L2P.0020/I The scheduler module shall be interfaced to the ingestion<br />

module to order the query of data streams.<br />

SR.INT.L2P.0030/I The ingestion module shall be interfaced with the system controller,<br />

delivering error message.<br />

SR.INT.L2P.0040/I The ingestion module shall be interfaced with the quality check<br />

module to control every data file.<br />

SR.INT.L2P.0050/I A configuration file shall contain all the parameters needed by<br />

the quality check module for each data stream.<br />

SR.INT.L2P.0060/I The quality check module shall be interfaced with the system<br />

controller, delivering error message.<br />

SR.INT.L2P.0070/I The quality check module shall be interfaced with the merging<br />

module to complete necessary fields before the confidence calculation.<br />

SR.INT.L2P.0080/I A configuration file shall contain all the parameters needed by<br />

the confidence calculation module for each data stream.<br />

SR.INT.L2P.0090/I<br />

merging module.<br />

The confidence calculation module shall be interfaced with the<br />

SR.INT.L2P.0100/I The SSES assignation module shall be interface with the<br />

confidence calculation module<br />

4.1.2 Interfaces between L2P production and archiving systems<br />

SR.INT.ARC.0010/I The L2P production system shall be interfaced with the archiving<br />

system, delivering by ftp push and registering the L2P data files to the archive.<br />

© 2004 SOC Page 60 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.INT.ARC.0020/I The L2P products shall be delivered to IFREMER by CMS with<br />

respect to the protocol requirements defined for the dataflow control subsystem of the<br />

system controller (refer to section 1.4.1).<br />

SR.INT.ARC.0030/I The L2P products delivered by CMS shall be formatted as specified<br />

in annex 1 (“L2P product format”).<br />

The L2P files are sent in their final (GHRSST-PP) format. It is assumed by IFREMER that<br />

they are complete and well formatted when successfully received.<br />

4.1.3 Interfaces between archiving and dissemination systems<br />

SR.INT.DIS.0010/I The ftp push/pull system shall be interfaced with the archiving system<br />

to access online L2P and L4 products.<br />

SR.INT.DIS.0020/I The DODS server shall be interfaced with the archiving system to<br />

access online L2P and L4 products. This interface is described in DODS user manual<br />

v1.11 (http://www.unidata.ucar.edu/packages/dods/user/guide-html/).<br />

SR.INT.DIS.0030/I The Live Access Server (LAS) shall be interfaced with the archiving<br />

system to access online L4 products. This interface is LAS documentation<br />

(http://ferret.pmel.noaa.gov/Ferret/LAS/<strong>Document</strong>ation/).<br />

4.1.4 Interfaces between archiving and MDB generation systems<br />

SR.INT.MDB.0010/I The MDB records production system shall be interfaced with the<br />

archiving system, retrieving the L2P data files needed for the generation of the MDB<br />

records.<br />

4.1.5 Interfaces between archiving and L4 generation system<br />

SR.INT.L4A.0010/I The UHR/L4 production system shall be interfaced with the archiving<br />

system, retrieving the L2P data files needed for the generation of the UHR/L4 products<br />

and delivering the generated UHR/L4 products.<br />

SR.INT.L4A.0020/I This data exchange shall aim minimizing useless data copy, using<br />

direct access to data storage disks, taking advantage of the colocation of both systems<br />

on IFREMER local network.<br />

SR.INT.L4A.0030/I It is the responsibility of the archiving system to provide a list of all<br />

L2P and ancillary products eligible for the UHR/L4 product generation, depending on<br />

© 2004 SOC Page 61 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

what is available at the cut-off time.<br />

MED-SOC-RS-001_1<br />

Issue F<br />

SR.INT.L4A.0040/I The UHR/L4 products shall be formatted as netcdf files with respect<br />

to the format provided in Annex 2 (“L4 product format”).<br />

4.1.6 Interfaces within the archiving system<br />

SR.INT.ARC.0510/I The archiving system shall use the existing offline mass storage<br />

subsystem existing at CERSAT to archive the historical data stored online. The interface<br />

of this subsystem is described in ‘<strong>Document</strong> de specification Logiciel de l’Atelier<br />

d’Archivage’ (ref : WCOL-DSL-A-03-CS). This interface may need to be updated to take<br />

into account <strong>Medspiration</strong> specificities.<br />

SR.INT.ARC.0520/I The online storage system of the archiving system shall be<br />

accessible for the offline mass storage system to restore historical data.<br />

SR.INT.ARC.0530/I The archiving system may use the existing referencing subsystem<br />

existing at CERSAT. The interface of this subsystem is described in ‘<strong>Document</strong> de<br />

specification Logiciel de l’Atelier Reception’ (ref : WCOL-DSL-R-02-CS). This interface<br />

may need to be updated to take into account <strong>Medspiration</strong> specificities.<br />

SR.INT.ARC.0540/I The online space storage shall be accessible by the backup system<br />

available at IFREMER which shall perform a nightly saving of every new or changed files.<br />

SR.INT.ARC.0550/I The online space storage shall be accessible for the backup system<br />

available at IFREMER to restore saved files if needed.<br />

4.1.7 Interfaces between HR-DDS production and dissemination systems<br />

SR.INT.DDS.0010/I The HR-DDS production system (SOC) shall be interfaced with the<br />

DODS server (in the IFREMER dissemination system) to subset and download the<br />

relevant L2P products for the computation of the related HR-DDS.<br />

SR.INT.DDS.0020/I The HR-DDS production system (SOC) shall be interfaced with<br />

the DODS server (in the IFREMER dissemination system) to subset and download the<br />

relevant L4UHR products for the computation of the related HR-DDS..<br />

SR.INT.DDS.0030/I There shall be an ftp interface between SOC and CMS for the<br />

movement of netCDF format AATSR data.<br />

SR.INT.DDS.0040/I<br />

There shall be an interface between SOC and GDAC for the<br />

© 2004 SOC Page 62 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

movement of metadata records.<br />

MED-SOC-RS-001_1<br />

Issue F<br />

SR.INT.DDS.0050/I There shall be an interface between the AATSR mirroring<br />

service and the <strong>Medspiration</strong> AATSR netCDF converter, which shall be monitored by the<br />

DDS system controller.<br />

SR.INT.DDS.0060/I There shall be an interface between the AATSR netCDF<br />

Converter and the netCDF Atlantic extraction process, which shall be monitored by the<br />

DDS system controller.<br />

SR.INT.DDS.0070/I There shall be an interface between the netCDF Converter and<br />

the AATSR HR-DDS Processor, which shall be monitored by the DDS system controller.<br />

SR.INT.DDS.0080/I There shall be an interface between the AATSR HR-DDS Processor<br />

and the HR-DDS Parser and Metadata Processor, which shall be monitored by the DDS<br />

system controller.<br />

SR.INT.DDS.0090/I There shall be an interface between the L2P HR-DDS Processor<br />

and the HR-DDS Parser and Metadata Processor, which shall be monitored by the DDS<br />

system controller.<br />

SR.INT.DDS.0100/I There shall be an interface between the L4UHR HR-DDS<br />

Processor and the HR-DDS Parser and Metadata Processor, which shall be monitored by<br />

the DDS system controller.<br />

SR.INT.DDS.0110/I There shall be an interface between the HR-DDS Processing<br />

<strong>System</strong> and the HR-DDS Archive, which shall be monitored by the DDS system<br />

controller.<br />

SR.INT.DDS.0120/I There shall be an interface between the HR-DDS archive and<br />

the HR-DDS DODS server, which shall be monitored by the DDS system controller.<br />

SR.INT.DDS.0130/I There shall be an interface between the HR-DDS archive and<br />

the HR-DDS sftp server, which shall be monitored by the DDS system controller.<br />

4.1.8 Interfaces between IFREMER system controller and other systems<br />

SR.INT.CTR.0020/I All IFREMER systems shall be interfaced with the Ifremer system<br />

controller, delivering operation and error messages.<br />

© 2004 SOC Page 63 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.INT.CTR.0030/I Remote systems (SOC, CMS) shall deliver to the controller the<br />

relevant error and operation messages to be centralized in the operation log system. The<br />

messages shall be transmitted through ftp.<br />

The IFREMER system controller will not have any direct control on remote systems (at<br />

SOC or CMS).The only interaction is through the operation and error messages.<br />

SR.INT.CTR.0040/I The controller shall deliver to the <strong>Medspiration</strong> remote systems the<br />

error and operation messages related to the dataflow control (retransmission request,<br />

acknowledgement). The messages shall be transmitted through ftp.<br />

SR.INT.CTR.0050/I Error and operation messages shall be formatted as specified in<br />

annex 4 (“Operation and error messages”).<br />

SR.INT.CTR.0060/I All systems under IFREMER control shall be able to return to the<br />

controller a return status after they have completed their task.<br />

4.2 External<br />

4.2.1 Satellite data providers<br />

SR.EXT.ING.0010/I The ingestion module shall query the provider to check the<br />

availability of data files from provider according to Table 5-1, Table 5-2 and Table 5-3<br />

SR.EXT.ING.0020/I The ingestion module shall download the available data files<br />

from data provider according to the protocol defined in Table 5-1, Table 5-2 and Table<br />

5-3<br />

SR.EXT.ING.0030/I Error message generated by the ingestion module shall be sent<br />

to the data provider and the system controller.<br />

SR.EXT.ING.0040/I The configuration files (see SR.QLC.0020 and SR.CFD.0370)<br />

will be received from GHRSST-PP by email (see SR.CFI.L2P.0010). They will be<br />

replaced manually after a coherency check.<br />

Comment: Configuration file format is still to be defined.<br />

SR.EXT.L2P.0010/I SSES will be provided by the GHRSST (see SR.CFI.L2P.0020)<br />

They will include a couple of error statistics (bias and standard deviation) per source and<br />

per confidence level value.<br />

Comment: The missing value to be used during the initiation phase remain to be defined.<br />

© 2004 SOC Page 64 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.EXT.SYS.00xx/I The MMR_DSD records shall be sent by email to the GHRSST-<br />

PO.<br />

SR.EXT.SYS.00xx/I<br />

appendix A6<br />

The format of the MMR_DSD record is described in GDS<br />

4.2.2 In situ data provider<br />

SR.EXT.MDB.0010/I The MDB records production system shall be interfaced with the<br />

CORIOLIS data centre to access the in-situ data.<br />

In the frame of MEDSPIRATION, the CORIOLIS data centre will be considered as the<br />

only source of in-situ data. However, CORIOLIS is a gateway to GTS data and MEDS<br />

data.<br />

SR.EXT.MDB.0020/I The in situ data shall be delivered daily by CORIOLIS in netCDF<br />

format.<br />

This format will be defined in close collaboration with the CORIOLIS team.<br />

SR.EXT.MDB.0030/I The data files shall be directly readable through the IFREMER local<br />

network (no transmission) since the CORIOLIS data centre is hosted by IFREMER.<br />

4.2.3 GDAC<br />

SR.EXT.CTR.0210/I The L2P products shall be delivered to GDAC by ftp push as soon<br />

as they have been ingested in the archiving system.<br />

The UHR/L4 products shall not be delivered to GDAC since they are regarded as a<br />

purely <strong>Medspiration</strong> product in the SoW.<br />

SR.EXT.CTR.0220/I The L2P products shall be delivered to GDAC by IFREMER with<br />

respect to the protocol requirements defined for the dataflow control subsystem of the<br />

system controller (refer to section 1.4.1).<br />

SR.EXT.CTR.0230/I The L2P products shall be formatted as described in annex 1.<br />

4.2.4 GDAC MDB<br />

SR.EXT.CTR.0310/I The <strong>Medspiration</strong> system shall be interfaced with the the GHRSST-<br />

© 2004 SOC Page 65 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

PP MDB at GDAC.<br />

MED-SOC-RS-001_1<br />

Issue F<br />

SR.EXT.CTR.0320/I The MDB records shall be formatted as XML files with respect to the<br />

DTD provided in Annex 3 (“MDB records format”).<br />

SR.EXT.CTR.0330/I The generated MDB data records shall be sent to the GHRSST-PP<br />

MDB through ftp.<br />

It is stated in GDS that both email or ftp may be used for the delivery. Only ftp will be<br />

considered within <strong>Medspiration</strong> since it provides better control on delivery checking and<br />

delays, and for consistency with delivery of MMR-FR and Error logs.<br />

SR.EXT.CTR.0340/I An error message (event code 9) shall be received from the central<br />

error log system if a <strong>Medspiration</strong> MDB record is rejected by the GHRSST MDB.<br />

4.2.5 Master Metadata Repository<br />

SR.EXT.CTR.0410/I The <strong>Medspiration</strong> system shall be interfaced with the Master<br />

Metadata Repository <strong>System</strong>.<br />

SR.EXT.CTR.0420/I Each MMR-FR shall be formatted in XML format, following the DTD<br />

in annex 5 (“format of MMR files”).<br />

SR.EXT.CTR.0430/I Each MMR-FR shall be delivered to the Master Metadata Repository<br />

by ftp push.<br />

It is stated in GDS that both email or ftp may be used for the delivery. Only ftp will be<br />

considered within <strong>Medspiration</strong> since it provides better control on delivery checking and<br />

delays, and for consistency with delivery of MDB records and Error logs.<br />

4.2.6 GDS error log system<br />

SR.EXT.CTR.0510/I The <strong>Medspiration</strong> system shall be interfaced with the central<br />

GHRSST-PP error log system, providing it with error messages or receiving error<br />

messages.<br />

SR.EXT.CTR.0520/I Each error log will be formatted in XML format, following the DTD in<br />

annex 4 (“Operation and error messages”).<br />

No format is specified in the GDS (although WP-ID1.3.1 refers to Appendix A7 which only<br />

specifies error codes) for the error messages. Information such as the name of the data<br />

file rejected or ingested shall be included within the error message.<br />

© 2004 SOC Page 66 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.EXT.CTR.0530/I Each error message shall be delivered to the GHRSST-PP error log<br />

system through ftp.<br />

It is stated in GDS that email shall be used for the delivery. We suggest to use ftp instead<br />

since it provides better control on delivery checking and delays, and for consistency with<br />

delivery of MDB records and MMR-FR.<br />

4.2.7 <strong>User</strong> interfaces<br />

SR.EXT.DIS.0010/I <strong>User</strong>s shall access to the DODS server for L2P, L4 (IFREMER) or<br />

HR-DDS (SOC) product subsetting and download using a DODS client (available on<br />

DODS web site, at http://www.unidata.ucar.edu/packages/dods/) or using a web form<br />

which shall be available with each server (at SOC and IFREMER). The web form comes<br />

with the DODS installation package.<br />

SR.EXT.DIS.0020/I <strong>User</strong>s shall access to the Live Access Server for L4 product<br />

subsetting and download using a web interface available with the IFREMER LAS<br />

installation.<br />

SR.EXT.DIS.0030/I <strong>User</strong>s shall be able to download L2P and L4 products from the<br />

IFREMER ftp site using a standard ftp client.<br />

SR.EXT.DIS.0040/I Ftp push from IFREMER to operational users shall be possible.<br />

<strong>User</strong>s shall register by email to the <strong>Medspiration</strong> help desk providing an ftp server<br />

address with write permission. The products shall be pushed following the protocol<br />

defined in section 1.4.1.<br />

4.2.8 DDS external interfaces<br />

SR.EXT.DDS.0010/I There shall be an ftp interface between SOC and ESA for the<br />

movement of ATS_NR__2P files.<br />

SR.EXT.DDS.0020/I There shall be an interface between SOC and GHRSST-PP for<br />

the movements of ATS_NR__2P files.<br />

SR.EXT.DDS.0030/I There shall be an open interface between SOC and the Internet<br />

for the movement of HR-DDS data via DODS.<br />

SR.EXT.DDS.0040/I There shall be an open interface between SOC and GHRSST-<br />

PP for the movement of HR-DDS data via sftp.<br />

© 2004 SOC Page 67 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.EXT.DDS.0050/I There shall be an interface between the HR-DDS <strong>System</strong> and<br />

GDAC for the movement of metadata.<br />

SR.EXT.DDS.0060/I There shall be an interface between the HR-DDS <strong>System</strong><br />

Controller and GDAC for the movement of ERRLOG records.<br />

© 2004 SOC Page 68 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

5. PERFORMANCE REQUIREMENTS<br />

5.1 Scenario Definitions<br />

SR.PER.SYS.0010/A The L2P data files shall ideally be delivered within 6 hours after<br />

acquisition. It is recognised this may be not possible depending on the L2 data providers<br />

on which the <strong>Medspiration</strong> system has no control and thus it is just an ideal target.<br />

SR.PER.SYS.0011/I<br />

L2P must be produced in near real time.<br />

SR.PER.SYS.0015/I A cutoff time (Ω) must be specified after which L2P data are no<br />

longer eligible for inclusion within the analysis. This must take account the time taken to<br />

process L2P data (∆Time L2P ), the time required to perform the analysis (∆T analysis ) itself<br />

and the Maximum desirable output time delay (T output ) using:<br />

Ω =<br />

(Eqn. 2.3.1)<br />

Toutput<br />

− ∆Tanalysis<br />

− ∆TL2P<br />

For example: if T output = 12:00 at T+1, ∆T analysis =4h and, ∆T L2P =2h then Ω= 06:00 at T+1.<br />

This means that L2P SST data for day T can be accepted until day T+1 at 06:00 UTC.<br />

Comment: After this cutoff time data concerning the analysis at day T will be considered<br />

as too late and a message error will be sent.<br />

SR.PER.SYS.0020/A The UHR/L4 data files shall be provided within 12 hours after the<br />

end of the processing window, that is to say before T0 + 36h when processing the (T0 -><br />

T0 + 24h) UHR/L4 product.<br />

SR.PER.SYS.0030/A The L2P MMR_FR metadata records shall be provided to MMR<br />

within 60 minutes after L2P production.<br />

SR.PER.SYS.0040/A The MDB records shall be delivered within 24 hours after they<br />

are delivered by the in situ data center.<br />

SR.PER.SYS.0050/A The L2P MMR_FR rejected by the MMR shall be less than 2%.<br />

SR.PER.SYS.0060/I<br />

instant<br />

The MMR_DSD describe the availability of data files at any given<br />

SR.PER.SYS.0070/I A too long delay between two data files availability of a same data<br />

© 2004 SOC Page 69 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

stream shall raise an error message to the GHRSST-PP and the data provider. This<br />

delay shall be parameterised for each data stream.<br />

SR.PER.SYS.0080/I<br />

better.<br />

All SST data products are requested to an accuracy of 0.4K or<br />

SR.PER.DDS.0010/I HR-DDS MMR_FR files shall be provided to MMR within 60<br />

minutes of granule generation.<br />

SR.PER.DDS.0020/I HR-DDS MMR-FR file rejection rate shall be less than 2%<br />

SR.PER.DDS.0030/I HR-DDS granules shall be available via DODS within 3 hours of<br />

L4UHR / L2P file generation.<br />

SR.PER.DDS.0050/I The ESA server ftp-ops.pde.envisat.esa.int shall be deemed not<br />

to provide an acceptable service if the average bandwidth available to SOC is less than<br />

50Kbytes/sec and/or more than 10% of interface attempts result in a timeout.<br />

SR.PER.DDS.0060/I<br />

less than 98%.<br />

The AATSR mirroring service shall have a user up-time of no<br />

SR.PER.DDS.0070/I<br />

than 98%.<br />

The DODS and sftp servers shall have a user up-time of no less<br />

5.2 Characterisation of L2 input data<br />

This section is a review of L2 data requested by the GHRSST-PP.<br />

5.2.1 Satellite SST<br />

Table 5-1: Satellite SST data streams used by <strong>Medspiration</strong>.<br />

Name sensor resolution Coverage <strong>Data</strong> provider/delivery<br />

ATS_NR_2P AATSR 1 km Atlantic ftp pull from SOC<br />

AVHRR16-L<br />

AVHRR 2 km Mediterranean ftp pull from PODAAC<br />

AVHRR17-L<br />

AVHRR16-L<br />

AVHRR17-L<br />

(NARSST)<br />

AVHRR 2 km European seas<br />

(RGD)<br />

Internal from<br />

EUMETSAT/CMS<br />

© 2004 SOC Page 70 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Name sensor resolution Coverage <strong>Data</strong> provider/delivery<br />

AVHRR16-G<br />

AVHRR 9 km Global ftp pull from PODAAC<br />

AVHRR17-G<br />

SEVIRI SEVIRI 0.1 deg. Atlantic Internal from<br />

EUMETSAT/CMS<br />

AMSRE AMSR-E 0.25 deg. Global ftp pull from REMSS<br />

TMI TRMM-TMI 0.25 deg. Global 40N-40S ftp pull from REMSS<br />

5.2.2 Auxiliary satellite data<br />

Table 5-2: Auxiliary satellite data stream used by <strong>Medspiration</strong>.<br />

Name Sensor parameter resolution Coverage <strong>Data</strong> provider and<br />

delivery<br />

NISE SSM/I Fract. Sea<br />

Ice<br />

25 km Northern and<br />

Southern<br />

hemisphere<br />

ftp pull from NSIDC<br />

SEVIRI-SSI SEVIRI 3h SSI 0.1 deg. Atlantic Internal from<br />

EUMETSAT/CMS<br />

AMSRE-WSP AMSR-E Wind speed 0.25 deg. Global ftp pull from REMSS<br />

AMSRE-ICE AMSR-E Ice flag 0.25 deg. Global ftp pull from REMSS<br />

AMSRE-RAIN AMSR-E Rain rate 0.25 deg. Global ftp pull from REMSS<br />

TMI-WSP TRMM-TMI Wind speed 0.25 deg. Global 40N-40S ftp pull from REMSS<br />

TMI-RAIN TRMM-TMI Rain rate 0.25 deg. Global 40N-40S ftp pull from REMSS<br />

AVHRR-AOD AVHRR AOD 1. deg. Global 70S-70N ftp pull from NESDIS<br />

AVHRR-AOD AVHRR AOD 2 km Mediterranean ftp pull from PODAAC<br />

AVHRR-AOD AVHRR AOD 9 km Global ftp pull from PODAAC<br />

The auxiliary data provided with a SST source are given in italics. They should be given<br />

© 2004 SOC Page 71 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

priority over any external sources, apart in case of known deficiencies.<br />

Issue F<br />

5.2.3 NWP model outputs<br />

Table 5-3: NWP output data stream used by <strong>Medspiration</strong>.<br />

Name Parameter resolution coverage <strong>Data</strong> provider and<br />

delivery<br />

ECMWF-ICE<br />

Fract. Sea<br />

Ice<br />

0.5 deg. globe pull from ECSS<br />

ECMWF-SSI 6h SSI 0.5 deg. globe pull from ECSS<br />

ECMWF-U<br />

ECMWF-V<br />

U-wind<br />

component<br />

V-wind<br />

component<br />

0.5 deg. globe pull from ECSS<br />

0.5 deg. globe pull from ECSS<br />

5.2.4 Reference data sets<br />

The following reference data sets have been identified<br />

• 10 day minimum climatologic SST<br />

• MODIS/Terra land cover<br />

• Most recent SST analysis : to be provided by GHRSST; meanwhile, a 10 day mean<br />

climatology will be used.<br />

• Upwelling maps (EUR/ATLUPW and EUR/MEDUPW): they will be derived from the<br />

Landmask (LPDAAC/LANDMASK) at 10 km resolution over the Atlantic and 2 km<br />

resolution over the Mediterranean. A coastal area will be defined as<br />

upwelling_distance (km) away from coastline, upwelling_distance should be distinct<br />

in the Atlantic and in the Mediterranean.<br />

• Contamination map : equivalent to the land mask at the micro-wave SST resolution<br />

over the Atlantic.<br />

© 2004 SOC Page 72 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Table 5-4 Reference data sets used by <strong>Medspiration</strong>.<br />

Issue F<br />

name Parameter resolution coverage <strong>Data</strong> provider and<br />

delivery<br />

CMS-mini<br />

CMS-mean<br />

(substitute to<br />

most recent<br />

SST analysis)<br />

LMSK +<br />

upwelling map+<br />

contamination<br />

map<br />

10 day mini<br />

climatic SST<br />

10day mean<br />

climatic SST<br />

Land cover<br />

classification<br />

0.1 deg. globe Internal from<br />

EUMETSAT/CMS<br />

0.1 deg. globe Internal from<br />

EUMETSAT/CMS<br />

1km globe ftp pull from LPDAAC<br />

5.3 Characterisation of in situ data<br />

This section is a review of the in situ data requested by the GHRSST-PP against what is<br />

available at the CORIOLIS data centre. Discrepancies between the requirements and<br />

the actual available data are identified and solutions are proposed to provide in the end<br />

the best set of MDB records.<br />

The requirements in the GDS are :<br />

1. sources : CORIOLIS, MEDS, additional in situ sources if relevant [RD-1: 1.2]<br />

2. sensors : moored and drifting buoys, ships [RD-3: WP-ID3.1.1]<br />

3. SST measurements : SST1m [RD-3: WP-ID3.1.1]<br />

4. ancillary data : as specified in appendix A3-1 [RD-1: 1.2]<br />

A first characterisation of the available data in the CORIOLIS database has been<br />

performed with the CORIOLIS staff to assess, and eventually refine, these requirements<br />

and to quickly identify eventual problem areas.<br />

1. sources : MEDS data provided through the GTS-PP are collected by CORIOLIS<br />

(when not already obtained from the GTS). CORIOLIS can thus be identified as the<br />

single in situ data provider which will significantly reduce the complexity of the<br />

© 2004 SOC Page 73 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

<strong>Medspiration</strong> external interfaces. In addition, this will give access to some other data<br />

sources such as Argo or the in situ data collected over the Mediterranean sea during<br />

the MFSTEP experiment (from September 2004 for 6 months).<br />

2. The data identified in the CORIOLIS database is shown in Table 5-5:<br />

Table 5-5 In situ data for the Mediterranean available through Coriolis<br />

Sensor Depth Age Count<br />

(Mediterranean<br />

in 2003)<br />

Comment<br />

Moored<br />

From<br />

Daily (GTS)<br />

Very few in Mediterranean sea<br />

buoys<br />

surface<br />

666<br />

Floats<br />

> 5 m Near real<br />

Not requested but good quality<br />

(ARGO)<br />

time<br />

Drifting<br />

From<br />

Daily (GTS)<br />

not available in<br />

~1000 over global ocean<br />

buoys<br />

surface<br />

2003<br />

CTD (ships)<br />

From<br />

Near real<br />

High quality but still very few<br />

surface,<br />

time<br />

measurements<br />

every meter<br />

XBT (ships) > 5 m,<br />

Often<br />

1015<br />

Quality of measurements between<br />

every meter<br />

highly<br />

[0, 5m] uncertain, often discarded<br />

delayed<br />

TSG (ships) > 5 m Often<br />

highly<br />

delayed<br />

141118 Quality uncertain (particularly on<br />

VOS = 70% of measurements)<br />

3. SST measurements : It can observed from the above table that selecting only<br />

SST1m measurements will lead to ignore many other data sources. It is suggested to<br />

relax this constraint down to 5 m (the closer measurement to the 1 m depth will be<br />

selected anyway for each profile) which could allow to consider also the float<br />

measurements which quality may be valuable for the MDB.<br />

4. The meteorological data are currently not available within the CORIOLIS data. It was<br />

agreed with CORIOLIS head that they will be collected from the GTS by the<br />

CORIOLIS staff (when simultaneously available with a SST measurement) and<br />

© 2004 SOC Page 74 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

ingested in the database. Not all the parameters listed in the GDS table A4.1.2 will<br />

be available, only what is distributed through the GTS will be retrieved.<br />

© 2004 SOC Page 75 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

6. EXPANDABILITY REQUIREMENTS<br />

6.1 General<br />

SR.EXP.SYS.0010/I Configuration files shall be used to implement configurable code<br />

as input parameters are expected to evolve.<br />

SR.EXP.SYS.0020/I The system shall be extensible and flexible to accommodate new<br />

satellite sensors and the potential failure of others. Software shall not be dependent on<br />

any single input data stream and be capable of ingesting new data streams with minimum<br />

additional development.<br />

SR.EXP.SYS.0030/I Extensions are envisaged, in a further stage (not within<br />

<strong>Medspiration</strong>), to provide UHR products for the Nordic Seas 48N-75N 12W-70 E 2-3 km<br />

and Black Sea 40N-48N 27 E–42 E 2-3 km<br />

6.2 Generation of L2P<br />

SR.EXP.L2P.0010/I It is foreseen that data streams will change throughout the life of<br />

the GHRSST-PP. The GDS must be capable of bringing new data streams in and out of<br />

the processing system without affecting other components of the system.<br />

SR.EXP.L2P.0020/I Future revisions of the GDS will most likely use near<br />

contemporaneous wind speed observations from other satellite instruments.<br />

SR.EXP.L2P.0030/I<br />

modified.<br />

It is expected that the methodology used to derive IPCV may be<br />

SR.EXP.L2P.0040/I The GDS procedures for the derivation of IPCV and MPCV<br />

values described in GDS WP-ID2.1.1.14 are quite basic. It is clear that significant<br />

improvements can be made leading to a better quality of pixel error statistics. It is<br />

foreseen that this component of the GDS will be upgraded following further research and<br />

development within the GHRSST-PP Diagnostic <strong>Data</strong> Set.<br />

SR.EXP.L2P.0050/I It is foreseen that minor modifications to the procedures used to<br />

control the quality of input data and derive L2P data files may be required once significant<br />

experience is gained with the GDS.<br />

SR.EXP.L2P.0060/I As experience is gained in the distribution and application of<br />

GDS data products, data product format changes are foreseen based on user<br />

requirements and data delivery constraints.<br />

© 2004 SOC Page 76 of 193


<strong>Medspiration</strong><br />

MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Issue F<br />

SR.EXP.L2P.0070/I The GDS reference processor system should be used to test all<br />

proposed upgrades so that the implications of the proposed changes may be evaluated<br />

prior to operational implementation.<br />

6.3 Diagnostic data set<br />

SR.EXP.DDS.0010/I The HR-DDS system shall be capable of ingesting new data<br />

sources with minimum effort.<br />

SR.EXP.DDS.0020/I The HR-DDS system shall be capable of producing data for any<br />

geographical region; however, total data volume may be limited by <strong>Medspiration</strong> to 10<br />

Gigabytes per month as necessary.<br />

SR.EXP.DDS.0030/I<br />

stream.<br />

The HR-DDS system not be dependant on any specific L2 data<br />

SR.EXP.DDS.0040/I The HR-DDS code shall not contain any ‘magic numbers’,<br />

editable configuration files shall be used.<br />

SR.EXP.DDS.0050/I No data source other than ATS_NR__2P shall be required to be<br />

mirrored on the AATSR mirroring service.<br />

SR.EXP.DDS.0060/I The Atlantic Extractor Process shall be capable of selecting data<br />

from additional rectangular regions within the limits already set, as defined by the<br />

GHRSST-PP Problem Resolution Board.#.<br />

© 2004 SOC Page 77 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

7. SECURITY REQUIREMENTS<br />

SR.SEC.SYS.0010/I<br />

with use of L2P.<br />

There are no data product dissemination limitations associated<br />

Comment: However some inputs are password protected for restricted use within the<br />

GHRSST community.<br />

SR.SEC.DDS.0010/I The AATSR mirroring service shall employ authorisation<br />

procedures no weaker than the ftp-ops.pde.envisat.esa.int system.<br />

© 2004 SOC Page 78 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

8. DESIGN REQUIREMENTS<br />

SR.DES.SYS.0010/I<br />

The system shall be developed in a robust manner.<br />

SR.DES.SYS.0020/I<br />

The system shall be built in a modular way.<br />

SR.DES.SYS.0030/I<br />

The system design should focus on reliability and simplicity.<br />

SR.DES.SYS.0040/I Configuration file shall be used to implement configurable code<br />

as input parameters are expected to evolve.<br />

SR.DES.SYS.0050/I The system should be able to ingest new data streams or<br />

remove existing data streams with minimum code modification.<br />

SR.DES.SYS.0060/I<br />

Software shall not be dependent on any single data stream.<br />

© 2004 SOC Page 79 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

9. RELIABILITY, AVAILABILITY AND MAINTAINABILITY<br />

REQUIREMENTS<br />

SR.REL.SYS.0010/I The system shall run automatically in near real time without human<br />

intervention except for monitoring and maintenance tasks.<br />

SR.REL.ARC.0020/I The archiving system shall provide a backup service to restore data<br />

lost in case of a disk crash or any hardware failure causing no recoverable damage to<br />

the archived data.<br />

© 2004 SOC Page 80 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

10. TESTING REQUIREMENTS<br />

SR.TST.SYS.0020/I In order to verify that each processing centre has implemented<br />

the GDS correctly a reference GDS data processor will be used to process a set of data<br />

products using a GDS test data set.<br />

SR.TST.SYS.0030/I<br />

generated.<br />

Error message indicating data set rejection shall be correctly<br />

SR.TST.SYS.0040/I<br />

Error message shall be delivered correctly to data provider.<br />

SR.TST.SYS.0050/I<br />

ERRLOG.<br />

Error messages shall be delivered correctly to the GHRSST-PP<br />

SR.TST.SYS.0060/I MMR_DSD records shall be prepared and registered correctly<br />

within the GHRSST-PP MMR system.<br />

SR.TST.SYS.0070/I<br />

L2 data files shall be correctly renamed.<br />

SR.TST.SYS.0080/I L2P output data products based on the GDS test input data set<br />

produced at each RDAC and GDAC should be identical to those produced by the GDS<br />

reference data processor.<br />

SR.TST.SYS.0090/I L2P measurement data shall be identical to the input SST data<br />

but with additional error and QC).<br />

SR.TST.SYS.0100/I<br />

manner.<br />

MMR_FR metadata records shall be submitted in a timely<br />

© 2004 SOC Page 81 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

11. SAFETY REQUIREMENTS<br />

There are no specific human safety requirements for <strong>Medspiration</strong>.<br />

© 2004 SOC Page 82 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

12. CUSTOMER FURNISHED ITEMS<br />

SR.CFI.SYS.0010/I GHRSST-PP will provide a DVD containing 2 days of L2 data<br />

files having global coverage from each L2 satellite data stream considered by the<br />

GHRSST-PP. Both polar orbiting and geostationary observations will be included.<br />

SR.CFI.SYS.0020/I GHRSST-PP will provide a DVD containing 1 week of in situ data<br />

straddling the same period as the data sets described in SR.CFI.SYS.0010<br />

SR.CFI.SYS.0030/I GHRSST-PP will provide an example GHRSST-PP Match-Up<br />

<strong>Data</strong>base with entries for each sensor.<br />

SR.CFI.SYS.0040/I GHRSST-PP will provide example GHRSST-PP MDB data<br />

records, netCDF format L2P, SSTfnd and HR-DDS<br />

SR.CFI.L2P.0010/I The GHRSST-PP Project Office shall update the configuration<br />

files for quality checks of ingested data and these shall be sent by email to the person<br />

responsible for the ingestion and L2P processor at CMS. (see SR.EXT.ING.0040).<br />

SR.CFI.L2P.0020/I The GHRSST-PP Project Office shall every month supply<br />

updated SSES by email to the person responsible for the L2P processor at CMS (see<br />

SR.EXT.L2P.0010).<br />

SR.CFI.DDS.0010/I ESA shall maintain the ftp-ops.pde.envisat.esa.int service for the<br />

duration of <strong>Medspiration</strong> in an acceptable state.<br />

SR.CFI.DDS.0020/I<br />

DDS sites.<br />

GHRSST-PP shall provide SOC with the locations of the HR-<br />

SR.CFI.DDS.0030/I Only the GHRSST-PP Problem Resolution Board shall be able to<br />

adjust the HR-DDS locations after SRR.<br />

SR.CFI.DDS.0040/I ESA shall provide <strong>Medspiration</strong> access to the ftpops.pde.envisat.esa.int<br />

server<br />

© 2004 SOC Page 83 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

13. GLOSSARY<br />

AATSR<br />

AMSR-E<br />

AMSRE-ICE<br />

AOD<br />

ARGO<br />

AVHRR<br />

CERSAT<br />

CMS<br />

CNR<br />

CORIOLIS<br />

CTD<br />

DDS<br />

DODS<br />

DTD<br />

DV<br />

ECMWF<br />

ERRLOG<br />

ESA<br />

EUDAC<br />

EURDAC<br />

FNMOC<br />

GDAC<br />

GDS<br />

GHRSST-PP<br />

Advanced Along-Track Scanning Radiometer<br />

Advanced Microwave scanning radiometer for Earth Observation<br />

auxiliary fractional sea ice field<br />

Aerosol Optical Depth<br />

Global array of temperature/salinity profiling floats<br />

Advanced very high resolution radiometer (NASA)<br />

Centre ERS d'Archivage et de Traitement - French ERS<br />

Processing and Archiving Facility<br />

Le Centre de Météorologie Spatiale<br />

Consiglio Nazionale delle Ricerche<br />

A system for operational oceanography under development in<br />

France to monitor and forecast the ocean behaviour.<br />

Conductivity/Temperature/Depth instruments<br />

(High resolution) Diagnostic data set<br />

Distributed Oceanographic <strong>Data</strong> <strong>System</strong><br />

<strong>Document</strong> Type Definition<br />

Diurnal Variability<br />

European Centre for Medium-Range Weather Forecasts<br />

Error log<br />

European Space Agency<br />

European data assembly centre<br />

European regional data assembly centre<br />

Fleet Numerical Meteorology and Oceanography Center<br />

Global data assembly centre<br />

GHRSST-PP data processing specification<br />

GODAE high resolution sea surface temperature pilot project<br />

© 2004 SOC Page 84 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

GODAE<br />

GTS<br />

HR-DDS<br />

IFREMER<br />

IPCV<br />

IPO<br />

IR<br />

LAS<br />

LPDAAC/LANDMASK<br />

MDB<br />

MDR<br />

MEDS<br />

MEDSPIRATION<br />

MFSTEP<br />

MMR<br />

MMR_DSR<br />

MMR_FR<br />

MMR-DSD<br />

MODIS<br />

MPCV<br />

MPS<br />

NISE<br />

NWP<br />

NWP-ICE<br />

Global ocean data assimilation experiment<br />

Global Telecommunication <strong>System</strong><br />

High-resolution diagnostic data set<br />

Institut français de recherche pour l'exploitation de la mer<br />

(French Research Institute for Exploitation of the Sea)<br />

Infrared proximity confidence values<br />

International project office<br />

Infrared<br />

Live Access Server<br />

Land Processes <strong>Data</strong> Active Archive Centre / LANDMASK<br />

Match-up data base<br />

Match-up data record<br />

Marine Environmental <strong>Data</strong> Service<br />

A European Contribution to the Global Ocean <strong>Data</strong> Assimilation<br />

Experiment Gridded High Resolution Sea Surface Temperature<br />

Pilot Project<br />

Mediterranean forecasting system toward environmental<br />

prediction<br />

Master metadata repository<br />

MMR data set record<br />

MMR file record<br />

Master metadata repository data set description<br />

Moderate Resolution Imaging Spectroradiometer (Aqua, Terra)<br />

Microwave proximity confidence values<br />

<strong>Medspiration</strong> processing system<br />

Reference sea ice data sets<br />

Numerical Weather Prediction<br />

NWP fractional sea ice field<br />

© 2004 SOC Page 85 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

OI<br />

QC<br />

RD<br />

RDAC<br />

SEVIRI<br />

SOC<br />

SOW<br />

SR<br />

SSES<br />

SSI<br />

SSM/I<br />

SST<br />

TRMM-TMI<br />

TSG<br />

UHR<br />

XBT<br />

Optimal interpolation<br />

Quality control<br />

Referenced <strong>Document</strong>s<br />

Regional data assembly centre<br />

Spinning Enhanced Visible and Infrared Imager (Meteosat<br />

Second Generation)<br />

Southampton Oceanographic Centre<br />

Statement of Work<br />

<strong>System</strong> <strong>Requirements</strong><br />

Sensor specific error statistics<br />

Surface solar irradiance<br />

Special Sensor Microwave Imager (Defense Meteorological<br />

Satellite Program)<br />

Sea surface temperature<br />

Tropical Rainfall Measuring Mission Microwave Imager<br />

Thermosalinograph<br />

Ultra-high resolution<br />

Expendable bathythermograph<br />

.<br />

© 2004 SOC Page 86 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

This Page Is Intentionally Blank<br />

© 2004 SOC Page 87 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

APPENDIX A TRACE BETWEEN USER REQUIREMENTS AND SYSTEM REQUIREMENTS<br />

This appendix contains two large tables which facilitate cross-referencing between the three reference documents [RD-1, RD-2 and RD-3] and the<br />

system requirements presented in the main body of this document. Table A1 segments the text in the reference documents (except for sections that are<br />

clearly out of the scope of <strong>Medspiration</strong>) and assigns unique identifiers to each statement, alongside the document page number and any relevant table,<br />

rule or figure numbers. The right hand column of this large table shows the SRs which can be traced back to the statement. Alternatively the statement<br />

is judged to be one which is intended to provide background information, or (in the case of [RD-3]) is out of the scope of <strong>Medspiration</strong>. In this case the<br />

entry is not a software requirement (NASR).<br />

Table A2 provides the reverse reference, from a systematic list of SRs back to the IDs of the reference document statements (as defined in Table A1)<br />

from which they are derived or which have influenced their definition. This arrangement using two tables should facilitate the following:<br />

• Checks to ensure that the system functionality required by the reference documents is fully reflected in the <strong>System</strong> requirements.<br />

• A means of rapidly cross-referencing those parts of the reference documents which relate to each of the system requirements.<br />

Not all SRs have a trace back to the reference documents. For these, their origin should be apparent from their context in this document. In many<br />

cases they relate to internal requirements such as internal interfaces or system controllers which are essential for the specification of a consistent<br />

overall system. Where there are discrepancies between the SR and the reference to which it is traced, this should be explained in the comments<br />

following the SR in the main body of this document, or will be discussed in the Design Justification File (DJF).<br />

© 2004 SOC Page 88 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Table A1. Trace between the documented user requirements and the system requirements specified in this document<br />

Key to abbreviations used in the SR entry column:<br />

NASR<br />

ATPD<br />

Not a system requirement (Text in the reference document which may be explanatory or background, and<br />

where there is nothing mentioned which contributes to defining the system)<br />

Not a system requirement, but concerned with testing the system and thus may be relevant to the Acceptance Test<br />

Procedures <strong>Document</strong><br />

Out of Scope<br />

Format<br />

Specifies requirements for a part of GHRSST-PP activity that lies outside the scope defined by the Statement of Work<br />

Contains information about detailed formatting of data files, reference variables, configuration files etc.<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

Statement of Work<br />

3 1 Introduction<br />

3 sow-001 The international GODAE steering committee initiated the international GODAE High Resolution SST Pilot Project<br />

(GHRSST-PP) to develop a system that will deliver a new generation of global coverage high resolution (better than 10 km<br />

and ~6 hourly) SST data products in real time that are tuned to the requirements of modern operational oceanography.<br />

NASR (not a<br />

software<br />

requirement)<br />

3 sow-002 GHRSST-PP data products will be derived by combining complementary Level-2 (L2) satellite and in situ observations to NASR<br />

improve spatial coverage, temporal resolution, cross-sensor calibration stability and SST product accuracy through synergy.<br />

© 2004 SOC Page 89 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

3 sow-003 Due to the large volumes of data that are considered by the GHRSST-PP and the demanding data product delivery<br />

SR entry<br />

NASR<br />

constraints, the GHRSST-PP is designed as a distributed processing system.<br />

3 sow-004 The strategic implementation concept of regional/global task sharing (RGTS) by regional data product assembly centres<br />

NASR<br />

(RDAC) interconnected by dedicated, fast data connections has been adopted to underpin the practical implementation of<br />

the GHRSST-PP data processing system.<br />

3 sow-005 RDAC ingest, quality control and merge together existing L2 satellite and in situ data sources to generate regional area<br />

NASR<br />

coverage quality controlled data products in realtime.<br />

3 sow-006 Every 6 hours, these data products are consolidated at RDAC and sent to a Global <strong>Data</strong> Analysis Centre (GDAC) where<br />

NASR<br />

they are integrated together with other RDAC data streams having an identical data format but covering different<br />

geographical areas to provide global coverage data products.<br />

3 sow-007 The <strong>Medspiration</strong> project will provide a European RDAC service to the GHRSST-PP during 2004-2006 by generating and<br />

disseminating European and Atlantic area SST data products according to the GHRSST-PP Detailed Processing Model<br />

SYS.0020<br />

ING.0130<br />

(DPM) [AP-1].<br />

3 sow-008 In addition, <strong>Medspiration</strong> will serve the direct needs of the European SST user community. SYS.2050<br />

3 1.1 The GHRSST-PP data processing model<br />

3 sow-009 The combined <strong>Medspiration</strong> DPM1 [AP-1], developed by the International GHRSST-PP Science Team, provides a<br />

NASR<br />

complete technical description of the data processing tasks required to generate GHRSST-PP data products.<br />

3 sow-010 Before SST data originating from different sources can be properly assimilated into ocean model systems or analysed<br />

NASR<br />

together to provide new SST data products that capitalise on their synergy, a bias and rms error estimate is required for<br />

each [pixel] measurement. This is a fundamental component of the GHRSST-PP processing model.<br />

3 sow-011 Error estimates for each L2 SST measurement are based on the statistical analysis of a match up database (MDB) NASR<br />

© 2004 SOC Page 90 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

containing satellite and in situ observations.<br />

3 sow-012 A quantitative confidence value is derived that is then used to assign statistical bias and rms. error estimate to each<br />

NASR<br />

measurement.<br />

3 sow-013 The outputs of this process are L2 pre-processed (L2P) products for each data file ingested by the processor. NASR<br />

3 sow-014 Operational ocean modelling users that have requested [AP-2] this type of SST data product may access L2P data<br />

NASR<br />

products directly in real time.<br />

3 sow-015 However, the majority of SST data users request [AP-2] complete SST fields, free of gaps caused by clouds, rain or lack of<br />

NASR<br />

data coverage, at regular intervals of between 6 and 24 hours.<br />

3 sow-016 Some require an estimate of the sub-surface SST field that cannot be directly measured using current satellite techniques NASR<br />

3 sow-017 whereas others require an estimate of the surface skin temperature that is in direct contact with the overlying atmosphere<br />

NASR<br />

and subject to considerable diurnal variability.<br />

3 sow-018 To address these needs, L2P data are used in an optimal interpolation scheme that is designed to NASR<br />

3 sow-019 (a) account for differences in spatial and temporal sampling characteristics of each data stream, NASR<br />

3 sow-020 (b) account for gaps in coverage due to the presence of cloud, rain or lack of data, NASR<br />

3 sow-021 (c) pre-condition data to account for SST diurnal variability and sensor bias, NASR<br />

3 soo-022 (d) retrieve an estimate of the subsurface temperature field (referred to as the foundation temperature, SSTfnd2). NASR<br />

3 sow-023 In addition, a measure of diurnal variability within the data product time domain will be derived to accompany the SSTfnd<br />

NASR<br />

estimate<br />

3 sow-024 and an estimate of the skin temperature of the ocean (SSTskin) will be retrieved based on parameterisations. NASR<br />

4 sow-025 L2P and SSTfnd SST products have been designed to satisfy the requirements of operational ocean forecast and<br />

NASR<br />

prediction systems.<br />

4 1.2 Objectives of the <strong>Medspiration</strong> project<br />

© 2004 SOC Page 91 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

4 sow-026 <strong>Medspiration</strong> will be the European RDAC service to the GHRSST-PP and provide SST data products to specific<br />

operational European users in near real time. The emphasis of the project is to use complementary satellite L2 SST and in<br />

situ data products together to deliver a new generation of SST data products. The <strong>Medspiration</strong> project consists of a data<br />

SR entry<br />

SYS.0010<br />

ARC.0010<br />

ARC.0040<br />

processing system, an off-line data archive and a data product dissemination service.<br />

4 sow-027 The following satellite data shall be used as input to the <strong>Medspiration</strong> software;<br />

4 sow-028 • AATSR EXT.ING.0010,<br />

DDS.1010<br />

4 sow-029 • AVHRR EXT.ING.0010<br />

4 sow-030 • TMI EXT.ING.0010<br />

4 sow-031 • AMSRE EXT.ING.0010<br />

4 sow-032 • AMSR Ignore<br />

4 sow-033 • MSG EXT.ING.0010<br />

4 sow-034 • SSM/I3 (L3 gridded SSM/I derived wind speed data as specified in [AP-2] required for derivation of L2P data products) EXT.ING.0010<br />

4 sow-035 These data are fully described in Table 4.2.1 of the <strong>Medspiration</strong> URD [AP-2]. NASR<br />

4 sow-036 The following in situ data shall be used as input to the <strong>Medspiration</strong> software: NASR<br />

4 sow-037 • Moored buoy, drifting buoy and ship SST observations obtained from the CORIOLIS and MEDS data centres. (4<br />

Additional in situ SST observations may also be used in the <strong>Medspiration</strong> software) These data are fully described in Table<br />

EXT.ING.0010<br />

EXT.MDB.0010<br />

4.2.1 of the <strong>Medspiration</strong> URD [AP-2].<br />

4 sow-038 Additional auxiliary and reference data that are required are fully described in [AP-1, Appendix A3-1]. NASR<br />

4 sow-039 The following products/services shall be developed and operated for 12 months: NASR<br />

4 sow-040 • Production and dissemination of L2P for the overall European RDAC, page 33 of <strong>Medspiration</strong> URD [AP-2] SYS.0050<br />

DIS.0010<br />

© 2004 SOC Page 92 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

ARC.0010<br />

4 sow-041 • Production and dissemination of SSTfnd at Ultra High Resolution (UHR) for the Mediterranean Sea5 coverage, page 34<br />

of the <strong>Medspiration</strong> URD [AP-2]. (The Mediterranean Sea has been selected as an initial priority area within the<br />

<strong>Medspiration</strong> project [AP-2] due to lower cloud coverage conditions compared to other suggested areas. Other UHR areas<br />

SYS.0080<br />

DIS.0010<br />

EXP.SYS.0030<br />

indicated in [AP-2] may also be considered at a later date.)<br />

6 1.4 Product Definitions<br />

6 1.4.1 GHRSST-PP L2P data products for the European RDAC<br />

6 sow-042 A full description of GHRSST-PP L2P data products can be found in [AP-1, page 86]. SYS.0052<br />

6 sow-043 L2P data products are generated for every L2 satellite SST data file that is ingested into the <strong>Medspiration</strong> system. Each<br />

L2P data file consists of the original L2 SST measurements together with a unique set of confidence data and a mean bias<br />

and rms. error estimate (known as sensor specific error statistics (SSES)) that must be derived for every [pixel]<br />

measurement using the GHRSST-PP MDB [AP-1, page 58].<br />

L2P.0010<br />

L2P.0020<br />

L2P.0040<br />

L2P.0050<br />

L2P.0070<br />

6 sow-044 SSES will be provided to the contractor by the GHRSST-PP. CFI.L2P.0020<br />

6 sow-045 L2P data files are formatted as netCDF data files according to the GHRSST-PP format fully documented in [AP-1, page<br />

INT.ARC.0030<br />

100].<br />

6 sow-046 L2P data files are required by ocean modelling data assimilation systems SYS.0100<br />

6 sow-047 and for inclusion into GHRSST-PP and <strong>Medspiration</strong> L4 analysis procedures. SYS.2020<br />

6 sow-048 A GHRSST-PP metadata record, fully documented in [AP-1, page 144], will be created for each L2P data file and sent to<br />

the GHRSST-PP global metadata repository as described in [AP-1, page 144].<br />

SYS.0055<br />

ING.0030<br />

6 1.4.2 SSTfnd UHR data products for the Mediterranean Sea<br />

6 sow-049 A number of European SST data users [AP-2] have requested data products that have a higher spatial resolution [2-3km] NASR<br />

© 2004 SOC Page 93 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

than the standard 1/12º GHRSST-PP L4 SSTfnd data product specification [AP-1, page 61].<br />

6 sow-050 SSTfnd UHR data products will be generated by <strong>Medspiration</strong> for European users for the Mediterranean Sea as first<br />

priority6 [AP-2].<br />

6 sow-051 L4 SSTfnd UHR data products will be generated from L2P data products following the same methodology as for GHRSST-<br />

SYS.0080<br />

L4A.0010<br />

SYS.0083<br />

PP SSTfnd data products<br />

6 sow-052 allowing for variations in decorrelation length scales and a higher spatial grid specification. SYS.0082<br />

6 1.4.3 GHRSST-PP High Resolution Diagnostic <strong>Data</strong> Set (HR-DDS) granules<br />

6 sow-053 The <strong>Medspiration</strong> project shall extract GHRSST-PP HR-DDS data granules for its coverage area SYS.0070<br />

DDS.0010<br />

6 sow-054 and format them as netCDF data files that are described in [AP-1, page 88]. SYS.0071<br />

DDS.0020<br />

6 sow-055 HR-DDS data granules will be delivered to the GHRSST-PP HR-DDS system. SYS.0071<br />

DDS.0110<br />

6 sow-056 A GHRSST-PP metadata record, will be created for each HR-DDS data file and sent to the GHRSST-PP global metadata<br />

repository as described in [AP-1, page 152].<br />

SYS.0071<br />

DDS.0010<br />

6 1.4.4 GHRSST-PP Match up database (MDR) data records<br />

6 sow-057 The <strong>Medspiration</strong> project shall prepare and deliver MDB records as described in [AP-1, page 54] for the GHRSST-PP MDB<br />

SYS.0060<br />

system.<br />

6 sow-058 The format of GHRSST-PP MDB records is described in [AP-1, page 130] and is compatible with other international<br />

SYS.0061<br />

satellite and in situ MDB systems.<br />

7 1.5 The <strong>Medspiration</strong> system scope<br />

7 sow-059 The <strong>Medspiration</strong> demonstration system must be capable of providing an operational service that delivers SST data SYS.0015<br />

© 2004 SOC Page 94 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

products to European users and the GHRSST-PP GDAC.<br />

7 sow-060 It must be able to ingest, process and disseminate large volumes of satellite and in situ data in near real time. SYS.0015<br />

DIS.0020<br />

DIS.0030<br />

7 sow-061 It must also be capable of real time ftp push and pull high-volume data delivery services [AP-1]. DIS.0110<br />

DIS.0120<br />

7 sow-062 The scope of the <strong>Medspiration</strong> system should consider the data streams and data volumes outlined in [AP-2], Table 4.2.1. SYS.0030<br />

7 sow-063 Figure 1.5.1 shows a functional decomposition diagram of the major data processing activities within the <strong>Medspiration</strong><br />

SRD Fig 2.4<br />

9 1.7 Prerequisites<br />

project together with associated input and output products.<br />

9 The following key issues shall be taken into consideration by the Contractor when executing the work:<br />

9 sow-064 • The GHRSST-PP and <strong>Medspiration</strong> is focussed on the provision of an operational near real time generation of SST data<br />

SYS.0015<br />

products.<br />

9 sow-065 Timeliness of data production, dissemination SYS.0015<br />

9 sow-066 and reliability of the <strong>Medspiration</strong> software system are critical to the success of the service. DES.SYS.0030<br />

CTR.0110<br />

9 sow-067 • Where possible the <strong>Medspiration</strong> shall rely on the use of configuration files [lookup tables] and implement configurable<br />

code [AP-4, page 13] that allows the processing model to be tuned (e.g., based on experience decorrelation length scales<br />

and various input parameters are expected to evolve) without the need to modify the <strong>Medspiration</strong> system software.<br />

DES.SYS.0040<br />

EXP.DDS.0010<br />

EXP.DDS.0020<br />

EXP.DDS.0030<br />

EXP.DDS.0040<br />

9 sow-068 • The optimal interpolation scheme used to derive L4 SSTfnd data products is expected to evolve as more experience is SYS.0082<br />

© 2004 SOC Page 95 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

gained through research and development within the GHRSST-PP.<br />

9 sow-069 The <strong>Medspiration</strong> L4 analysis system shall therefore be built in a modular manner (e.g. Mediterranean sea first) that allows<br />

for a limited amount of software adjustment without affecting the remainder of the <strong>Medspiration</strong> system.<br />

9 sow-070 • The <strong>Medspiration</strong> software shall be developed in a robust manner that is not dependent on any single data stream (which<br />

may not be available due to failure)<br />

DES.SYS.0010<br />

DES.SYS.0020<br />

DES.SYS.0060<br />

EXP.L2P.0010<br />

9 sow-071 and be capable of ingesting new data streams as they become available with minimum additional software development. DES.SYS.0050<br />

EXP.L2P.0010<br />

9 sow-072 • The <strong>Medspiration</strong> service shall be promoted. NASR<br />

9 sow-073 Publicity material NASR<br />

9 sow-074 including scientific publication, if generated, NASR<br />

9 sow-075 shall be directed towards non earth observational specialists with emphasis toward the application of <strong>Medspiration</strong> data<br />

NASR<br />

products and services.<br />

9 sow-076 • The project shall be undertaken by a European consortium containing partners familiar with: NASR<br />

9 sow-077 • Large scale NRT software and system development NASR<br />

9 sow-078 • Operational dissemination of satellite data sets, NASR<br />

10 2.1 Customer furnished Items (CFI)<br />

10 sow-079 A GHRSST-PP Reference Test <strong>Data</strong> Set (TDS) provided by the GHRSST-PP will be used to develop and test services<br />

CFI.SYS.0010<br />

10 The TDS will include the following:<br />

implemented in each GHRSST-PP RDAC. The intention is to provide this data set as a DVD.<br />

10 sow-080 • CFI 1: 2 days of L2 data files having global coverage from each L2 satellite data stream considered by the GHRSST-PP<br />

CFI.SYS.0010<br />

as defined in [AP-1, Appendix 3]. Both polar orbiting and geostationary observations will be included.<br />

10 sow-081 • CFI 2: 1 week of in situ data straddling the same period as the data sets described in Part 1. CFI.SYS.0020<br />

© 2004 SOC Page 96 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

10 sow-082 • CFI 3: An example GHRSST-PP Match-Up <strong>Data</strong>base7 with entries for each sensor [AP-1]. CFI.SYS.0030<br />

10 sow-083 • CFI 4: Example GHRSST-PP MDB data records, netCDF format L2P, SSTfnd and HR-DDS CFI.SYS.0040<br />

10 sow-084 granule data files and associated metadata records. CFI.SYS.0040<br />

10 sow-085 The GHRSST-PP TDS DVD will be available to the Contractor at the <strong>Medspiration</strong> project KO. CFI.SYS.0010<br />

10 sow-086 SSES files will be provided every month as a CFI by the GHRSST-PP Science Team coordinated by the International<br />

CFI.L2P.0020<br />

GHRSST-PPO Project Office and made available to the Contractor through ESA.<br />

10 2.2 Work breakdown<br />

10 sow-087 The <strong>Medspiration</strong> project will be subdivided into two distinct phases: NASR<br />

10 sow-088 • Phase 1: <strong>System</strong> Prototype Development (12 months) following the procedures laid out in [AP-3] see PMP<br />

10 sow-089 • Phase 2: An update of the system (3 months) and a sustained Service Demonstration (12 months including the system<br />

see PMP<br />

update phase).<br />

10 sow-090 At all stages of the project, ESA and the GHRSST-PP will monitor the progress of the project. At the end of Phase 1, an<br />

see PMP<br />

Acceptance Review will take place to assess the prototype system and a decision to continue into Phase 2 of the project<br />

will be taken. A complete service assessment will take place at the end of Phase 2.<br />

10 2.3 Project management<br />

10 sow-091 A Project Management Plan (PMP) shall be delivered by the Contractor at KO. PMP<br />

10 sow-092 Monthly report shall be delivered to ESRIN the first working day of each month by E-mail. PRP<br />

10 sow-093 Progress meetings shall be held according to the following timetable: NASR<br />

10 • KO: ESRIN<br />

10 • KO + 2 month: SRR, Contractor<br />

10 • KO + 4 Month: PDR, ESRIN<br />

10 • KO + 6 Months: CDR, ESRIN<br />

© 2004 SOC Page 97 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

10 • KO + 9 Months: QR, ESRIN<br />

10 • KO + 12 months: AR, contractor<br />

10 • KO + 15 months: <strong>System</strong> update, ESRIN<br />

10 • KO + 18 months: <strong>User</strong> feedbacks, Contractor<br />

10 • KO + 24 months, <strong>Medspiration</strong> Symposium, ESRIN<br />

<strong>User</strong> <strong>Requirements</strong> <strong>Document</strong><br />

Background to GHRSST and its requirements<br />

5 URD-001 The Global Ocean <strong>Data</strong> Assimilation Experiment (GODAE) was initiated in 1997 by the Ocean Observing Panel for Climate<br />

NASR<br />

(OOPC), in concert with the Global Climate Observing <strong>System</strong> (GCOS) and the Global Ocean Observing <strong>System</strong> (GOOS),<br />

as an experiment in which a comprehensive integrated observing system will be established and operated for several<br />

years. From 2003 to 2007 GODAE will provide a global system of observations, communications, modelling and<br />

assimilation which will deliver regular, comprehensive information on the state of the oceans. Information generated by<br />

GODAE will be made widely available for operational applications in regional and local situations calling for timely<br />

knowledge and forecasting of the ocean state.<br />

5 1 URD-002 Sea surface temperature (SST) measured from Earth Observation Satellites in considerable spatial detail and at high<br />

NASR<br />

frequency, is increasingly required for use in the context of operational monitoring and forecasting of the ocean, for<br />

assimilation into coupled ocean-atmosphere model systems and for applications in short-term numerical weather prediction<br />

and longer term climate change detection. The international GODAE steering committee focused attention on the centrality<br />

of SST measurements to the success of GODAE by initiating the GODAE High Resolution SST Pilot Project (GHRSST-<br />

PP).<br />

5 7 URD-003 The purpose of the GHRSST-PP is to develop an operational demonstration system that will deliver a new generation of NASR<br />

© 2004 SOC Page 98 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

global coverage high-resolution (better than 10 km and ~6 hourly) SST data products<br />

5 URD-004 GHRSST-PP data products will be derived by combining complementary Level-2 (L2) satellite and in situ observations in<br />

NASR<br />

real time to improve spatial coverage, temporal resolution, cross sensor calibration stability and SST product accuracy.<br />

5 11 URD-005 A full description of the GHRSST-PP project is given in the GHRSST-PP Development and Implementation Plan (GDIP) NASR<br />

5 5 URD-006 the GHRSST-PP is designed as a distributed processing system. NASR<br />

6 1 URD-007 The strategic implementation concept of regional/global task sharing (RGTS) by regional data product assembly centres<br />

NASR<br />

(RDAC) interconnected by dedicated, fast data connections has been adopted to underpin the practical implementation of<br />

the GHRSST-PP data processing system outlined in Figure 1.<br />

6 4 URD-008 RDAC ingest, quality control and merge together existing L2 satellite and in situ data sources to generate regional area<br />

NASR<br />

coverage quality controlled data products in real-time.<br />

6 6 URD-009 Every 6 hours, these data products are consolidated at RDAC and sent to a Global <strong>Data</strong> Analysis Centre (GDAC) NASR<br />

6 7 URD-010 where they are integrated together with other RDAC data streams having an identical data format but covering different<br />

NASR<br />

geographical areas to provide global coverage data products.<br />

6 9 URD-011 The GHRSST-PP Development and implementation Plan [RD-2] provides a complete description of the RGTS system and<br />

NASR<br />

the GHRSST-PP.<br />

6 1 URD-012 The combined Detailed Processing Model (DPM) 1 and Input/Output <strong>Data</strong> Definition document (IODD) [RD-3], developed<br />

NASR<br />

by the International GHRSST-PP Science Team, provide a complete technical description of the data processing tasks<br />

required to generate GHRSST-PP data products. Note that the GHRSST-PP Processing Model will be discussed in detail<br />

at the 4th GHRSST-PP workshop, to be held in September 2003 at JPL, USA. Following this meeting, changes to the<br />

current version of the GPM (the <strong>Medspiration</strong> DPM) are anticipated. It is expected that members of the <strong>Medspiration</strong><br />

consortium will attend the 4th GHRSST-PP workshop.<br />

6 1 URD-013 Before SST data originating from different sources can be properly assimilated into ocean model systems or analysed NASR<br />

© 2004 SOC Page 99 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

together to provide new SST data products that capitalise on their synergy, a bias and rms. error estimate is required for<br />

each [pixel] measurement. This is a fundamental component of the GHRSST-PP processing model.<br />

6 4 URD-014 Error estimates for each L2 SST measurement are based on the statistical analysis of a match up database (MDB)<br />

NASR<br />

containing satellite and in situ observations.<br />

6 5 URD-015 A quantitative confidence value is derived that is then used to assign statistical bias and rms. Error estimate to each<br />

NASR<br />

measurement.<br />

6 7 URD-016 The outputs of this process are L2 pre-processed (L2P) products for each data file ingested by the processor. NASR<br />

6 8 URD-017 Operational users that have requested [RD-13] these sparse grid SST data products may access L2P data products<br />

NASR<br />

directly in real time.<br />

6 URD-018 However, the majority of SST data users request [RD-13] complete SST fields, free of gaps caused by clouds, rain or lack<br />

NASR<br />

of data coverage, at regular intervals of between 6 and 24 hours.<br />

6 URD-019 Some require an estimate of the sub-surface SST field that cannot be directly measured using current satellite techniques NASR<br />

6 URD-020 whereas others require an estimate of the surface skin temperature that is in direct contact with the overlying atmosphere<br />

NASR<br />

and subject to considerable diurnal variability.<br />

6 URD-021 To address these needs, L2P data are used in an optimal interpolation scheme (e.g., Reynolds and Smith, 1994; Reynolds<br />

NASR<br />

et al, 2002; Guan and Kawamura, 2003) that is designed to<br />

6 URD-022 (a) account for differences in spatial and temporal sampling characteristics of each data stream, NASR<br />

6 URD-023 (b) account for gaps in coverage due to the presence of cloud, rain or lack of data, NASR<br />

6 URD-024 (c) account for SST diurnal variability and retrieve an estimate of the subsurface temperature field (referred to as the<br />

NASR<br />

foundation temperature, SSTfnd2),<br />

6 URD-025 (d) derive a measure of diurnal variability to accompany the SSTfnd estimate and NASR<br />

6 URD-026 (e) retrieve an estimate of the skin temperature of the ocean (SSTskin). NASR<br />

© 2004 SOC Page 100 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

6 URD-027 L2P and SSTfnd SST products have been designed to satisfy the requirements of operational ocean forecast and<br />

SR entry<br />

NASR<br />

prediction systems.<br />

6 URD-028 The data generated and maintained at both RDAC and GDAC is interfaced by a suite of services and tools that collectively<br />

NASR<br />

provide user information and services (UIS).<br />

6 URD-029 These include an active web service where data may be viewed, ordered and accessed, an archive service, a<br />

NASR<br />

comprehensive metadata repository and a data product validation data base.<br />

6 URD-030 The UIS is intended to serve the needs of non-operational scientific user groups. NASR<br />

6 URD-031 In contrast the specific and often unique requirements of operational users are met by the GHRSST-PP Application <strong>User</strong><br />

NASR<br />

Services (AUS). Within the AUS, a number of systems (e.g., MERCATOR, FOAM, ECMWF) will access and utilise<br />

GHRSST-PP data products in a test mode to assess their usefulness and measure the impact of their use against existing<br />

operational practices.<br />

6 URD-032 Only through this kind of user feedback can a sustainable SST data product service be justified NASR<br />

7 GHRSST-PP implementation n/a<br />

7 URD-033 The GHRSST-PP envisages that this work will be supported through Regional Task Sharing Projects that underpin each<br />

NASR<br />

RDAC centre. Four GHRSST-PP RDAC projects are currently in preparation and are fully described in the GDIP. These<br />

are:<br />

7 URD-034 1. The <strong>Medspiration</strong> project (MSP) serving the Atlantic, Mediterranean and European Coastal seas area sponsored by the<br />

NASR<br />

European Space Agency.<br />

7 URD-035 2. The New Generation SST project (NGSST) serving the western Pacific area sponsored by the Japanese Space Agency. NASR<br />

7 URD-036 3. The SeasNET program of the Institut de Recherche pour le Development (IRD) serving the tropical oceans. The<br />

NASR<br />

SeasNET project is based in France.<br />

7 URD-037 4. The Production of Enhanced Multi-sensor SST analysis project (PEMSA) serving the regional needs of the American NASR<br />

© 2004 SOC Page 101 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

region.<br />

7 URD-038 An initial GHRSST-PP GDAC centre is in preparation at the Jet propulsion Laboratory Physical Oceanography <strong>Data</strong> Active<br />

NASR<br />

Archive Centre (PO.DAAC) that is linked to the US-GODAE data server system (http://www.usgodae.org). The GDAC will<br />

host the GHRSST-PP metadata repository that is used to document all data sets that are used within the GHRSST-PP as<br />

described in the GPM [RD-3].<br />

7 URD-039 The US-GODAE server is dedicated to serving large data sets to operational ocean models within the GODAE framework<br />

NASR<br />

and to other scientific users. It forms the main hub of the GODAE data sharing project and is well connected with the<br />

operational ocean modelling community in Europe.<br />

7 URD-040 It is expected that in the coming years an equivalent European GDAC centre may be established providing European<br />

NASR<br />

autonomy as a mirrored system.<br />

7 URD-041 In addition, the extensive experience and capability of the PO.DAAC will ensure that all GHRSST-PP users are provided<br />

NASR<br />

with excellent support, documentation and data access.<br />

7 GHRSST-PP management n/a<br />

7 URD-042 The planning, management, implementation and monitoring of all aspects of the GHRSST-PP are the responsibility of the<br />

NASR<br />

GHRSST-PP Science Team. As stated in the consensus statement of The Ocean Observing <strong>System</strong> for the 21st Century<br />

(Koblinsky and Smith, 2001), the GHRSST-PP Science Team has the implementation oversight for the high resolution SST<br />

analysis network component of an integrated global observation strategy. The GHRST-PP constitutes the first step towards<br />

such an integrated system of SST measurements.<br />

7 URD-043 The Science Team is formed as an international group having a broad experience in all aspects of the GHRSST-PP<br />

NASR<br />

activities and convenes at least once per year to review the progress of the project. It plays a particularly important role in<br />

the following areas:<br />

7 URD-044 1. Reviewing the evolution of GHRSST-PP objectives/goals, maintenance of the GHRSST-PP Strategic Plan and, NASR<br />

© 2004 SOC Page 102 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

GHRSST-PP Development and Implementation Plan (GDIP), the GHRSST-PP Processing Model (GPM);<br />

7 URD-045 2. Coordination of the international consortium that will undertake the development and implementation of the GHRSST-<br />

NASR<br />

PP, including its final transition into an operational system;<br />

7 URD-046 3. Providing advice and guidance on scientific and technical innovations relevant to the GHRSST-PP; NASR<br />

7 URD-047 4. Providing a formal body that can liase and interact with national agencies in order to implement the GHRSST-PP. NASR<br />

7 URD-048 Starting in September 2003, an international GHRSST-PP project office (GHRSST-PO) will become operational at the Met<br />

NASR<br />

Office, UK taking responsibility for the day-to-day coordination of all aspects of the project.<br />

7 1.1 <strong>Document</strong> Scope n/a<br />

7 URD-049 ESA has integrated a <strong>Data</strong> <strong>User</strong> <strong>Element</strong> (DUE) within the second phase of the Earth Observation Envelope Programme<br />

NASR<br />

(EOEP), which will cover the time frame 2003 to 2007. The DUE provides continuity for the activities that have to date<br />

been carried out under the <strong>Data</strong> <strong>User</strong> Programme (DUP).<br />

7 URD-050 Consultations with Member States during the preparation of the second period of the DUP indicated that a majority would<br />

NASR<br />

like DUP activities to be integrated within the EOEP.<br />

8 URD-051 The second period of DUP has therefore been limited to three years in order to facilitate the incorporation of its activities<br />

NASR<br />

within the second phase of the EOEP.<br />

8 URD-052 The activities initiated within the DUE provide a bridge between scientific and methodological investigations carried out<br />

NASR<br />

within AO projects and the more commercially oriented activities to be performed under market development, an element<br />

already part of the EOEP. A major benefit is that a single programmatic approach can be followed for both the Product &<br />

Service development and the market development activities.<br />

8 URD-053 This document is the <strong>User</strong> <strong>Requirements</strong> <strong>Document</strong> (URD) for a European operational service called <strong>Medspiration</strong> that will<br />

NASR<br />

generate and disseminate SST data products in the European area under the DUE for European users and the GHRSST-<br />

PP.<br />

© 2004 SOC Page 103 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

8 URD-054 During the preparation of the <strong>Medspiration</strong> URD, European user requests for GHRSST-PP SST data products have been<br />

SR entry<br />

NASR<br />

solicited and collated by members of the GHRSST-PP Science team coordinated by the GHRSST-PP Science Team<br />

Chairman.<br />

8 URD-055 Together, this group acts as an external component of the ESA Expert Support Laboratory (ESL). NASR<br />

8 URD-056 The <strong>Medspiration</strong> URD is the first of a series of background documents that define what data products will be produced by<br />

NASR<br />

<strong>Medspiration</strong>, the detailed processing model (DPM) that will be used (which in the case of <strong>Medspiration</strong> is the GHRSST-PP<br />

processing model), the input and output definitions for the service (IODD) and, the project validation plan (PVP) that will<br />

coordinate the validation of <strong>Medspiration</strong> data products and services.<br />

8 URD-057 Collectively these documents form the primary inputs to the ESA <strong>Data</strong> <strong>User</strong> <strong>Element</strong> project planning and implementation<br />

NASR<br />

process.<br />

8 1.2 Objectives of the <strong>Medspiration</strong> project<br />

8 URD-058 There is no body or project presently existing in Europe able to provide a European contribution to GHRSST-PP by fulfilling<br />

sys.0010<br />

the tasks of a European area RDAC (EURDAC). The aim of the <strong>Medspiration</strong> project is to provide this service to the<br />

GHRSST-PP and directly serve the interests of European users with dedicated SST data products.<br />

8 URD-059 <strong>Medspiration</strong> will implement a finite demonstration service for the European area based on the GHRSST-PP GPM [RD-3]<br />

which has been developed by the International GHRSST-PP Science Team.<br />

8 URD-060 The <strong>Medspiration</strong> service will deliver GHRSST-PP SST data products to specific operational and scientific users in the<br />

European area and to the GHRSST-PP in real time and as an ftp pull service to registered clients.<br />

sys.0020,<br />

sys.0010<br />

sys.0030,<br />

sys.0060,<br />

sys.0070<br />

DIS.0120<br />

CTR.0570<br />

8 URD-061 In addition, <strong>Medspiration</strong> will generate and deliver ultra-high resolution (~2km spatial resolution data products) for specific sys.0040<br />

© 2004 SOC Page 104 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

European users following the same methodology described in the GPM.<br />

8 URD-062 The <strong>Medspiration</strong> project will focus on the following objectives: see following<br />

8 URD-063 • The implementation of the GHRSST-PP data processing model to provide operational data products covering the Atlantic<br />

Ocean and European Seas to the GHRSST-PP and European users;<br />

sys.0020,<br />

sys.0030<br />

sys.0070<br />

8 URD-064 • The operational generation of ultra-high resolution data products (2-4km) for the Mediterranean Sea, and a finite<br />

sys.0040<br />

number/time period of user requested small regional areas of interest;<br />

8 URD-065 • The operational extraction and archive of GHRSST-PP Diagnostic <strong>Data</strong> Set (DDS) granules [RD-4] for the European area<br />

and European sensors (ENVISAT AATSR);<br />

SYS.0050<br />

DDS.0020<br />

ARC.0010<br />

8 URD-066 • Establish data product dissemination service including the operational delivery of SST products to specific European<br />

sys.0060<br />

operational users and to the GHRSST-PP GDAC;<br />

8 URD-067 • Conduct an evaluation of <strong>Medspiration</strong> data products and service (by solicited user feedback) including a proper<br />

validation of SST data products using high quality in situ observations provided by users that will form part of the GHRSST-<br />

PP satellite and in situ match up data base using existing regional in situ infrastructure.<br />

sys.0080<br />

sys.0090,<br />

sys.0100<br />

9 1.3 <strong>Document</strong> Content<br />

9 This document is organised into the following sections:<br />

9 • Part 1 is this introduction.<br />

9 • Part 2 provides a user profile for each of the primary users of the <strong>Medspiration</strong> service.<br />

9 • Part 3 provides a consolidated user request for <strong>Medspiration</strong> SST data products.<br />

9 • Part 4 provides a specification of user defined products and service requirements for the <strong>Medspiration</strong> service.<br />

9 1.4 Reference <strong>Document</strong>s<br />

© 2004 SOC Page 105 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

9 URD-O68 [RD-1] Donlon, C. J., The GODAE High Resolution Sea Surface Temperature Pilot Project: Strategy and Initial<br />

SR entry<br />

NASR<br />

Implementation Plan, available from the GHRSST-PP International Project Office, 2002.<br />

9 URD-069 [RD-2] Donlon, C. J., The GODAE High Resolution Sea Surface Temperature Pilot Project Development and<br />

NASR<br />

Implementation Plan (GDIP), available from the GHRSST-PP International Project Office, 2003b.<br />

9 URD-070 [RD-3] Donlon, C.J. et al., The Recommended GHRSST-PP In situ and Satellite Integration Processing Model Version 1<br />

NASR<br />

(ISDI-PM-v1), available from the GHRSST-PP International Project Office, 2003.<br />

9 URD-071 [RD-4] Pinnock, S. and C. J. Donlon, Implementation of the GHRSST-PP High Resolution Diagnostic <strong>Data</strong> Set, GHRSST-<br />

NASR<br />

PP <strong>Document</strong> reference GHRSST/14, available from the GHRSST-PP International Project Office, 2003.<br />

9 URD-072 [RD-5] Casey, K., Draft Report for the GHRSST-PP Reanalysis Technical Advisory Group (RAN-TAG), available from the<br />

NASR<br />

GHRSST-PP International Project Office, 2003.<br />

11 URD-073 2 GHRSST-PP <strong>User</strong> Profiles The following sections provide a summary user profile for the major users of SST data<br />

EXP.SYS.0030<br />

products in the European area. For each user, a short description of the main SST application is provided followed by a<br />

summary of the SST data product(s) that are currently used if appropriate. Finally, each user has specified the SST<br />

products that they would ideally like to use in their applications. Based on a consolidation of all user profiles, a number of<br />

optimal SST data products are specified in Section 3 that balance the needs of each user group against the need to<br />

provide an affordable near real-time service.<br />

33 3 Consolidated <strong>User</strong> <strong>Requirements</strong> for <strong>Medspiration</strong><br />

33 URD-074 The user requirements described in Section 2 of this document are broad and encompass a wide range of activities,<br />

NASR<br />

demand a large number of different data formats, delivery type and spatio-temporal resolution. In this section of the<br />

<strong>Medspiration</strong> URD these requirements are considered together and consolidated into a finite number of SST data products<br />

that can be used by the widest community of users without significantly compromising any application.<br />

33 The following criteria were used to consolidate <strong>Medspiration</strong> user requirements: NASR<br />

© 2004 SOC Page 106 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

33 URD-075 • The requirements of the GHRSST-PP NASR<br />

33 URD-076 • Number of individual user requests NASR<br />

33 URD-077 • The long term usefulness of the products NASR<br />

33 URD-078 • <strong>Data</strong> product coverage and spatial characteristics NASR<br />

33 URD-079 • <strong>Data</strong> product type and accuracy NASR<br />

33 URD-080 • Temporal characteristics NASR<br />

33 URD-081 • <strong>Data</strong> product format NASR<br />

33 URD-082 • <strong>Data</strong> product delivery and dissemination NASR<br />

33 3.1 <strong>Medspiration</strong> data product coverage<br />

33 3.1.1 Global coverage<br />

33 URD-083 European area users requiring global coverage SST data products should these data from the GHRSST-PP GDAC. It is<br />

not foreseen that <strong>Medspiration</strong> will produce global coverage SST data products. Clearly though, if GHRSST-PP is to<br />

generate a real time global coverage SST data product it is important that <strong>Medspiration</strong> delivers L2P SST data products in<br />

a timely manner and fulfil the role of a GHRSST-PP RDAC under the GHRSST-PP RGTS implementation model.<br />

SYS.0015,<br />

SYS.0050<br />

DIS.0120<br />

DIS.0110<br />

33 3.1.2 European GHRSST-PP RDAC coverage (EURDAC)<br />

33 URD-084 The core of <strong>Medspiration</strong> will be to generate SST products over the Atlantic Ocean and adjoining Seas which is influential<br />

for local forecasting activities in the European area.<br />

34 URD-085 The Atlantic product should encompass the area 70S to 90N (to the ice limit) and from 100W to 45E (to include the<br />

adjacent seas) at a resolution of 1/12º (~10km) as shown in Figure 3.1.2.1. At this scale all data input sources used by<br />

SYS.0010,<br />

SYS.0050<br />

SYS.0051<br />

ING.0130<br />

GHRSST-PP are possible.<br />

34 URD-086 This specification corresponds to the “regional coverage” that is envisaged in the GHRSST Strategic Plan for a European<br />

NASR<br />

RDAC<br />

© 2004 SOC Page 107 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

34 URD-087 and it encompasses the EUMETSAT Ocean and Sea Ice Satellite and Application Facility (SAFOI) Atlantic region that will<br />

SR entry<br />

NASR<br />

provide a significant data input to the <strong>Medspiration</strong> project.<br />

34 T.3.1.2.1 URD-088 Table 3.1.2.1 describes the consolidated user request for GHRSST-PP RDAC SST data products.<br />

34 T.3.1.2.1 URD-089 GHRSST-PP EURDAC. L2P: 6 hourly SYS.0050<br />

34 T.3.1.2.1 URD-090 EURDAC (EU users) SSTfnd: daily analysis Out of scope<br />

34 T.3.1.2.1 URD-091 GHRSST-PP EURDAC GHRSST-PP DDS data granules SYS.0070<br />

34 T.3.1.2.1 URD-092 EURDAC MDB data records SYS.0060<br />

34 3.1.3 <strong>Medspiration</strong> Ultra-High (UHR) resolution coverage<br />

34 URD-093 The potential zones of interest for ultra-high resolution SST data product coverage are the Mediterranean Sea, the Black<br />

Sea, the Baltic Sea, the North Sea, the Greenland Sea and the Gulf of Mexico. These areas correspond to particular<br />

EXP.SYS.0030<br />

SYS.0080<br />

interests of the European countries for modelling ocean and air-sea interaction processes at fine resolution. In these cases<br />

the nominal space resolution is 2-3 km.<br />

34 URD-094 For UHR products, the use of Microwave SST data as part of the input data is excluded within 50 km of the coast (this is a<br />

minimum exclusion due to side-lobe contamination).<br />

CFD.0100<br />

CFD.0110<br />

CFD.0120<br />

34 URD-095 Figures 3.1.3.1-3 describe the areas requested for <strong>Medspiration</strong> UHR data coverage data products. see below<br />

34 F.3.1.3.1 URD-096 Figure 3.1.3.1 The <strong>Medspiration</strong> Mediterranean Sea product grid domain will cover from 30N to 46N and from 6W to 36.5E<br />

at 2 km spatial resolution.<br />

35 F.3.1.3.2 URD-097 Figure 3.1.3.2 The <strong>Medspiration</strong> Nordic Seas product grid domain will cover from 48N to 75N and from 12W to 70E at 2 km<br />

SYS.0081<br />

ING.0130<br />

NASR<br />

spatial resolution.<br />

35 F.3.1.3.3 URD-098 Figure 3.1.3.3 The <strong>Medspiration</strong> Black Sea product grid domain will cover from 40N to 48N and from 27E to 42E at 2 km<br />

NASR<br />

spatial resolution.<br />

© 2004 SOC Page 108 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

35 URD-099 Table 3.1.3.1 describes the consolidate user request for <strong>Medspiration</strong> data product ultra high resolution coverage based on<br />

SR entry<br />

PER.SYS.0080<br />

the user requirements set out in Section 2 of this document. All SST data products are requested to an accuracy of<br />

0.4K or better. The ECMWF request for data products having an accuracy of 0.1K is exceptionally demanding and should<br />

be considered as an “ultimate” target for satellite derived SST data products.<br />

36 T. 3.1.3.1 URD-100 Mediterranean Sea; L2P,SSTskin and SSTfnd SYS.0050<br />

SYS.0080<br />

36 T. 3.1.3.1 URD-101 Black Sea; L2P and`SSTfnd NASR<br />

36 T. 3.1.3.1 URD-102 Nordic seas; L2P, SSTskin, & SSTfnd NASR<br />

37 4 Product and Service Specifications<br />

37 4.1 Processing system<br />

37 URD-103 The system should generate the SST products as described in Tables 3.1.2.1 and 3.1.3.1. SYS.0050 to<br />

SYS.0083<br />

37 URD-104 In addition, <strong>Medspiration</strong> must contribute to the building of the GHRSST-PP Diagnostic <strong>Data</strong> Set as defined in [RD-2] and<br />

SYS.0070<br />

[RD-4].<br />

37 URD-105 The system should be extensible and flexible in order to accommodate new satellite sensors and the potential failure of<br />

EXP.SYS.0020<br />

others.<br />

37 URD-106 The <strong>Medspiration</strong> processing system must conform to the GHRSST-PP processing model [RD-3] if regional data products<br />

SYS.0020<br />

generated at different GHRSST-PP RDAC are to be effectively combined together as a single global coverage data<br />

product at GHRSST-PP GDAC.<br />

37 4.2 Input data definitions<br />

37 URD-107 The input data sets to the <strong>Medspiration</strong> data processing chain will be level 2 SST, wind speed and surface solar irradiance<br />

SYS.0030<br />

(SSI) data products that are available in real time.<br />

© 2004 SOC Page 109 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

37 URD-108 Table 4.2.1 provides a summary of the data streams that should be considered by the <strong>Medspiration</strong> data processing<br />

SR entry<br />

SYS.0030<br />

system. All of these data sets are available in real time and have either no data usage agreement or a special data<br />

agreement that has been negotiated for the GHRSST-PP project.<br />

37 URD-109 In some cases data providers have stipulated that all data must be used for the sole purpose of the GHRSST-PP. As<br />

SYS.0040<br />

37 4.3 <strong>Data</strong> products<br />

<strong>Medspiration</strong> intends to generate and disseminate combined L2 data products (as opposed to the native L2 data streams<br />

themselves) there are no data product dissemination limitations associated with use of these L2 data streams.<br />

3 URD-110 The <strong>Medspiration</strong> Project must be capable of insertion into the implementation strategy of GHRSSTPP, becoming a<br />

SYS.0010<br />

Regional <strong>Data</strong> Assembly Centre and capable of contributing regional coverage GHRSST-PP EURDAC area L2P data<br />

products (as defined in the GPM) to Global <strong>Data</strong> Analysis Centres in real time. In addition, European area coverage L4<br />

analysed data products should be produced to satisfy European user needs. The required data products are:<br />

37 URD-111 • Collated products (L2P) for each input sensor. L2P data products should be made available to European users in real<br />

time via special agreement. L2P data products for a single sensor should be consolidated to provide a single L2P data file<br />

SYS.0050<br />

SYS.0100<br />

for each sensor at 6 hourly intervals. These data should also be passed to the GHRSST-PP GDAC in near real time.<br />

37 URD-112 • A static archive of L2P data products that can (a) serve the needs of the GHRSST-PP reanalysis project and (b) be used<br />

as the starting point for any reprocessing effort if required. Because the data quality flagging for L2P data is experimental,<br />

it is important for the pilot project phase of GHRSST to be able to revisit the flagging criteria following quality analysis,<br />

SYS.0090<br />

SYS.2040<br />

ARC.0010<br />

which is facilitated by preserving a static archive.<br />

37 URD-113 • Level 4 analysed SSTfnd for Ultra High Resolution areas. The analysed product is the result of combining all available<br />

SYS.0080<br />

SST observations using an optimal interpolation (OI) scheme described in [RD-3].<br />

37 URD-114 • GHRSST-PP diagnostic data set granules as described in [RD-4] SYS.0070<br />

SYS.0071<br />

© 2004 SOC Page 110 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

37 URD-115 • Satellite and in situ match up data sets to be entered into the GHRSST-PP MDR system according to [RD-3] as available. SYS.0060<br />

SYS.0061<br />

37 URD-116 Several intermediate data products may be produced but these will be internal to the <strong>Medspiration</strong> project and will not be<br />

NASR<br />

available to the user community.<br />

38 T.4.2.1 URD-117 ATS_NR__2P: AATSR, 1km, (SSTskin) SYS.0030<br />

38 T.4.2.1 URD-118 ATS_MET_2P: AATS,R 10 arc min, (SSTskin) SYS.0030<br />

38 T.4.2.1 URD-119 AVHRR16-L: AVHRR NOAA16, 9km, (MCSST) SYS.0030<br />

38 T.4.2.1 URD-120 NARSST: AVHRR NOAA16+17, 2km, (SSTsubskin) SYS.0030<br />

38 T.4.2.1 URD-121 AVHRR17-L: AVHRR NOAA17, 9km, (MCSST) SYS.0030<br />

38 T.4.2.1 URD-122 ATLSST: GOESE+MSG SEVIR,I 0.1º, (SSTsubskin,) SYS.0030<br />

38 T.4.2.1 URD-123 ATLSSI: GOESE+MSG SEVIR,I 0.1º, (SSI) SYS.0030<br />

38 T.4.2.1 URD-124 AMSRE: AMSR-E, 25 km grid, (SSTsubskin, Wind speed, water vapour, cloud liquid water vapour, rain rate) SYS.0030<br />

38 T.4.2.1 URD-125 AMSR: AMSR, 25 km grid, (SSTsubskin) NASR<br />

39 T.4.2.1 URD-126 TMI: TRMM-TMI, 25 km grid, (SSTsubskin, Wind speed, water vapour, cloud liquid water vapour, rain rate) SYS.0030<br />

39 T.4.2.1 URD-127 SSMI: SSM/I F13, F14 and F15, (25 km grid, Wind speed, water vapour, cloud liquid water vapour, rain rate) SYS.0030<br />

39 T.4.2.1 URD-128 ECMWF NWP Wind speed and SSI, 0.5º grid, (Wind speed, heat fluxes and SSI) SYS.0030<br />

39 T.4.2.1 URD-129 CORIOLIS: In situ SST observations, Point source, (In situ ocean and atmospheric parameters) SYS.2060<br />

39 T.4.2.1 URD-130 MEDS: In situ observations, Point source, (In situ ocean and atmospheric parameters) SYS.2060<br />

40 4.4 Product dissemination<br />

40 URD-131 It is expected that <strong>Medspiration</strong> will provide an operationally sustained service for a 12 month period following a 12 month<br />

SYS.0015<br />

preparation and prototyping phase.<br />

40 URD-132 It is recommended that <strong>Medspiration</strong> data products should be made available for public release to obtain the widest SYS.0110<br />

© 2004 SOC Page 111 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

possible application, user feedback and validation data input at the earliest opportunity.<br />

40 URD-133 At all stages of the project, close user collaboration is required especially with the GHRSST-PP Science Team. SYS.0020<br />

SYS.0021 (See<br />

also CFIs in<br />

chapter 12)<br />

GDS (version 1, revision 1.1; 7 Jan 2004)<br />

1.0 Introduction<br />

17 GDS-001 The primary aim of the Global Ocean <strong>Data</strong> Assimilation Experiment (GODAE) High Resolution Sea Surface Temperature<br />

NASR<br />

Pilot Project (GHRSST-PP) is to develop and operate an operational demonstration system that will deliver high-resolution<br />

(better than 10 km and ~24 hours) global coverage SST data products for the diverse needs of GODAE and the wider<br />

scientific community. A new generation of SST data products will be derived and served to the user community by<br />

combining complementary satellite and in situ SST observations in near real time. A full description of the GHRSST-PP<br />

project is given in the GHRSST-PP Development and Implementation Plan (GDIP) which can be obtained from the<br />

GHRSST-PP project web server located at http://www.ghrsst-pp.org.<br />

17 GDS-002 Figure 1.1 provides a simplified schematic overview of the GDIP. The GHRSST-PP is based on a distributed system in<br />

NASR<br />

which the data processing operations that are necessary to operationally generate and distribute high resolution SST data<br />

sets having global coverage are shared by Regional <strong>Data</strong> Assembly Centres (RDAC). RDAC ingest, quality control and<br />

merge existing satellite and in situ SST data sources that are then used together to generate regional coverage quality<br />

controlled SST data products to the same specification (called L2P products), in real-time. RDAC data products are then<br />

assembled together at Global <strong>Data</strong> Analysis Centres (GDAC) where they are integrated and analysed to provide L4 global<br />

coverage data products.<br />

© 2004 SOC Page 112 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

18 GDS-003 Each RDAC implements a processing system described in the GDS within a regionally funded project that is common to all<br />

SR entry<br />

NASR<br />

RDAC. Five GHRSST-PP RDAC projects are currently in preparation and are fully described in the GDIP. These are:<br />

18 GDS-004 1. The New Generations SST project (NGSST) serving the western Pacific area replaced by a approximately defined by<br />

NASR<br />

the Geostationary Meteorological Satellite (GMS, now replaced by GOES-9) footprint. The NGSST project is based in<br />

Japan.<br />

18 GDS-005 2. The European <strong>Medspiration</strong> Project (MSP) serving the Atlantic area and European shelf seas. SYS.0010<br />

18 GDS-006 3. The Ocean Forecasting Australia (OFA) BlueLink project serving the regional needs of the Australian region. NASR<br />

18 GDS-007 4. The Survey of the Environment Assisted by Satellite (SEASnet) program of the IDD serving the tropical oceans. The<br />

NASR<br />

SEASnet project is based in France.<br />

18 GDS-008 5. A project serving the SST needs of the USA under the US National Ocean Partnership Program (NOPP) under the<br />

NASR<br />

general title of ‘SST for GODAE’.<br />

18 1 GDS-009 A GDAC centre is currently in preparation that will be implemented as a joint system by the Jet propulsion Laboratory<br />

NASR<br />

Physical Oceanography <strong>Data</strong> Active Archive Center (PO.DAAC) and the US-GODAE data server system at Monterey<br />

(http://www.usgodae.org).<br />

18 4 GDS-010 The PO.DAAC will be responsible for the data management of the GDAC including metadata, data serving and user<br />

NASR<br />

interactions and US-GODAE will be responsible for the archive of GHRSST-PP data sets and the production of a global<br />

analysed SST fields initially using the US-Navy system.<br />

18 7 GDS-011 The USGODAE server is dedicated to serving large data sets to operational ocean models. In addition, the extensive<br />

NASR<br />

experience and capability of the PO.DAAC will ensure that GHRSST-PP scientific users are provided with excellent<br />

support, documentation and data access.<br />

18 GDS-012 In parallel, a second backup GDAC will be developed as part of the European Marine Environment and Security in the<br />

SYS.0010<br />

European Area (MERSEA) initiative (see http://www.ifremer.fr/merseaip/objectives.htm).The EU RDAC project<br />

© 2004 SOC Page 113 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

<strong>Medspiration</strong> forms the initial pilot study for the MERSEA GDAC which will provide an operational backup processing<br />

system to the US GDAC described above. MERSEA aims to develop a European system for operational monitoring and<br />

forecasting on global and regional scales of the ocean physics, biogeochemistry and ecosystems. The prediction time<br />

scales of interest extend from days to months. This integrated system will be the Ocean component of the future European<br />

Global Monitoring for Environment and Security (GMES) system. An initial configuration of the MERSEA EU GDAC is<br />

expected in 2006/7.<br />

18 GDS-013 Figure 1.2 shows the configuration of the GHRSST-PP Regional/Global task sharing system that will implement the GDS.<br />

In addition to the global processor system, the GDAC also hosts the GDS (international) system-wide metadata repository<br />

(MMR), matchup database system (MDB), High Resolution Diagnostic <strong>Data</strong> Set (HR-DDS) and main product archive.<br />

RDAC will also maintain their own specific archives of L2P data products, HR-DDS, MDB and ultra-high resolution (~2km<br />

spatial grid) data products as required but the MMR will always reside at the GDAC.<br />

SYS.0010<br />

SYS.0060<br />

SYS.0070<br />

SYS.0090<br />

SYS.0100<br />

ARC.0010<br />

19 The GHRSST-PP <strong>Data</strong> Processing Specification (GDS)<br />

19 1 GDS-014 The GHRSST-PP <strong>Data</strong> Processing Specification (GDS) is a recommended common data processing specification that<br />

NASR<br />

should be implemented at each GHRSST-PP RDAC and GDAC. It defines clearly the input and output data specifications,<br />

data processing procedures, algorithms and data product file formats that are used within the GDS and are thus common<br />

to each GHRSST-PP RDAC and GDAC. This is a prerequisite if the GHRSST-PP Global/Regional task sharing<br />

implementation framework is to function efficiently.<br />

19 7 GDS-015 The GDS specifies in detail the minimum data processing that should be performed at each RDAC and GDAC centre to<br />

SYS.0020<br />

ensure that data and data products can be used and exchanged with confidence. For example, a common processing<br />

description is necessary to simplify documentation of data, facilitate exchange by sharing a common data format agreed by<br />

RDAC, GDAC and users, to avoid significant duplication of effort, to minimise reformatting of different data products<br />

© 2004 SOC Page 114 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

derived by RDAC and to ease the integration of RDAC data to provide global coverage data sets at GDAC centres. Lowlevel<br />

data products generated according to the GDS will form a direct input to the GHRSST-PP reanalysis project (RAN).<br />

These operationally produced data products will then be improved by using additional data that are only available in a<br />

delayed mode together with extensive quality control procedures. The GHRSST-PP RAN Strategy is described in Casey et<br />

al. (2003).<br />

19 GDS-016 GHRSST-PP RDAC centres must implement data processing procedures that account for specific aspects of regional<br />

SYS.0010<br />

coverage input data products (e.g., geostationary imagers). RDAC must also provide additional data products and services<br />

to satisfy regional user requirements (e.g., regionally specific analysed data products or ultrahigh resolution data products)<br />

that will require research and development of new analysis, quality control, and data provision procedures. Regional<br />

extensions to the GDS are both necessary and critical to the preservation of regional identity and the proper evolution of<br />

the GDS; without evolution, the data products and services provided by the GHRSST-PP are unlikely to satisfy developing<br />

user demands and concerns.<br />

20 GDS-017 It is not possible to specify exactly how the final version of the GDS will be implemented and operated without first having<br />

an initial ‘Version-1’ system in place to make an informed assessment. Neither is it desirable to have a GDS that is rigid<br />

SYS.0021<br />

DES.SYS.0040<br />

without the ability to innovate based on experience. The first version of the GDS (v1.0) is focused on an initial data<br />

processing specification with specific emphasis on implementing:<br />

20 GDS-018 (a) An operational data exchange and delivery system connecting data providers, RDAC, GDAC and user communities by<br />

SYS.0015<br />

2005,<br />

20 GDS-019 (b) A number of common format (L2P) sensor specific near real time SST data products SYS.0050<br />

20 GDS-020 (c) A suite of regional L4 data products based on different analysis procedures following current state-of-the-art knowledge<br />

SYS.0080<br />

that can be upgraded and refined based on experience. The analysis procedures should as far as possible capitalise on<br />

the synergy benefits of complementary measurements (accuracy, susceptibility to atmospheric clouds and aerosol loading,<br />

© 2004 SOC Page 115 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

spatial and temporal resolution etc.)<br />

20 GDS-021 (d) An initial Global analysis system at the GHRSST-PP GDAC, NASR<br />

20 GDS-022 (e) An initial implementation of the GHRSST-PP High Resolution Diagnostic <strong>Data</strong> Set (HR-DDS) system and, SYS.2080<br />

SYS.2081<br />

SYS.2082<br />

20 GDS-023 (f) A suite of initial test data products to be available to selected users at each RDAC and GDAC by mid 2004. CFI.SYS.0010<br />

20 GDS-024 The GDS is split into several work package (WP) sections that are supported by extensive technical Appendices. This<br />

NASR<br />

format provides a framework that preserves the readability of the processing specification while providing extensive<br />

technical references that can be easily maintained and updates without affecting the overall structure of the GDS. While<br />

the interface parameters for each work package will remain relatively static, considerable flexibility within a WP is<br />

maintained by this type of approach. Work packages develop a modular approach with clearly defined input and output<br />

parameters that greatly assist in the development of large multiinstitute/ national projects such as the GHRSST-PP.<br />

20 <strong>Document</strong> Scope<br />

20 GDS-025 This document is the GDS version 1 which is based on the First Report of the GHRSST-PP In situ and Satellite <strong>Data</strong><br />

NASR<br />

Integration Technical Advisory Group (ISDI-TAG, Wick et al, 2002) and many subsequent discussions at the Third<br />

GHRSST-PP Workshop, held at ESA/ESRIN, Frascati, Italy in December 2002 (Donlon, 2003b) and further refined at the<br />

4th GHRSST-PP Science Team Meeting, Los Angeles, USA in September 2003. It represents a consensus opinion of the<br />

GHRSST-PP community of how to pursue the optimal combination of satellite and in situ data streams within a globally<br />

distributed operational system to provide a new generation of global coverage SST data products. Much of the document is<br />

dedicated to issues of data exchange, management and operational considerations. For certain international groups<br />

(particularly the European <strong>Medspiration</strong> project), it provides a reference baseline that will be implemented and innovated<br />

by their project.<br />

© 2004 SOC Page 116 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

20 GDS-026 The GDS will evolve as each of the GHRSST-PP RDAC projects gain experience and knowledge throughout the Pilot<br />

SR entry<br />

NASR<br />

Project and a significant scientific upgrade of the processor is foreseen following the successful commission of the v1.0<br />

GDS. This is thus a working document written primarily for the RDAC and GDAC teams as they physically realise, develop<br />

and, refine the GHRSST-PP demonstration system. The document content will be constantly modified as initial approaches<br />

are tried and refined by the RDAC teams. But, in order to maintain consistency with funding agencies, published releases<br />

of the GDS will occur at regular intervals and represent reference baselines.<br />

21 7 GDS-027 These will carry an integer revision number (e.g., GDS-v1.0, GDS-v2.0 etc). As such, each may be considered a technical<br />

NASR<br />

reference manual for the GHRSST-PP for a given time window based on the best current knowledge of the GHRSST-PP<br />

community, the Science Team, data providers and RDAC/GDAC projects. The GDS will be modified, refined and updated<br />

and published electronically on a regular basis by the GHRSST-PP International Project Office (GHRSST-PO) as RDAC<br />

projects implement the GDS.<br />

23 2 Notations, conventions and definitions used by the GDS.<br />

23 GDS-028 The following sections describe the notations and conventions that are used throughout the GDS documentation. RDAC<br />

SYS.0020<br />

and GDAC implementation projects are expected to adhere to the nomenclature and style of the GDS in their own<br />

documentation as much as possible. This will greatly facilitate exchange of documentation between each centre.<br />

23 2.1 Flow diagram symbols<br />

GDS-029 Table 2.1.1 Symbols used in GDS flow and functional breakdown diagrams SYS.0020<br />

2.2 Definition of data processing levels<br />

GDS-030 The GDS uses the definitions provided in Table 2.2.1 when referring to data processing levels SYS.0020<br />

2.3 GDS data processing window specifications<br />

24 GDS-031 GDS data analysis processing activities are linked to Analysed Product Processing Window (APPW) periods within a 24<br />

L4A.0020<br />

hour period that are defined in Table 2.3.1. A cutoff time (Ω) must be specified after which L2P data are no longer eligible<br />

© 2004 SOC Page 117 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

for inclusion within the analysis. This must take account the time taken to process L2P data (DTime L2P ), the time required<br />

to perform the analysis (DT analysis ) itself and the Maximum desirable output time delay (T output ) using (Eqn. 2.3.1)<br />

2.4 GHRSST-PP definitions of sea surface temperature<br />

24 GDS-032 2.4.1 to 2.4.5 are physical definitions of temperature variables See SRD<br />

section2.1.1<br />

26 GDS-033 2.4.6. The diurnal cycle/variation of SST (DV) For the GDS, a diurnal cycle refers to changes in vertical and horizontal<br />

See<br />

SRD<br />

distribution of SST throughout a 24 hour period and thus includes warm stratified layers and cool skin effects. Cool skin<br />

section2.1.1<br />

effects are typically more pronounced at night due to radiative cooling of the sea surface but may also occur during the day<br />

when the wind is light following a significant rainfall that may leave a cool freshwater layer on the surface of the ocean<br />

2.4.7 GDS operational output grid specifications<br />

26 T.2.4.7.1 GDS-034 Table 2.4.7.1 describes the GDS preliminary output grid specification that is applicable to standard L4SSTfnd global<br />

Out of scope<br />

coverage data products produced by the GDS<br />

26 T.2.4.7.2 GDS-035 Ultra-high resolution SSTfnd data products (L4FNDUHR) will be similar to L4FND data products but produced on a smaller<br />

L4A.0050<br />

grid specification described in Table 2.4.7.2.<br />

27 T.2.4.7.3 GDS-036 High Resolution Diagnostic <strong>Data</strong> Set (HR-DDS) will be constructed by remapping L2P data streams onto a uniform grid as<br />

specified in Table 2.4.7.3.<br />

DDS.0090<br />

DDS.0030<br />

GDS-037 2.5.Missing <strong>Data</strong> Values In all GHRSST-PP data products missing values are denoted by the integer value -9999 Format<br />

GDS-038<br />

2.6 Sea ice value In all GHRSST-PP data products sea ice is denoted by the integer value 5555. The corresponding<br />

value for the fractional sea ice coverage can be found in the associated L2P and L4 product confidence data record.<br />

3.0 Overview of the GHRSST-PP Processing Specificationv1.0 (GDS)<br />

Format<br />

31 GDS-039 This section of the GDS provides a general overview of the processing model introducing how data will flow within the<br />

NASR<br />

processor, the processor input and output data definitions (IODD), and the work packages (WP) that will be used to<br />

© 2004 SOC Page 118 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

implement the processor. It may be used as a “map” of the GDS and identifies the interfaces between each specific work<br />

package. The work packages themselves are expanded in technical detail within subsequent sections and associated<br />

appendices of the GDS<br />

31 GDS-040 Figure 3.1.1 General overview of the GHRSST-PP GDS. Input data sources are shown in red, output parameters are<br />

NASR<br />

shown in green and GDS processing steps with a further breakdown are shown in light blue. <strong>Data</strong>base storage is shown in<br />

dark blue. For each output parameter a corresponding metadata record is generated and delivered to the GHRSST-PP<br />

Master Metadata Repository (MMR) system. The MMR system has been omitted to preserve clarity in this figure<br />

32 Summary description of the GDS<br />

32 GDS-041 The GDS is designed to produce SST data products that satisfy the requirements of operational ocean forecast and<br />

NASR<br />

prediction systems. Based on a user requirements survey (see the proceedings of the 3rd GHRSST-PP workshop and the<br />

<strong>Medspiration</strong> <strong>User</strong> <strong>Requirements</strong> <strong>Document</strong> for example, both available from the GHRSST-PO or http://www.ghrsstpp.org),<br />

the main requirements are<br />

32 GDS-042 1. SST data are required operationally in near real time and should ideally be available within 6 hours of data acquisition at<br />

SYS.0015<br />

the satellite platform.<br />

32 GDS-043 2. An error estimate (bias and standard deviation) and confidence data (such as an indication of atmospheric aerosol<br />

presence at the time of measurement) for each SST measurement including a bias and SD. is required by users and by<br />

L2P.0040<br />

L2P.0050<br />

the GHRSST-PP as a necessary precursor to data analysis.<br />

32 GDS-044 3. A global ocean coverage analysed SST data product providing a best measure of the SSTfnd is required at a minimum<br />

NASR<br />

interval of 24 hours having a maximum spatial grid size of 1/12° (latitude x longitude).<br />

32 GDS-045 4. Coastal modelling groups require an ultra-high spatial resolution SSTfnd and SSTskin data products on a grid cell size<br />

of 1-2km and a timeliness of ~6 hourly.<br />

SYS.0080<br />

L4A.0020<br />

32 Figure 3.1.1 shows a functional breakdown diagram of the GDS that identifies the flow of data, inputs and output channels SRD Fig 2.4<br />

© 2004 SOC Page 119 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

and data. The GDS is split into work package (WP) sections to aid development and understanding of the system. The<br />

major work package components of the processing specification are shown in light blue together with their associated input<br />

and output data parameters that are shown in red and green respectively. The flow of data and information between each<br />

work package is indicated by the direction of connecting arrows.<br />

32 GDS-046 The GDS data processing flow is designed to: see below<br />

32 GDS-047 (a) Ingest and QC input SST data sets. L2P.0010<br />

L2P.0020<br />

32 GDS-048 (b) Assign pixel confidence data, auxiliary data (wind speed, Surface Solar Irradiance (SSI) and, Aerosol Optical Depth<br />

(AOD)) to each SST measurement.<br />

32 GDS-049 (c) Assign error estimates (called Sensor Specific Error Statistics, SSES) to each SST measurement and format as a<br />

GHRSST-PP L2P output data file. L2P data sets should be available to the RDAC user community and GDAC processor<br />

L2P.0030<br />

L2P.0040<br />

L2P.0050<br />

PER.SYS.0010<br />

ideally within 6 hours of data reception at the satellite sensor.<br />

32 GDS-050 (d) Collocate L2P data records with in situ measurements and enter these data into the GHRSST-PP Match-up<br />

<strong>Data</strong>base (MDB).<br />

32 GDS-051 (e) Extract High Resolution Diagnostic <strong>Data</strong> Set (HR-DDS) data granules from L2P data and submit these data to the<br />

GHRSST-PP HR-DDS system.<br />

SYS.0060<br />

SYS.2060<br />

SYS.0070<br />

SYS.2080<br />

DDS.0030<br />

INT.DDS.0010<br />

32 GDS-052 (f) Consolidate L2P SST data sets every 24 hours and analyse all data within an Analysis Product Processing Window<br />

SYS.0080<br />

(APPW) to provide a ‘best estimate’ of the SSTfnd. Format analysed SST data sets as GHRSST-PP L4 data files and<br />

archive. L4 data products should be available to the user community no later than 12 hours following the end of an APPW.<br />

The analysis procedures, than must be developed, should aim to capitalise on the synergy benefits of complementary<br />

© 2004 SOC Page 120 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

infrared and microwave SST measurements with particular emphasis on data combinations that lead to better accuracy,<br />

minimise the impacts of cloud/rain contamination and atmospheric aerosol loading, optimise spatial and temporal sampling<br />

resolution and address diurnal variability amongst others.<br />

33 GDS-053 (g) Extract High Resolution Diagnostic <strong>Data</strong> Set (HR-DDS) data granules from L4 data and submit these data to the<br />

GHRSST-PP HR-DDS system.<br />

SYS.0070<br />

SYS.2080<br />

DDS.0040<br />

INT.DDS.0020<br />

33 GDS-054 (h) For each data processing operation, create unique metadata records and submit these to the GHRSST-PP Master<br />

Metadata Repository (MMR).<br />

SYS.0055<br />

CTR.0530<br />

33 GDS-055 (i) Archive L2P and L4 data in a GDAC archive for use in the GHRSST-PP Reanalysis Project (RAN). SYS.0090<br />

ARC.0010<br />

33 GDS-056 Referring to Figure 3.1.1, RDAC first ingest, in real time, regional coverage satellite and in situ SST and auxiliary data<br />

streams from a variety of different data providers (WP-ID1 in Figure 3.1.1).<br />

SYS.2010<br />

L2P.0010<br />

33 GDS-057 Each input satellite SST data stream is quality controlled to check for gross errors, consistency and timelines. SYS.2010<br />

L2P.0020<br />

33 GDS-058 In the case of MW SST data sets, the use of level-2 (L2) as a descriptor of the data processing level is incorrect as these<br />

NASR<br />

data are typically L3 gridded fields. Thus, in the context of L2P data descriptions, the input data are actually L2/L3 but in<br />

order to keep the terminology simple for L2P data files throughout the GDS the term L2P will be used for both L2 IR and L3<br />

MW SST data sets.<br />

33 GDS-059 <strong>Data</strong> that are unacceptable for further use in the GDS are rejected at this point. ARC.0010<br />

33 GDS-060 Each SST data file is then reformatted to a common GDS pre-processed data format (WP-ID2 in Figure 3.1.1) referred to<br />

L2P.0060<br />

as L2P and described in Appendix A1, that is designed to allow easy sharing of data across the GHRSST-PP RDAC,<br />

© 2004 SOC Page 121 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

provide a data product suitable for use by GHRSST-PP data analysis systems and ocean model data assimilation<br />

schemes, the conform to the GODAE data sharing project.<br />

33 GDS-061 L2P data products are the lowest-level, common format data products produced by the GHRSST-PP and provide the<br />

NASR<br />

“building blocks” for all other higher level data products.<br />

33 GDS-062 Each L2P data product contains original SST measurements together with confidence fields and an error estimate (called<br />

SSES and described ion the following paragraphs). L2P data products consist of the original SST measurements (that are<br />

not re-gridded or modified) together with additional confidence and error estimates.<br />

L2P.0030<br />

L2P.0040<br />

L2P.0050<br />

33 GDS-063 A separate L2P data product is produced for each sensor at each RDAC. L2P.0070<br />

33 GDS-064 Operational users that have requested this type of SST data product (see the <strong>Medspiration</strong> <strong>User</strong> <strong>Requirements</strong> <strong>Document</strong><br />

SYS.2050<br />

(URD), available from the GHRSST-PP Project Office) may access L2P data products directly in real time from RDAC and<br />

GDAC.<br />

33 GDS-065 In addition, L2P data products provide a direct input to the GHRSST-PP Reanalysis project (Casey et al., 2003) and form<br />

NASR<br />

the principal data archive for the GHRSST-PP.<br />

33 GDS-066 A bias and standard deviation (sd.) error estimate is required for each [pixel] measurement before SST data originating<br />

L2P.0050<br />

from different sources can be properly assimilated into ocean model systems or analysed together to provide new SST<br />

data products. Assigning error estimates to SST measurements is a fundamental requirement of the GDS. Two types of<br />

error assignment are possible within a L2P data set:<br />

33 GDS-067 1. Error estimates provided by a data provider as part of the L2/L4 input data L2P.0050<br />

33 GDS-068 2. Statistical error estimates calculated from a match up database containing satellite and drifting/moored buoy in situ SST<br />

L2P.0050<br />

observations.<br />

33 GDS-069 In the case of (2), there are clearly insufficient buoy observations to assign a unique error estimate for every satellite<br />

NASR<br />

measurement and instead, the available data spanning a particular time-window must be used to assign a statistical error<br />

© 2004 SOC Page 122 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

estimate. This is performed as a four stage process (WP-ID3 in Figure 3.1.1):<br />

34 GDS-070 1. A quantitative “confidence value”, having a value spanning 0 (no data, measurement is sea ice) to 5 (highest confidence<br />

in the SST measurement), is derived for each pixel measurement based primarily on an estimate of the surface wind<br />

L2P.0040<br />

CFD.0010<br />

speed, proximity to other cloud contaminated/rain contaminated pixels and, proximity to a SST reference climatology.<br />

Confidence values are optimized for likely errors from each sensor type (MW and IR).<br />

34 GDS-071 2. L2P observations are then matched to near contemporaneous quality-controlled in situ SST observations in situ data<br />

SYS.0060<br />

and the resulting data record (together with any additional auxiliary data) are stored in a relational database system called<br />

the GHRSST-PP Match-up <strong>Data</strong>base (MDB).<br />

34 GDS-072 3. Error estimates for each SST measurement are then derived by analysing the MDB for a given time window (e.g., 1<br />

Out of scope<br />

week, 1 month, etc.). For each sensor and for each confidence value (1-5) a Sensor Specific Error Statistic (SSES) is<br />

produced (i.e., for each sensor and confidence value all data are analysed to compute a mean bias and standard<br />

deviation).<br />

34 GDS-073 4. The SSES bias and standard deviation computed in (3) above is then assigned to all [sensor specific] L2P data having<br />

L2P.0050<br />

an equal confidence value.<br />

34 GDS-074 The format of data records held within the MDB is described in Appendix A4 and has been designed to be compatible with<br />

format<br />

other SST MDB (e.g., the Miami pathfinder MDB) in order to take advantage of these data and to further contribute to their<br />

development.<br />

34 GDS-075 GHRSST-PP L2P data products will serve the requirements of some operational ocean modelling groups in real time. SYS.0100<br />

34 GDS-076 However, many user applications and operational modelling groups request complete SST fields, free of gaps caused by<br />

NASR<br />

clouds, rain or lack of data coverage, at regular intervals of between 6 and 24 hours.<br />

34 GDS-077 Most users require an estimate of the sub-surface SST field that is free of diurnal variability (the SSTfnd) that cannot be<br />

NASR<br />

directly measured using current satellite techniques.<br />

© 2004 SOC Page 123 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

34 GDS-078 Others require an estimate of the surface skin temperature that is more directly related to the overlying atmosphere and<br />

SR entry<br />

NASR<br />

subject to considerable diurnal variability<br />

34 GDS-079 To address these needs, L2P data will used in an advanced analysis scheme (e.g., Reynolds and Smith, 1994; Reynolds<br />

SYS.2030<br />

et al, 2002; Guan and Kawamura, 2003, Murray et al. 2002, Murray et al. 1994, Fieguth et al., 1998; 2000, Menemenelis et<br />

al. 1997) that is designed to:<br />

34 GDS-080 (a) Provide a daily global coverage combined SSTfnd product that builds on the synergy between complementary SST<br />

Out of scope<br />

data streams having a grid cell size of 1/12° latitude x longitude (GDS-v1.0),<br />

34 GDS-081 (b) Provide error estimates for each analysed grid cell, L4A.0051<br />

34 GDS-082 (c) Account for differences in spatial and temporal sampling characteristics of each data stream, L4A.0052<br />

34 GDS-083 (d) Account for gaps in coverage due to the presence of cloud, rain or lack of data, L4A.0053<br />

34 GDS-084 (e) Account for SST diurnal variability and retrieve an estimate of the subsurface temperature field (SSTfnd), L4A.0054<br />

34 GDS-085 (f) Provide a measure of diurnal variability within the data product time domain to accompany the SSTfnd estimate and, L4A.0055<br />

34 GDS-086 (g) Provide a best estimate of the skin temperature of the ocean (SSTskin) for each grid cell. L4A.0055<br />

35 GDS-087 At 24 hour intervals, called Analysed Product Processing Windows (APPW) defined in Table 2.3.1, global coverage L2P<br />

Out of scope<br />

data products are collated together at GDAC centres and used in an analysis procedure that will generate GHRSST-PP L4<br />

analysed data products (WP-ID4 in Figure 3.1.1).<br />

35 GDS-088 L4 SST data products include the foundation SST (SSTfnd) including an estimate of SSTskin diurnal variability (magnitude<br />

SYS.2030<br />

and phase). Each L4 data product is formatted to a common GDS L4 data format (Appendix A1) that contains the analysis<br />

at each grid cell together with quality control and error statistic information.<br />

35 GDS-089 L4 data products will be available in real time to the GHRSST-PP user community no later than 12 hours after an APPW<br />

SYS.2050<br />

(see Table 2.3.1). MDB data records should be prepared for each L4 data product and sent to the GHRSST-PP MDB<br />

system for product validation activities (WP-ID5).<br />

© 2004 SOC Page 124 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

35 GDS-090 Global analysis will take place at the GHRSST GDAC although regional analysis systems will be used at RDAC. NSAR<br />

35 GDS-091 The GDS also prepares and delivers L2P and GHRSST-PP Analysed L4 data products for use within the GHRSST-PP<br />

SYS.2080<br />

high resolution diagnostic data set (HR-DDS), fully described in reference document GHRSST/14 “The HR-DDS<br />

implementation plan” and available at http://www.ghrsst-pp.org.<br />

35 GDS-092 Note that the GHRSST-PP HR-DDS is used for diagnostic analysis and validation of the GHRSST-PP GDS by the<br />

NASR<br />

GHRSST-PP community and is completely separate from the GHRSST-PP Match-up <strong>Data</strong>base (MDB) system.<br />

35 GDS-093 The HR-DDS consists of satellite SST and auxiliary data (wind speed and solar radiation where possible) that are<br />

DDS.0030<br />

extracted from data streams and re-gridded to a resolution of 0.01° x 0.01° latitude x longitude over small well defined<br />

geographical areas called HR-DDS granules (see Appendix A5) that are distributed globally.<br />

35 GDS-094 The HR-DDS provides a satellite only (at least in the GDS) data resource that allows easy inter-comparison of separate<br />

DDS.0091<br />

data streams in an operational manner.<br />

35 GDS-095 It is implemented as a shared data resource where data access to will be through a Live Access Server (or variant)<br />

system.<br />

SYS.2081<br />

SYS.2082<br />

35 GDS-096 The primary outputs of the HR-DDS are: NASR<br />

35 GDS-097 Operational assessment of relative satellite SST bias changes (typically caused by an environmental event (e.g., volcanic<br />

NASR<br />

eruption, atmospheric dust) or sensor problems)<br />

35 GDS-098 To test, validate and refine the methods and data products that are produced by the GDS itself. NASR<br />

35 GDS-099 Provide a focus for commissioning the GDS and the production of on-going metrics required to assess the performance of<br />

NASR<br />

the processor.<br />

35 GDS-100 Each GDS work package has been designed to encapsulate a distinct suite of activities, common to all RDAC or GDAC<br />

NASR<br />

that has a definite input and output definition (IODD). In this way, the exchange and use of all data products within the<br />

GDS is greatly simplified by referring to the work package interface IODD provided in section 3.2.<br />

© 2004 SOC Page 125 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

36 GDS-101 3.2 Input and output data definitions (IODD) for the GDS Appendix A3 provides a complete description of the input data<br />

SR entry<br />

format<br />

streams for the GDS and the main outputs of the GDS are listed in Table 3.2.1<br />

39 Work Package number: WP-ID1 Ingestion of <strong>Data</strong> Streams<br />

39 Leader: GHRSST-PO and <strong>Data</strong> Management sub-group (Contact: Ed Armstrong, ed@seastar.jpl.nasa.gov)<br />

39 Aim : To configure the GDS to successfully ingest near real time data streams from different data providers.<br />

39 Objectives:<br />

39 WP1-001 1. Specify the data streams that will be considered by the GDS. NASR<br />

39 WP1-002 2. Specify a data provider responsible for each data stream. NASR<br />

39 WP1-003 3. Specify the primary and any secondary GDAD/RDAC entry points for each data stream. NASR<br />

39 WP1-004 4. Specify the limitations to the use of data arising from specific agreements pertaining to data use within the GHRSST-PP. NASR<br />

39 WP1-005 5. To specify the GHRSST-PP Master Metadata Repository <strong>Data</strong> Set Description (MMR_DSD) records for each data<br />

NASR<br />

stream.<br />

39 WP1-006 6. Specify the data transfer mechanism from data provider to GHRSST-PP primary entry point. NASR<br />

39 Description:<br />

39 WP1-007 This WP is dedicated to the ingestion of satellite and auxiliary data streams (including in situ and NWP fields) within the<br />

NASR<br />

GHRSST-PP GDS. It describes how each data stream is documented within the GHRSST-PP data management system<br />

and provides guidance for RDAC/GDAC on the essential data ingestion/exchange/provision control mechanisms within the<br />

GDS.<br />

39 WP1-008 For each data stream, a MMR_DSD record will be generated and registered by the GHRSST-PO and data management<br />

ING.0030<br />

sub-group to which all other L2P MMR File records (MMR_FR) that describe individual data files are registered in WP-ID2.<br />

39 WP1-009 If data are rejected by the GDS then an appropriate error message should be sent to the data provider stating why the data ING.0020<br />

© 2004 SOC Page 126 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

have been rejected. The error message should be copied to the ERRLOG (errlog@ghrsst-pp.org).<br />

SR entry<br />

ING.0100<br />

QLC.0010<br />

39 WP1-001 Inputs:<br />

39 WP1-010 1. <strong>Data</strong> product definitions. NASR<br />

39 WP1-011 2. <strong>Data</strong> provider definitions and restrictions. NASR<br />

39 WP1-012 3. RDAC/GDAC data usage descriptions. NASR<br />

39 WP1-013 4. <strong>Data</strong> access and use agreements. NASR<br />

39 WP1-014 5. MMR_DSD records. NASR<br />

39 Outputs:<br />

39 WP1-015 1. E-mail error message sent data provider explaining data ingestion/format problems. ING.0100<br />

39 WP1-016 2. E-mail to the GHRSST-PP ERRLOG explaining data ingestion/format problems. QLC.0010<br />

39 WP1-017 3. MMR_DSD records for each data stream used by the GDS. ING.0030<br />

39 Acceptance tests:<br />

39 WP1-018 1. Error message indicating data set rejection is correctly generated. ATPD<br />

40 WP1-019 2. Error- message is correctly delivered to data provider. ATPD<br />

40 WP1-020 3. Error messages are correctly delivered to the GHRSST-PP ERRLOG. ATPD<br />

40 WP1-021 4. MMR_DSD records prepared and registered correctly within the GHRSST-PP MMR system. ATPD<br />

40 WP1-022 5. <strong>Data</strong> files are correctly renamed on acceptance into the GDS. ATPD<br />

40 WP1-023 Figure 4.1 provides a functional breakdown diagram that identifies the main data processing tasks within WP-ID1. These<br />

NASR<br />

are:<br />

40 WP1-024 1. Establish real time operational access to data streams at RDAC and GDAC centres as appropriate for the processing<br />

centre (i.e. regional or global).<br />

ING.0010<br />

EXT.ING.0010<br />

© 2004 SOC Page 127 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

EXT.ING.0020<br />

40 WP1-025 2. Create a Master Metadata Repository <strong>Data</strong> Set Description (MMR_DSD) record within the GHRSST-PP MMR for each<br />

data stream described in Appendix A3.<br />

ING.0030<br />

MDB.0530<br />

40 WP1-026 3. Evaluate each data files for integrity when ingested at an RDAC or GDAC facility. ING.0060<br />

CTR.0120<br />

CTR.0110<br />

40 WP1-027 4. Determine if the data file should be accepted into the GDS. If not, then raise an error report and send this to the data<br />

provider and to the GDS ERRLOG.<br />

QLC.0010<br />

EXT.ING.0030<br />

40 WP1-028 5. Rename the data file by prefixing the RDAC identification code to the filename. ING.0040<br />

40 WP1-029 Each sub task is described in the following sections. NASR<br />

40 F4.1 WP1-030 Figure 4.1 Functional breakdown of GDS WP-ID1 identifying each sub task. NASR<br />

41 WP-ID1.1.1 Access to input data streams<br />

41 WP1-031 The GDS does not consider L0, L1a or L1B data streams (although L1B data streams may be available for certain HR-<br />

NASR<br />

DDS sites if RDAC are able to supply these data).<br />

41 WP1-032 Satellite SST, auxiliary satellite data (including wind speed and surface solar irradiance (SSI)), in situ observations and<br />

L2P.0010<br />

NWP data are all required by the GDS to construct L2P data products and to produce L4 analysed SST data product<br />

outputs.<br />

41 WP1-033 Satellite data are ingested at RDAC and in some cases at GDAC centres. L2P.0010<br />

41 WP1-034 Each data stream will be formatted according to a well specified format provided by each data provider. format<br />

41 WP1-035 Note that data streams from the same sensor may have different formats depending on the data provider. format<br />

41 WP1-036 Table 4.1 provides an overview of the data sets that are available to the GHRSST-PP during the period 2003-2007<br />

categorised by data type.<br />

SRD 5.2 Tables<br />

5-1, 5-2 & 5-3<br />

© 2004 SOC Page 128 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

41 T4.1 WP1-037 Table 4.1 Real Time global and regional satellite and in situ data streams available to the GHRSST-PP during the period<br />

SR entry<br />

NASR<br />

2003-2007. In situ data streams are those providing data within the upper 8 m of the water column. Acronym definitions<br />

are given in Appendix A2 Table A2.2.<br />

42 WP1-038 Appendix A3 describes in detail the data that will be used by the GDS. Appendix A3.1 defines the satellite SST data<br />

NASR<br />

streams, Appendix A3.2 the auxiliary satellite data streams, Appendix A3.3 the in situ data streams and Appendix A3.4 the<br />

NWP/Forecast fields.<br />

42 WP1-039 A unique identification code for each data stream is given in Appendix A3 that will be referred to throughout the GDS. NASR<br />

42 WP1-040 Also specified are the general characteristics of each data stream including the data provider, data delivery mechanism,<br />

NASR<br />

any data agreement that is applicable together with the primary and secondary RDAC/GDAC entry points.<br />

42 WP1-041 All of this information will be used by the GHRSST-PO and the GHRSST-PP <strong>Data</strong> management sub-group to generate<br />

NASR<br />

MMR_DSD records for each data stream that is used by the GDS according to the specifications provided in Appendix A6<br />

Table A6.2.1.<br />

42 WP-ID1.1.2 Preparation of Master Metadata Repository (MMR) data set description records (MMR_DSD)<br />

42 WP1-042 The data streams within the GDS are dynamic in the sense that new data sets will become available, certain data sets may<br />

NASR<br />

take priority in given circumstances (e.g., aerosol contamination) and regional preferences based on the ease of data<br />

access will exist at RDAC.<br />

42 WP1-043 The GHRSST-PP Master Metadata Repository (MMR) is an online dynamic database that contains a metadata record for<br />

NASR<br />

every data file that is produced by the GHRSST-PP. It provides a major source of information for RDAC and GDAC<br />

describing the availability of data files at any given instant in time. It also provides a catalogue for the GHRSST-PP<br />

archive.<br />

42 WP1-044 The GHRSST-PP MMR system and the format of MMR_DSR and associated MMR File Records (MMR_FR),<br />

ING.0050<br />

implementation of the MMR and delivery of MMR data records by RDAC and GDAC is described in Appendix A6.<br />

© 2004 SOC Page 129 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

42 WP1-045 The GHRSST-PO in collaboration with individual data providers, RDAC and GDAC will prepare and maintain MMR_DSD<br />

SR entry<br />

CTR.0530<br />

records within the MMR system for each data stream used in the GDS.<br />

42 WP1-046 The information within the MMR_DSD must be kept current as these data will be used to monitor the availability of data,<br />

CTR.0530<br />

problems and issues that are relevant to operations and data quality.<br />

42 WP-ID1.2 Evaluate the integrity of data files<br />

42 WP1-047 When a data file enters a GDAC or RDAC processing facility following transfer from the data provider, it typically resides<br />

NASR<br />

on a temporary staging area at an RDAC or GDAC centre. In some cases, the delivery may be incomplete; data<br />

connections could be lost, corruption of the data may have occurred during delivery, or the original file was incorrectly<br />

generated.<br />

42 WP1-048 The aim of the GDS at this stage is to identify a data delivery/data format problem and to initiate a dialog with the data<br />

provider to establish a solution as soon as possible.<br />

42 WP1-049 Individual RDAC and GDAC are responsible for implementing local procedures that are able to access data and a wide<br />

ING.0060<br />

CTR.0120<br />

CTR.0180<br />

variety of tools will be used that are specific to each RDAC.<br />

42 WP1-050 RDAC/GDAC must also implement local procedures to assess the integrity of satellite data files, delivery protocols and<br />

reliability noting that these are likely to vary according to the data provider (i.e. the same data may be provided in a<br />

different format) and in time (data providers change file format, sensor is down, communication. problems).<br />

42 WP1-051 RDAC systems should focus on operational reliability and simplicity rather than elaborate procedures that are likely to<br />

CTR.0140<br />

CTR.0150<br />

CTR.0170<br />

DES.SYS.0030<br />

introduce unnecessary complexity into the data processing system with minimal impact.<br />

43 WP1-052 As a guide, the first GDS processing steps must evaluate the timeliness and integrity of each data file based on the<br />

see below<br />

following questions:<br />

43 WP1-053 1. What is the status of the satellite system at present Are there health warnings or environmental problems affecting the<br />

CFI.L2P.0010<br />

quality of data This information should be present in the MMR_DSD record.<br />

© 2004 SOC Page 130 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

43 WP1-054 2. Was data transfer successful and is the data file complete File sizes and CRC checks for example, will establish the<br />

SR entry<br />

CTR.0120<br />

situation.<br />

43 WP1-055 3. Is the data file correctly formatted according to the specifications provided by the data provider The data format could<br />

L2P.0020<br />

be randomly checked at the RDAC.<br />

43 WP1-056 4. Was the delivery timely ING.0110<br />

43 WP-ID1.3 Admit data file into the GDS system<br />

43 WP1-057 Based on the assessment made by RDAC in WP-ID1.2 a decision is made to either accept a data stream or to reject the<br />

L2P.0020<br />

data stream from the GDS. The GDS specifies the following rules for this purpose:<br />

43 Rule<br />

WP1-058<br />

A data file should be admitted into the GDS processing system by an RDAC or GDAC only if the data file has been verified<br />

QLC.0045<br />

1.3.1<br />

as complete, correctly formatted and that the delivery was timely with respect to the L2P delivery criteria of ‘ideally<br />

available within 6 hours of acquisition at the satellite platform’ and the analysed product processing window (APPW)<br />

defined in Table 2.3.1.<br />

43 WP-ID1.3.1 Inform data provider and ERRLOG of data delivery/ingestion problem.<br />

43 WP1-059 If a data file has been rejected by an RDAC/GDAC, an error must be raised and an e-mail message sent to the data<br />

ING.0100<br />

provider that explains why the data file has been rejected from the GHRSST-PP GDS. These messages are used to:<br />

43 WP1-060 (a) Establish a healthy feedback dialog with data providers NASR<br />

43 WP1-061 (b) To provide information to the GHRSST-PP that can be used to monitor GDS operations. NASR<br />

43 WP1-062 A separate error message should be sent for each data access attempt (i.e., if a second download is attempted and<br />

ING.0120<br />

also fails two ERRLOG messages are generated). The <strong>Data</strong> provider contact details and point of contact for error<br />

reports can be found in the L2P MMR_DSD records associated with each satellite data stream.<br />

43 WP1-063 A copy of the error message should be sent via e-mail to the GHRSST-PP ERRLOG system using the address<br />

QLC.0010<br />

errlog@ghrsst-pp.org. The subject of the message should be written as follows:<br />

© 2004 SOC Page 131 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

44 WP1-065 GDS ERROR message:where is defined in Appendix A2.1 and<br />

is defined in Appendix A7. The format of ERRLOG messages is provided in Appendix A7.<br />

SR entry<br />

EXT.CTR.0520<br />

<br />

44 WP-ID1.4 Rename data file to GHRSST-PP GDS specification.<br />

44 WP1-066 In order to identify a data file that has been admitted into the GDS system by an RDAC or GDAC, the GDS recommends<br />

ING.0040<br />

that each data file should be renamed by prefixing the processing centre code to the original filename according to the<br />

following filename specification:<br />

44 WP1-067 -GDS- ING.0040<br />

44 WP1-068 where is defined in Appendix A2 Table A2.1. In the example above, JAP-GDS-20010507tm<br />

ING.0040<br />

would refer to a REMSS TMI bmaps_03 data file (20010507tm) that has been checked and admitted into the GDS at the<br />

Japanese RDAC (JAP).<br />

44 WP1-069 Renaming a data file in this way provides a simple method to identify data that have passed the initial data delivery tests<br />

NASR<br />

from those that have not while preserving the original data product filename. Although data files within this part of the GDS<br />

may have no visibility beyond the RDAC processor itself, the purpose of renaming files to a common specification is to<br />

facilitate error analysis and problem solving. If problems are found at a higher level within the GDS processing chain,<br />

accepted GDS files may be differentiated from other data files and traced back to RDAC/GDAC centres by simply noting<br />

the filename.<br />

GDS-WP-ID2<br />

WP-ID2: Generation of GHRSST-PP Level-2 Pre-processed <strong>Data</strong> files (L2P).<br />

45 To process input data files and produce GHRSST-PP L2P SST data files in a timely manner.<br />

45 Objectives:<br />

45 WP2-001 1. To extract confidence data provided with each input SST data file. QLC.0050<br />

© 2004 SOC Page 132 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

QLC.0060<br />

QLC.0065<br />

45 WP2-002 2. To quality control input SST data and derive GHRSST-PP pixel confidence data. QLC.0010<br />

45 WP2-003 3. To determine if SST data are suitable for further use in the GDS. If not inform the data provider and generate an error<br />

QLC.0010<br />

message for the GHRSST-PP ERRLOG explaining why these data have not been used in the GDS.<br />

45 WP2-004 4. Extract auxiliary data (wind speed, surface solar irradiance and aerosol optical depth) and collocate with SST data. MRG.0010<br />

45 WP2-005 5. Create a GHRSST-PP netCDF L2Pdata file. Optional fields will only be included prior to a cut-off time at which point the<br />

L2P.0055<br />

L2P data will be shipped ‘as is’.<br />

45 WP2-006 6. For each L2P data file, prepare and submit L2P MMR_FR to the GHRSST-PP MMR. CTR.0530<br />

45 WP2-007 7. Submit appropriate error messages to the GHRSST-PP ERRLOG log (errlog@grsst-pp.org). CTR.0550<br />

CTR.0580<br />

45 Description:<br />

45 WP2-008 This part of the GDS is dedicated to the production of L2P data files that consist of native SST observations together with<br />

NASR<br />

additional pixel confidence data and auxiliary data stored in GDS netCDF format.<br />

45 WP2-009 L2P data products form the basic “building blocks” for all other GHRSST-PP data products and should ideally be made<br />

NASR<br />

available to the GHRSST-PP operational user community in real time within 6 hours after the reception of data at the<br />

satellite sensor.<br />

45 WP2-010 Each input SST data file is assessed pixel by pixel, using gross error checking rules. L2P.0020<br />

45 WP2-011 A configuration file is used to store all QC parameter limits and thresholds that will be revised as required. QLC.0020<br />

45 WP2-0120 Confidence data are extracted from native SST data streams if present and they are derived using auxiliary and reference<br />

data streams and sensor specific error statistics (SSES, generated by WP-ID7).<br />

L2P.0050<br />

CFD.0030<br />

45 WP2-013 For every L2P file that is generated, a L2P MMR_FR must be created and registered under the appropriate MMR_DSD at ING.0030<br />

© 2004 SOC Page 133 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

the GHRSST-PP MMR system as soon as possible.<br />

SR entry<br />

CTR.0550<br />

CTR.0580<br />

45 WP2-014 As the GHRSST-PP intends to generate global 1/12° L4 analysed SST data output from L2P data, high resolution ) input<br />

L2P.0090<br />

data sets may be averaged to a coarser resolution no greater than 1/12° in order to reduce data volume. It should be<br />

stressed that this is highly undesirable as it will impact the usefulness of L2P data sets within the coastal zone<br />

where 1km data may already be too coarse in certain areas characterised by dynamic temperature structures.<br />

45 WP2-015 GHRSST-PP High Resolution Diagnostic <strong>Data</strong> Set (HR-DDS) granules are extracted from L2P data, formatted and<br />

DDS.0030<br />

delivered to the GHRSST-PP HR-DDS system.<br />

45 WP2-016 If satellite SST data are unsuitable for further use in the GDS then a message is sent to the data provider and to the<br />

GHRSST-PP ERRLOG log (errlog@ghrsst-pp.org) informing why these data have not been used.<br />

ING.0100<br />

QLC.0035<br />

46 Inputs:<br />

46 WP2-017 1. Infrared and Microwave satellite SST data streams. L2P.0010<br />

46 WP2-018 2. Auxiliary satellite and NWP data streams. L2P.0010<br />

46 WP2-019 3. Reference data streams.<br />

46 WP2-020 4. GDS L2P configuration files. QLC.0020<br />

46 WP2-021 5. Error checking rules. QLC.0020<br />

46 Outputs:<br />

46 WP2-022 1. GHRSST-PP L2P data files. L2P.0070<br />

46 WP2-023 2. L2P MMR_FR metadata record. IMG.0030<br />

46 WP2-024 3. Error message sent to data provider and to the GHRSST-PP ERRLOG log (errlog@ghrsst-pp.org) as required. IMG.0100<br />

46 WP2-025 4. HR-DDS data granules extracted from L2P data. DDS.030<br />

46 Acceptance tests:<br />

© 2004 SOC Page 134 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

46 WP2-026 1. L2P output data products based on the GDS test input data set produced at each RDAC and GDAC should be<br />

SR entry<br />

ATPD<br />

identical to those produced by the GDS reference data processor.<br />

46 WP2-027 2. L2P measurement data are identical to the input SST data but with additional error and QC). ATPD<br />

46 WP2-028 3. HR-DDS data extraction is correct and data delivery interface is functional. ATPD<br />

46 WP2-029 4. Gross error checking procedures successfully flag unsuitable data and admit useful data with minimal data loss. ATPD<br />

46 WP2-030 5. Error messages to data provider and ERRLOG are correct and promptly delivered. ATPD<br />

46 WP2-031 6. MMR_FR metadata records are submitted in a timely manner. ATPD<br />

46 WP2-032 7. MMR_FR metadata records are accurate. ATPD<br />

46 Metric for performance assessment:<br />

46 WP2-033 1. L2P data files are correctly formatted. ATPD<br />

46 WP2-034 2. L2P data are produced in a timely manner (ideally within 6 hours from reception at satellite platform). ATPD<br />

46 WP2-035 3. L2P MMR_FR metadata records provided and ingested at MMR in real time. MMR rejection rate < 2%. ATPD<br />

CTR.0130<br />

46 WP2-036 4. <strong>Data</strong> provider promptly notified in case of problems with data stream. ATPD<br />

46 WP2-037 5. HR-DDS L2P extraction and format is correct. ATPD<br />

46 WP2-038 6. ERRLOG messages are correct and only reject problematic data. ATPD<br />

46 WP2-039 The purpose of this work package is to derive Level-2 Pre processed GHRSST-PP data files (L2P) through a process of<br />

NASR<br />

quality control and data merging. The principles for setting the quality flags and pre-processing algorithms for assigning<br />

quality flags are outlined in this section.<br />

46 WP2-040 The L2P processor identifies and flags problematic satellite SST data at an early stage of the GDS processing chain<br />

NASR<br />

leaving as much time as possible to correct faults and make use of the erroneous data later on in the processing system.<br />

46 WP2-041 The main difference between the input SST data file and the output L2P data file is that additional confidence data and L2P.0040<br />

© 2004 SOC Page 135 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

sensor specific error estimates for each pixel value have been included and the original SST data reformatted to netCDF<br />

format.<br />

SR entry<br />

L2P.0050<br />

L2P.0060<br />

46 WP2-042 No modification to the original SST data values has taken place during the conversion process. L2P.0080<br />

47 Footnote<br />

7<br />

WP2-043 Note that in the case of 1km input SST data streams, averaging of these data is permitted to provide data at a resolution<br />

of no greater than 1/12° latitude x longitude. In this case, there may be some modification to the original input data.<br />

L2P.0090<br />

47 WP2-044 L2P data products will be available to the operational user community in real time and ideally, within 6 hours following<br />

PER.SYS.0010<br />

reception at the satellite platform.<br />

47 WP2-045 However, as the GHRSST-PP intends to generate global 1/12° L4 analysed SST data output from L2P data files, high<br />

L2P.0090<br />

resolution input data sets may be averaged to a coarser resolution no greater than 1/12°.<br />

47 Figure<br />

WP2-046 Functional breakdown of WP-ID2 identifying each major sub task NASR<br />

5.1<br />

47 WP2-047 A functional breakdown diagram for WP-ID2 is shown in Figure 5.1 (additional functional breakdown diagrams are<br />

NASR<br />

provided for WP-ID2.1, WP-ID2.2 and WP-ID2.5 in the appropriate section below).<br />

47 WP2-048 In a series of pre-processing and quality control procedures, microwave and Infrared Satellite SST data are merged<br />

L2P.0030<br />

together with reference data and auxiliary data including satellite and NWP Wind speed, sea ice concentration, surface<br />

solar irradiance and aerosol optical depth estimates.<br />

47 WP2-049 If data processing problems are encountered the SST data file is rejected from the GDS, (e.g., values are outside of<br />

expected range, overwhelming cloud contamination) an error message is sent to the satellite SST data provider point of<br />

ING.0100<br />

QLC.0035<br />

contact (as defined in the GHRSST-PP MMR_DSD record associated with the SST data file) explaining why the SST data<br />

have not been used in the GDS. A copy of the message is also sent to the GHRSST-PP error log system (errlog@ghrsstpp.org).<br />

48 WP2-050 If the SST data are accepted into the GDS, L2P additional pixel confidence data are then derived for every SST L2P.0040<br />

© 2004 SOC Page 136 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

measurement. L2P confidence data are required to:<br />

48 WP2-051 1. Derive a quantitative ‘confidence value’ for each SST measurement that is used to identify reduced quality data. NASR<br />

48 WP2-052 2. Derive and interpret of SSES values assigned to each SST measurement. NASR<br />

48 WP2-053 3. Force parameterisations that estimate diurnal variability. NASR<br />

48 WP2-054 Using confidence data, Sensor Specific Error Statistics (SSES) are assigned to each SST measurement based on analysis<br />

L2P.0050<br />

of the GHRSST-PP Match-up <strong>Data</strong>base (MDB) system described in WP-ID3.<br />

48 WP2-055 A L2P data file is generated (format described in Appendix A1.4) which contains the original SST data together with pixel<br />

L2P.0060<br />

confidence data. L2P data files are formatted as GDS netCDF files (WP-ID1.2.3) that are fully described in Appendix A1.<br />

48 WP2-056 For each L2P data file a MMR_FR metadata record (described in Appendix A6) is generated and registered at the MMR<br />

ING.0030<br />

under the appropriate MMR_DSD.<br />

48 WP2-057 L2P data files and pixel confidence data are identical in format for all input SST data streams with differs between the IR<br />

L2P.0060<br />

and MW SST data exist. Using a single data format for all SST data files greatly facilitates SST inter-comparison and<br />

merging activities within the GHRSST-PP.<br />

48 WP-ID2.1 Pre-processing and merging of auxiliary data.<br />

48 WP2-058 This work-package describes in detail common pre-processing operations performed on each IR and MW SST data file.<br />

MRG.0020<br />

Figure 5.2.1 provides a functional breakdown diagram that identifies each common pre-processing pixel based operation.<br />

These are:<br />

48 WP2-059 Check that the SST value lies within acceptable limits, if not then reject the data file QLC.0070<br />

48 WP2-060 Compute the pixel measurement time ING.0070<br />

48 WP2-061 Assign a fractional sea ice value if required MRG.0030<br />

48 WP2-062 Assign a wind speed value MRG.0040<br />

48 WP2-063 Assign a surface solar irradiance (SSI) value MRG.0050<br />

© 2004 SOC Page 137 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

48 WP2-064 Assign an Atmospheric Optical Depth (AOD) value MRG.0060<br />

48 WP2-065 The specific thresholds, limits and processing rules that are used to QC each data set are stored in a system wide L2P<br />

QLC.0020<br />

configuration file. This allows easy modification of critical values and other configurable parameters without the need for<br />

software code changes.<br />

48 WP2-066 As auxiliary data files are only used as indicators within the GDS, only simple QC tests are specified. NASR<br />

48 WP2-067 It is left to the RDAC and GDAC to determine the level of QC that is required to use these data with confidence. NASR<br />

48 WP2-068 Different pre-processing operations must be applied to IR and MW SST data and to each auxiliary data file. NASR<br />

49 Figure<br />

WP2-069<br />

Fig.5.2.1 Functional breakdown diagram of pre-processing and merging operations common to IR and MW SST data files.<br />

NASR<br />

5.2.1<br />

A L2P configuration file is used to store appropriate thresholds and reference values.<br />

49 WP-ID2.1.1 SST Climatology limit check<br />

49 WP2-070 This test is used to set the L2P confidence data record variable DT_min defined in Table A1.2.2 in Appendix A1. It<br />

identifies data that fall outside of acceptable range of global SST values and identifies unrealistic SST values.<br />

QLC.0070<br />

QLC.0080<br />

49 The GDS specifies the following rules:<br />

49 Rule<br />

WP2-071<br />

Rule 2.1.1a: A valid SST observation must lie within a temperature range of LowSSTLimit < SST < HighSSTLimit. If the<br />

QLC.0070<br />

2.1.1a<br />

pixel SST value is out of these limits, the observation must be rejected from the final L2P data file.<br />

QLC.0040<br />

49 Rule<br />

WP2-072<br />

Rule 2.1.1b: The difference between the pixel SST value (SSTobs) and the reference SST fied, SSTref (defined in<br />

QLC.0080<br />

2.1.1b<br />

Appendix A3, Table A3.1.1), DT_min is derived using: DT_min = SSTobs - SSTref (Eq. 2.1.1) The<br />

DT_min value should be inserted into the DT_min field of the L2P confidence data record for the pixel in question.<br />

49 WP2-074 This reference SST is SST1m derived as a 10 day “coldest” climatology from Pathfinder SST data products. SST values<br />

QLC.0080<br />

for a given location and season that are significantly beneath this climatological value indicate the presence of cloud<br />

contamination.<br />

50 WP-ID2.1.2 Compute SST measurement time<br />

© 2004 SOC Page 138 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

50 WP2-075 Most SST data files provide the time at which a SST measurement was acquired but some do not. The GHRSST-PP GDS<br />

SR entry<br />

ING.0070<br />

specifies that the time of measurement should be extracted or computed to the nearest minute and then coded as<br />

continuous UTC time coordinates using the following rules:<br />

50 Rule<br />

WP2-076<br />

Rule 2.1. 2a The time of the earliest SST measurement within this data set should be coded as a continuous time<br />

ING.0080<br />

2.1.2a<br />

coordinate specified as minutes from 00:00:00 UTC January 1, 1981 (start of the useful AVHRR SST data record) and<br />

entered into the netCDF data file global variable Start_Time.<br />

50 Rule<br />

WP2-077<br />

Rule 2.1.1.2b The time of SST measurement should be coded as the deviation in minutes from the netCDF data file global<br />

ING.0090<br />

2.1.1.2b<br />

variable Start_Time and entered into the L2P confidence record variable TimeDeviation.<br />

50 WP-ID2.1.3 Assign fractional sea ice value if required<br />

50 WP2-078 Some SST data are contaminated in part or wholly by sea ice. This procedure is used to set the L2P confidence data<br />

record variable FractionalSeaIce defined in Table A1.2.2 in Appendix A1.<br />

MRG.0070<br />

MRG.0080<br />

MRG.0090<br />

50 WP2-079 Some input SST data streams provide a flag to indicate that the SST measurement is contaminated by sea ice (e.g.,<br />

MRG.0090<br />

AMSR-E). In cases where no such data are present in the input data stream a reference sea ice data product (defined in<br />

Appendix A3) should be used to assess the likelihood of sea ice contamination. The GDS specifies the following rules:<br />

50 Rule<br />

WP2-080<br />

Rule 2.1.3a If an input data set pixel fractional sea ice estimate exists, this should be used to set the FractionalSeaIce<br />

MRG.0070<br />

2.1.3a<br />

flag of the L2P confidence data record.<br />

50 Rule<br />

WP2-081<br />

Rule 2.1.3b If an input data set pixel sea ice flag exists (i.e. indicating sea ice but not ther fractional amount of coverage),<br />

MRG.0080<br />

2.1.3b<br />

this should be used to set the FractionalSeaIce flag of the L2P confidence data record to 100% according to the code<br />

values defined in Table A1.4.2 in Appendix A1. If the flag is a binary flag. a check should be made against the reference<br />

fractional sea Ice data set to determine if the pixel is likely to be 100% contaminated and the fraction reported by the<br />

reference data set should be set if it is < 100%.<br />

© 2004 SOC Page 139 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

50 Rule<br />

WP2-082<br />

Rule 2.1.3c If an input data set pixel sea ice flag does not exist, and the pixel is located in or close to a region that may be<br />

MRG.0090<br />

2.1.3c<br />

ice contaminated, the reference sea ice data sets defined in Appendix A3 should be used to determine the value of the<br />

L2P confidence flag FractionalSeaIce.<br />

50 Rule<br />

WP2-083<br />

Rule 2.1.3d The L2P confidence bitfield FractionalSeaIce_source must be set to indicate the data source used to set<br />

MRG.0110<br />

2.1.3d<br />

FractionalSeaIce using the values specified in Table 2.1.3.<br />

50 Rule<br />

WP2-084<br />

Rule 2.1.3e: If the SST input data set is derived from a microwave sensor and a sea ice flag is present in the data stream,<br />

MRG.0120<br />

2.1.3e<br />

bit 4 of the L2P confidence data variable MW_SST_flags should be set for this pixel.<br />

51 Table<br />

WP2-085 Code value for the source of the data used to assign the L2P confidence value FractionalSeaIce_source MRG.0110<br />

2.1.3<br />

51 WP-ID2.1.4 Assign wind speed data to L2P confidence record<br />

51 WP2-086 Wind speed measurements are required within the GDS to derive L2P SST proximity confidence values, for use in diurnal<br />

NASR<br />

variability correction schemes and to interpret the relationship between satellite and in situ SST data.<br />

51 WP2-087 At low wind speeds, especially in clear sky conditions, stronger diurnal variability is expected leading to higher surface<br />

NASR<br />

layer temperature gradients and the potential for significant de-coupling of the skin/sub-skin SST from the SST at depth.<br />

51 WP2-088 Ideally a near contemporaneous wind speed measurement from satellite sensors should be used but this is impossible for<br />

NASR<br />

all sensors due to the limited number of satellite wind speed sensors available.<br />

51 WP2-089 As a surrogate for a measured wind speed value, NWP estimates may be used as an indication of the surface wind<br />

NASr<br />

speed.<br />

51 WP2-090 The L2P confidence data record variable Wspd_value defined in Table A1.2.2 contains a best estimate of the surface<br />

MRG.0130<br />

wind speed at the time of SST data acquisition.<br />

51 WP2-091 A near contemporaneous wind speed estimate (an NWP field may be used here) is assigned to this field as an indicator of<br />

MRG.0130<br />

the turbulent state of the air sea interface.<br />

© 2004 SOC Page 140 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

51 WP2-092 The source of wind speed information must be specified in the L2P confidence data variable Wspd_source (defined in<br />

Table 2.1.4) and the difference in acquisition time between SST observation and wind speed measure must be specified in<br />

SR entry<br />

see next five<br />

rows<br />

the L2P confidence variable Wspd_Dtime_from_SST.<br />

51 The GDS specifies the following rules:<br />

51 Rule<br />

WP2-093<br />

Rule 2.1.4a: A 10m surface wind speed value should be assigned to each SST measurement pixel using the GDS L2P<br />

MRG.0130<br />

2.1.4a<br />

confidence data variable Wspd_value described in Table A1.2.2. As a general rule for the GDS v1.0, only simultaneous<br />

wind speed measurements should be used otherwise NWP analyses should be used. Future revisions of the GDS will<br />

most likely use near contemporaneous wind speed observations from other satellite instruments.<br />

51 Rule<br />

WP2-094<br />

Rule 2.1.4b: The source of data used to set the L2P confidence data variable Wspd should be indicated in the confidence<br />

Format<br />

2.1.4b<br />

variable Wspd_source as defined in Table 2.1.4.<br />

requirement<br />

51 Rule<br />

WP2-095<br />

Rule 2.1.4c: In the case of microwave SST measurements, the use of a simultaneous microwave 10m wind speed<br />

MRG.0150<br />

2.1.4c<br />

measurements obtained from the same instrument providing the SST measurement may be used.<br />

51 Rule<br />

WP2-096<br />

Rule 2.1.4d: In the absence of a simultaneous surface wind speed measurement, an NWP estimated 10m surface wind<br />

MRG.0160<br />

2.1.4d<br />

speed (See Appendix A3, Table A3.1.1) should be used to set the L2P confidence data variable Wspd_value.<br />

52 Rule<br />

WP2-097<br />

Rule 2.1.4e: The difference in time expressed in hours between the time of SST measurement and the time of wind speed<br />

MRG.0170<br />

2.1.4e<br />

measurement should be entered into the L2P confidence data variable Wspd_Dtime_from_SST. A value of 25 should be<br />

used to indicate unknown.<br />

52 Table<br />

2.1.4<br />

WP2-098 Code value for the source of the data used to assign the L2P confidence value Wspd_source Format<br />

requirement<br />

52 WP2-099 The following QC checks should be applied to all input wind speed data sets. The tests applied are simple tests designed<br />

QLC.0090<br />

to verify that wind speed data are within realistic limits.<br />

52 WP2-100 RDAC and GDAC may choose to implement more sophisticated QC procedures if deemed necessary.<br />

© 2004 SOC Page 141 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

52 WP-ID2.1.4.1 Gross wind speed limit check<br />

52 WP2-101 This test identifies wind speed values that fall outside of acceptable range of global wind speed values. The GDS specifies<br />

NASR<br />

the following rule:<br />

52 Rule<br />

WP2-102<br />

Rule 2.1.4.1: A valid wind speed observation must lie within a wind speed range of 0 < u < HighWindSpeed. Wind<br />

QLC.0090<br />

2.1.4.1<br />

speed values that fall outside this range are considered erroneous and should not be used within the GDS.<br />

52 WP2-103 The HighWindSpeed value is set in a L2P configuration file maintained at the RDAC and GHRSST-PP International<br />

QLC.0120<br />

Project Office.<br />

52 WP-ID2.1.5 L2P Surface Solar Irradiance (SSI) value assignment<br />

52 WP2-104 SSI measurements are required within the GDS to asses the magnitude and variability of significant diurnal SST variations,<br />

NASR<br />

for use in diurnal variability correction schemes, as part of the L4 SST analysis procedures and to interpret the relationship<br />

between satellite and in situ SST data.<br />

52 WP2-105 Ideally a near contemporaneous SSI measurement from satellite sensors should be used but this is impossible for all<br />

NASR<br />

areas due to the limited number of geostationary satellite sensors available.<br />

52 WP2-106 As a surrogate for a measured SSI value, NWP estimates may be used. MRG.0190<br />

52 WP2-107 This test is used to set the L2P confidence data record variable SSI_value defined in Table A1.2.2 in Appendix A1. MRG.0180<br />

MRG.0190<br />

52 WP2-108 Near contemporaneous SSI measurement, coded according to the definition provided in Appendix A1 Table A1.2.2 should<br />

MRG.0180<br />

be assigned to the L2P confidence data a record field SSI_value.<br />

52 WP2-109 The source of SSI information must be specified in the L2P confidence data variable SSI_source (defined in Table 2.1.5)<br />

MRG.0200<br />

and the difference in acquisition time between SST observation and SSI measure must be specified in the L2P confidence<br />

variable SSI_Dtime_from_SST.<br />

53 WP2-110 The GDS specifies the following rules: see next 5 rows<br />

© 2004 SOC Page 142 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

53 Rule<br />

WP2-111<br />

Rule 2.1.5a: A 6 hourly integrated downwelling SSI measurement (e.g., derived from satellite measurements) should be<br />

MRG.0180<br />

2.1.5a<br />

assigned to each SST pixel value using the SSI_value L2P confidence data variable. The SSI measurement nearest in<br />

space and time before the input pixel SST value should be used.<br />

53 Rule<br />

WP2-112<br />

Rule 2.1.5b: If no SSI measurement is available, a 6 hourly integrated SSI value derived from an NWP system (See<br />

MRG.0190<br />

2.1.5b<br />

Appendix A3) nearest in space and time to the SST measurement should be used to set the value of SSI_value.<br />

53 Rule<br />

WP2-113<br />

Rule 2.1.5c: The source of data used to set the L2P confidence data variable SSI_value should be indicated in the<br />

MRG.0200<br />

2.1.5c<br />

confidence variable SSI_source as defined in Table 2.1.5.<br />

53 Rule<br />

WP2-114<br />

Rule 2.1.5e: The difference in time expressed in hours between the time of SST measurement and the mean time of SSI<br />

MRG.0210<br />

2.1.5e<br />

measurement/data validity should be entered into the L2P confidence data variable SSI_Dtime_from_SST. A value of 25<br />

should be used to indicate unknown.<br />

53 Table<br />

WP2-115 Code value for the source of the data used to assign the L2P confidence value SSI_source MRG.0200<br />

2.1.5<br />

53 WP2-116 The following QC checks should be applied to all input SSI data sets that are used at RDAC and GDAC. These tests are<br />

see following<br />

simple tests designed to verify that SSI data are within realistic limits. RDAC and GDAC may choose to implement more<br />

sophisticated QC procedures if deemed necessary.<br />

53 WP-ID2.1.5.1 Gross SSI limit check<br />

53 WP2-117 This test identifies data that fall outside of acceptable range of global SSI values. The GDS specifies the following rule: NASR<br />

53 Rule<br />

2.1.5.1<br />

WP2-118 Rule 2.1.5.1: A valid 6 hourly integrated SSI observation/forecast must lie within a SSI range of 0 < SSI <<br />

HighSSIValue. SSI values that fall outside this range should not be used within the GDS.<br />

QLC.0100<br />

53 WP2-119 The HighSSIValue value is set in the L2P configuration file maintained at the RDAC and GHRSST-PP International<br />

QLC.0120<br />

Project Office.<br />

53 P-ID2.1.6 L2P Aerosol Optical Depth (AOD) value assignment<br />

© 2004 SOC Page 143 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

53 WP2-120 This test is used to set the L2P confidence data record variable AOD_value defined in Table A1.2.2 in Appendix A1. A<br />

SR entry<br />

MRG.0220<br />

near contemporaneous AOD measurement, coded according to the definition provided in Appendix A1 Table A1.2.2 should<br />

be assigned to the L2P confidence data a record field AOD_value.<br />

53 WP2-121 The source of AOD information must be specified in the L2P confidence data variable AOD_source (defined in Table<br />

MRG.0240<br />

2.1.6) and the difference in acquisition time between SST observation and AOD measure must be specified in the L2P<br />

confidence variable AOD_Dtime_from_SST.<br />

53 The GDS specifies the following rules:<br />

54 Rule<br />

WP2-122<br />

Rule 2.1.6a: A L2P AOD measurement (e.g., derived from satellite measurements) should be assigned to each SST pixel<br />

MRG.0220<br />

2.1.6a<br />

value using the AOD_value L2P confidence data variable. The AOD measurement nearest in space and time to the input<br />

pixel SST value should be used.<br />

54 Rule<br />

WP2-123<br />

Rule 2.1.6b: If no AOD measurement is available, an AOD value derived from an NWP system or aerosol forecast (See<br />

MRG.0230<br />

2.1.6b<br />

Appendix A3) nearest in space and time to the SST measurement should be used to set the value of AOD_value.<br />

54 Rule<br />

WP2-124<br />

Rule 2.1.6c: The source of data used to set the L2P confidence data variable AOD_value should be indicated in the<br />

MRG.0240<br />

2.1.6c<br />

confidence variable AOD_source as defined in Table 2.1.6.<br />

54 Rule<br />

WP2-125<br />

Rule 2.1.6d: The difference in time expressed in hours between the time of SST measurement and the time of AOD<br />

MRG.0250<br />

2.1.6d<br />

observations/analysis should be entered into the L2P confidence data variable AOD_Dtime_from_SST. A value of 25<br />

should be used to indicate unknown.<br />

54 Table<br />

WP2-126 Code value for the source of the data used to assign the L2P confidence value AOD_source MRG.0240<br />

2.1.6<br />

54 WP2-127 The following QC checks should be applied to all input AOD data sets. NASR<br />

54 WP2-128 The tests applied are simple tests designed to verify that AOD data are within realistic limits. NASR<br />

54 WP2-129 RDAC and GDAC may choose to implement more sophisticated QC procedures if deemed necessary. EXP.SYS.0100<br />

© 2004 SOC Page 144 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

54 WP-ID2.1.6.1 Gross AOD limit check<br />

54 WP2-130 This test identifies AOD values that fall outside of acceptable range of global AOD values. The GDS specifies the following<br />

NASR<br />

rule:<br />

54 Rule<br />

WP2-131<br />

Rule 2.1.6.1: A valid AOD observation/estimate must lie within a range of LowAODLimit < u < HighAODLimit. AOD<br />

QLC.0110<br />

2.1.6.1<br />

values that fall outside this range are considered erroneous and should not be used within the GDS.<br />

54 WP2-132 The LowAODLimit and HighAODLimit values are set in a L2P configuration file maintained at the RDAC and GHRSST-<br />

QLC.0120<br />

PP International Project Office.<br />

54 WP-ID2.2 Derivation of L2P confidence data and assignment of Sensor Specific Error Statistics (SSES).<br />

54 WP2-133 Due to characteristic differences between SST data sets derived from infrared and microwave sensors, the optimal<br />

CFD.0050<br />

assignment of error and confidence data for each measurement is different.<br />

54 WP2-134 In the case of infrared SST data sets, data are stratified according to their proximity to cloud and by wind speed (used to<br />

CFD.0320<br />

identify conditions favourable to stratification).<br />

54 WP2-135 Bias and standard deviation errors are assigned based on an analysis of the GHRSST-PP match-up database (MDB). L2P.6020<br />

54 WP2-136 A proximity confidence value is specified for each pixel based on the “proximity” of that pixel to several different criteria<br />

CFD.0320<br />

known to have a derogatory effect on the final SST value that is reported.<br />

55 WP2-137 The L2P confidence data record variable Proximity_Confidence defined in Table A1.4.2 in Appendix A1 is used to<br />

CFD.0320<br />

statistically derive and assign Sensor Specific Error Statistics (SSES) error estimate for each SST measurement.<br />

55 WP2-138 A long time series of data will be developed within the GHRSST-PP Match-up database (MDB) (containing limited<br />

NASR<br />

geographical coverage of in situ SST and L2P data) that are analysed on a regular (weekly/monthly) basis according to<br />

sensor and Proximity_Confidence values to derive that provide a mean bias and standard deviation value that can then<br />

be applied to all other SST measurements having the same Proximity_Confidence value thereby maximising the use of<br />

the limited coverage of the in situ observations. The technique relies on the fact that SST measurements are generally of a<br />

© 2004 SOC Page 145 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

reduced quality (and are therefore likely to have larger errors) the closer they are to other measurements that have been<br />

flagged as erroneous.<br />

55 WP2-139 The GDS specifies separate schemes to derive Proximity_Confidence values for infrared and microwave SST data<br />

CFD.0050<br />

streams.<br />

55 WP2-140 For SST data derived from infrared radiometers, sub-pixel cloud and wind speed are the main factors influencing the<br />

NASR<br />

quality of SST measurements.<br />

55 WP2-141 However, extremely large errors may be associated with atmospheric aerosol contamination due to dust storms or volcanic<br />

NASR<br />

eruptions that have an immediate and widespread impact. This type of error is not considered in the GDS-v1.0 but will be<br />

included in a revised GDS once v1.0 is operating.<br />

55 WP2-142 For microwave SST data streams, the proximity to rainfall, side lobe contamination in the proximity of land, high wind<br />

CFD.0100<br />

speed regimes modifying the surface emissivity of seawater in the 6-11 GHz frequency, radio frequency interference and<br />

proximity to sea ice are the main factors influencing the Proximity_Confidence value and the quality of SST retrieval. The<br />

approach adopted for the GDS-v1.0 is to flag suspect data in close proximity to rain, land, sea ice and at high wind speeds.<br />

55 WP2-143 Bias errors are assigned based on an analysis of the GHRSST-PP MDB. Standard deviation error increments associated<br />

CFD.0020<br />

with each error source are statistically derived based on look-up tables generated by the data provider (Remote Sensing<br />

<strong>System</strong>s) and through analysis of the MDB.<br />

55 WP-ID2.2.1 Derivation of L2P confidence data and assignment of SSES for Infrared SST data sets.<br />

55 WP2-144 Figure 2.2.1.1 shows a functional breakdown diagram of the processing operations required to derive L2P confidence data<br />

NASR<br />

and assign SSES for infrared SST data files.<br />

55 WP2-145 Initially, a series of quality control checks are performed on pixel-by-pixel bases that are designed to reject suspect data<br />

QLC-0045<br />

from the GDS.<br />

55 WP2-146 Proximity confidence values are then derived and SSES are assigned to each measurement. SSES are fully described in see lower down<br />

© 2004 SOC Page 146 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

WP-ID3.<br />

55 WP-ID2.2.1.1 Cosmetic data test<br />

55 WP2-147 Some SST data streams contain pixel values that are cosmetic rather than measured in order to provide a complete data<br />

see WP2-149<br />

product. For example, the ENVISAT ATS_NR__2P data stream may contain cosmetically filled pixels. These data are<br />

used to fill in data gaps across the swath following rectification of the curved ground swath made by the conical scan<br />

geometry used by the instrument.<br />

56 WP2-148 The purpose of this test is to extract the status of an input SST flag (if such a flag is present in the native input data<br />

NASR<br />

stream) and to reject all data that have been derived rather than measured by a sensor.<br />

56 The following GDS rule is specified:<br />

56 Rule<br />

WP2-149<br />

Rule 2.2.1.1: If an input SST data stream includes information that indicates that a pixel value is cosmetic rather than<br />

QLC.0050<br />

2.2.1.1<br />

measured, these data should be rejected from the L2P data file.<br />

56 Figure<br />

WP2-150<br />

Functional breakdown diagram of data processing steps to produce L2P data records for infrared satellite SST<br />

NASR<br />

2.2.1.1<br />

data files.<br />

56 WP2-151 WP-ID2.2.1.2 Cloud flag test<br />

56 WP2-152 Many input SST data streams provide a flag or a data value indicating that the input pixel is cloudy. The GDS specifies the<br />

NASR<br />

following rule:<br />

56 Rule<br />

WP2-153<br />

Rule 2.2.1.2: If an input SST pixel data value or associated confidence data indicate the pixel to be cloudy, these data<br />

QLC.0060<br />

2.2.1.2<br />

should be rejected from the L2P data file.<br />

56 WP-ID2.2.1.3 Sun glint test<br />

56 WP2-154 This test is used to set the L2P confidence data record variable SunGlint defined in Table A1.2.2 in Appendix A1. CFD.0300<br />

56 WP2-155 Depending on the local wind speed and sensor-sun view-geometry, sun-glint areas may be present in satellite SST data<br />

NASR<br />

streams. Normal sun glint occurs in reflection from the oceans when the sun is behind and to the side of the satellite.<br />

© 2004 SOC Page 147 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

56 WP2-156 Sun glint may degrade the SST value retrieved by affecting the quality of the cloud mask used by an infrared radiometer<br />

and consequently these regions should be flagged.<br />

56 WP2-157 Note that for polar orbiting satellite systems the problem is quite limited, but for geosynchronous orbit sensors it is<br />

SR entry<br />

CFD.0300<br />

CFD.0310<br />

NASR<br />

unavoidable.<br />

57 The GDS specifies the following rules:<br />

57 Rule<br />

WP2-158<br />

Rule 2.2.1.3a: If an indication of sun-glint is provided with an input SST pixel value, this should be used to set the<br />

CFD.0300<br />

2.2.1.3a<br />

SunGlint flag of the corresponding L2P confidence data record.<br />

57 Rule<br />

WP2-159<br />

Rule 2.2.1.3b: If an indication of sun-glint is not provided with an input pixel value, an appropriate method should be used<br />

CFD.0310<br />

2.2.1.3b<br />

to set the SunGlint flag of the L2P confidence data record that is based on the geometry of sun and satellite position and<br />

surface roughness (e.g., Gardashov and Barla, 2001).<br />

57 WP-ID2.2.1.4 Derivation of Infrared Proximity Confidence Values (IPCV)<br />

57 WP2-160 The major source of error associated with infrared SST data sets is contamination by cloud due to poor cloud clearing<br />

CFD.0320<br />

techniques or particularly difficult environmental conditions (e.g., clouds are sub-pixel in size, SST is highly dynamic<br />

(coastal zone), thin cirrus cloud is present or low lying fog) that are difficult to detect using current algorithms. For the GDS<br />

v-1.0, IPCV are determined as a function of deviation from a minimum SST climatology (as cloud contaminated SST are<br />

cooler than surrounding waters) and distance from the nearest cloud flagged pixel (as smaller clouds are often found in<br />

close proximity to larger clouds). This type of approach is successfully used at the EUMETSAT O&SI SAF (Brisson et al.,<br />

2001) and as part of the US Navy NAVOCEANO operational SST processing system.<br />

57 WP2-161 The GDS-v1.0 IPCV scale is defined in Table 2.2.1.4. CFD.0320<br />

57 Table<br />

WP2-162 Infrared proximity confidence value (IPCV) scale definitions. CFD.0320<br />

2.2.1.4<br />

57 WP2-163 Figure 2.2.1.4.1 presents a schematic lookup table describing how IPCV values are assigned. In this figure, four DT_min CFD.0320<br />

© 2004 SOC Page 148 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

thresholds (IPCVThresh1 - IPCVThresh3) and two cloud proximity distances (IPCV_D1 and IPCV_D2) delineate the set of<br />

data that correspond to the confidence scale provided in Table 2.2.4.1.<br />

57 WP2-164 It is expected that the IPCV scale will be expanded to identify an additional scheme that can be used to identify areas of<br />

NASR<br />

diurnal stratification based on the difference between the previous analysis field and the observations and the SSI.<br />

57 WP2-165 IPCV values are derived by considering the following GDS rules that have been found sufficient to identify most poor<br />

NASR<br />

quality data (Brisson et al., 2001):<br />

57 Rule<br />

WP2-166<br />

Rule 2.2.1.4a: Typically, sub-pixel clouds often contaminate data immediately adjacent to cloudy areas. The spatial<br />

CFD.0340<br />

2.2.1.4a<br />

distance, D (expressed in km) of an SST observation to clearly flagged cloud contaminated areas is used to identify these<br />

measurements. Two threshold values set by the GDS, IPCV_D1 and IPCV_D2, for each L2P data stream. The distance<br />

IPCV_D2 is set on the assumption that cloudy pixels beyond this distance do not influence the SST measurement. The<br />

distance IPCV_D1 is set assuming that cloudy pixels where IPCV_D1 < nearest_cloudy_pixel < IPCV_D2 have a<br />

reasonable probability of being cloud contaminated. Pixels within a distance less than D1 are assumed to have a high<br />

probability of cloud contamination. D values will depend on the spatial resolution and the geometry of each individual<br />

satellite sensor and must be established for each sensor. For example, D values for geostationary observations of 5km will<br />

be different from polar orbiting radiometer data of 1.1km.<br />

57 Figure<br />

WP2-167<br />

Schematic diagram showing how IPCV values are derived based on the deviation from a minimum SST climatology<br />

CFD.0330<br />

2.2.1.4.1<br />

defining the coolest expected SST value and the distance from the nearest cloud flagged pixels.<br />

CFD.0320<br />

58 Rule<br />

WP2-168<br />

Rule 2.2.1.4b: The second IPCV test uses the deviation from a minimum SST climatological value defined in the L2P<br />

CFD.0350<br />

2.2.1.4b<br />

confidence data record variable DT_min discussed in Section WP-ID2.1. Three threshold values for DT_min will be set by<br />

the GDS L2P configuration file IPCVThresh1 - IPCVThresh3.<br />

58 Rule<br />

WP2-169<br />

Rule 2.2.1.4c: A static map of upwelling regions may be used to decide if a pixel is cloud contaminated in the case of IPCV<br />

CFD.0360<br />

2.2.1.4c<br />

value 6. Maps will be developed by RDAC as appropriate for each region of interest.<br />

© 2004 SOC Page 149 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

58 WP2-170 Note that the actual values for IPCVThres1-3 and IPCV_D1 and IPCV_D2 are expected to vary for each sensor (Brisson<br />

SR entry<br />

CFD.0370<br />

et al., 2001) and the exact value of these parameters is a subject of on-going research and development within the<br />

GHRSST-PP. IPCVThres1-3, IPCV_D1 and IPCV_D2 values will be maintained by the GHRSST-PP International Project<br />

Office and stored in the L2P configuration file that is specific to each RDAC or GDAC.<br />

58 WP2-171 The derivation of these values will initially use best estimates until a sufficiently large population of data within the<br />

CFD-0320<br />

GHRSST-PP MDB is available (e.g., those suggested by Brisson et al. (2001).<br />

58 WP2-172 The filename of this configuration file should be written into each L2P file using the optional global netCDF attribute<br />

“References” as described in Appendix A1.2.1.<br />

Format<br />

requirement<br />

58 WP2-173 It is expected that the methodology used to derive IPCV may be modified in the GDSv2. EXP.L2P.0030<br />

59 WP-ID2.2.1.5 Assign Sensor Specific Error Statistic<br />

59 WP2-174 This processing step is used to assign the L2P confidence data record variable bias and sd defined in Table A1.2.2 in<br />

Appendix A1.<br />

L2P.6010<br />

L2P.6020<br />

59 The GDS specifies the following rules:<br />

59 Rule<br />

WP2-175<br />

Rule 2.2.1.5a: If pixel bias error and sd error statistics are provided by a data provider, these values may be assigned to<br />

L2P.6010<br />

2.2.1.5a<br />

the L2P bias and sd variables respectively. In this case, the L2_native_bias and L2_native_sd flags of the<br />

corresponding L2P confidence data record should also be set as these error estimates are independent of GHRSST-PP<br />

methods.<br />

59 WP2-176 If native bias and sd error values are not available or an RDAC chooses to ignore the error statistics provided by a data<br />

provider, these must be assigned based on the L2P Proximity_Confidence variable using sensor specific error statistics<br />

L2P.6020<br />

CFD.0020<br />

(SSES) as described in WP-ID3. In this case the L2_native_bias and L2_native_sd. flags of the corresponding L2P<br />

confidence data record should be set to zero.<br />

59 The GDS specifies the following rule:<br />

© 2004 SOC Page 150 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

59 Rule<br />

WP2-177<br />

Rule 2.2.1.5b: If pixel bias error and sd error statistics are not available for SST measurements, these values should be<br />

L2P.6020<br />

2.2.1.5b<br />

assigned the most recent SSES bias and sd values for the each data stream and L2P confidence data record<br />

Proximity_Confidence value. In addition, the L2_native_bias and L2_native_sd. flags of the corresponding L2P<br />

confidence data record should be set to zero.<br />

59 WP2-178 It is expected that the IPCV and SSES schemes will develop considerably throughout the GHRSST-PP. EXP.L2P.0030<br />

59 WP-ID2.2.2 Quality control of satellite MW SST data sets and assignment of pixel confidence data.<br />

59 WP2-179 SSES and L2P confidence and data must be derived for every MW SST measurement contained within an input SST data<br />

stream.<br />

see<br />

lines<br />

following<br />

59 WP2-180 The processing is different from infrared data streams since the native MW fields already contain wind speed, rain, and<br />

CFD.0050<br />

other ancillary parameters required to calculate the coded bitfields identifying suspect pixels.<br />

59 WP2-181 Specific thresholds, limits, reference files, and processing rules that are used to process each data set are stored in a<br />

CFD.0220<br />

system wide L2P microwave L2P configuration file.<br />

59 WP2-1820 This allows easy modification of critical values and other configurable parameters without the need for software code<br />

CFD.0220<br />

changes.<br />

59 WP2-183 Figure 2.2.2.1 shows a functional breakdown diagram of the processing steps required to derive confidence data and<br />

NASR<br />

assign SSES to microwave SST measurements.<br />

59 WP2-184 The output of these processing steps is a GHRSST-PP L2P netCDF data file described in Appendix A1.4. format<br />

59 WP2-185 Table A1.4.2 in Appendix A1 describes the format of GDS L2P pixel confidence data records. format<br />

59 WP2-186 The following sections describe in detail the MW QC tests and pre-processing operations performed on each data file. NASR<br />

59 WP-ID2.2.2.1 SST Climatology limit check<br />

59 WP2-187 This test is identical to 2.1.1 described above. QLC.0070<br />

QLC.0080<br />

© 2004 SOC Page 151 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

60 Figure<br />

WP2-188<br />

Functional breakdown diagram of data processing steps to produce L2P data records for microwave satellite SST<br />

NASR<br />

2.2.2.1<br />

data files.<br />

60 WP-ID2.2.2.2 Test for possible sidelobe contamination<br />

60 WP2-189 Microwave SST observations within about 50 to 100 km from land and sea ice are affected by the warm emission of land<br />

NASR<br />

and ice entering the antenna side-lobes. The effect varies by region and season (e.g., in monsoon climates where more<br />

contamination is expected in wet ground conditions) and is difficult to quantify without having access to engineering data<br />

provided by the satellite sensor.<br />

60 WP2-190 It cannot be quantified using buoy match-ups in the coastal zone. NASR<br />

60 WP2-191 Instead, global reference maps identifying regions with persistent sidelobe contamination will be generated by Remote<br />

NASR<br />

Sensing <strong>System</strong>s to provide an estimate the retrieval error in coastal zones and near sea ice margins.<br />

60 WP2-192 The geolocation of each pixel will be used to look up possible contamination from the reference map. NASR<br />

61 WP2-193 Since the effect is different for the 10.7 GHz and 6.9 GHz retrievals, it must be calculated separately for TMI and AMSR-E<br />

NASR<br />

radiometers. We will approach this problem in two ways. Initially, TMI will be collocated with VIRS infrared SSTs while<br />

AMSR-E will be collocated with MODIS SST retrievals. The bias and variability in the STD will be calculated as the STD as<br />

a function of geographic location minus the minimum STD as a function of geographic location. This may be done using<br />

only retrievals within 200 km of land or ice or it may be done globally. There are several issues with this methodology that<br />

should be considered. There will be some overlap with dependencies on SST and wind speed. If the check is only used<br />

close to land, it will ignore possible side-lobe effects near frontal features, but if it is used globally, increased variability from<br />

the IR sensors due to water vapour or stray light will be included. It is expected that further work will be required to<br />

establish the optimum configuration of sidelobe contamination maps.<br />

61 The GDS specifies the following rule:<br />

61 Rule WP2-194 · Rule 2.2.2.2a: If a pixel is located within an area of known sidelobe contamination as defined by sensor specific CFD.0110<br />

© 2004 SOC Page 152 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

2.2.2.2a<br />

sidelobe contamination maps, bit 0 of the L2P confidence data variable MW_SST_flags should be set.<br />

61 Rule<br />

WP2-195<br />

· Rule 2.2.2.2b: If a sidelobe contamination map is not available, any pixel within sidelobe_distance_threshold<br />

CFD.0120<br />

2.2.2.2b<br />

(specified in the L2P configuration file in km) d should flagged as contaminated.<br />

61 WP-ID2.2.2.3 Strict test for possible rain contamination<br />

61 WP2-196 SST cannot be retrieved in raining conditions and data flagged as rain contaminated must be rejected from the L2P data<br />

QLC.0065<br />

file. The GDS specifies the following rule:<br />

61 Rule<br />

WP2-197<br />

· Rule 2.2.2.3: If a measurement is identified as rain contaminated in the native data input file, this measurement<br />

QLC.0065<br />

2.2.2.3<br />

should not be included in a L2P data file.<br />

61 WP-ID2.2.2.4 Relaxed test for possible rain contamination<br />

61 WP2-198 A recognized problem in the MW SST retrievals is unidentified rain contamination in the SST, often nearby raining pixels.<br />

NASR<br />

Low rain rates or small areas of rain within the large MW footprint are difficult to identify and in many cases, the quality of<br />

SST retrievals surrounding a rain flagged pixel are degraded by unresolved rain.<br />

61 WP2-199 For this reason, pixels within a defined radius relaxed_MW_rain_distance (specified in km and held within the L2P<br />

CFD.0140<br />

configuration file) of a raining pixel will be flagged as potentially rain contaminated.<br />

61 WP2-200 In addition, a comparison of the previous analysed SST values will be used to reject or accept the measurement for<br />

CFD.0150<br />

inclusion within a L2P file if the difference is greater than a predefined threshold specified in K and held in the L2P<br />

configuration file variable relaxed_rain_threshold.<br />

61 The GDS specifies the following rules:<br />

61 Rule<br />

WP2-201<br />

· Rule 2.2.2.4a: If a pixel lies within a radius of relaxed_MW_rain_distance specified in km of a clearly flagged rain<br />

CFD.0140<br />

2.2.2.4a<br />

contaminated pixel, bit 1 of the L2P confidence data variable MW_SST_flags should be set.<br />

61 Rule<br />

WP2-202<br />

· Rule 2.2.2.4b: If the difference between the most recent GHRSST-PP analysis SST and the potentially<br />

CFD.0150<br />

2.2.24b<br />

contaminated pixel is greater than relaxed_rain_threshold (K), the pixel should not be included in a L2P data file.<br />

© 2004 SOC Page 153 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

62 WP-ID2.2.2.5 TMI SST retrieval below 285 K<br />

62 WP2-203 TMI retrieves SST primarily from the 10.7 GHz channel, which has decreased sensitivity (and therefore increased error) at<br />

CFD.0160<br />

SSTs below 285K. Although the bias at lower temperatures is negligible, the standard deviation increases (as reflected in<br />

the SSES). TMI data having retrievals below 285K must be flagged according to the following GDS rule:<br />

62 Rule<br />

WP2-204<br />

· Rule 2.2.2.5: If a TRMM TMI SST measurement value is less than 285K, bit 2 of the L2P confidence data variable<br />

CFD.0160<br />

2.2.2.5<br />

MW_SST_flags should be set.<br />

62 WP-ID2.2.2.6 High Wind retrieval, above 12 ms -1<br />

62 WP2-205 Uncertainties in the emissivity model at high winds and increased errors in the directional model, result in larger errors<br />

CFD.0170<br />

above 12 ms -1 . These errors should not vary temporally or spatially since they are due to uncertainty in the estimate of<br />

emissivity. Although this increased error will be reflected in the SSES, some data users may chose to simply exclude these<br />

retrievals as suspect.<br />

62 The GDS specifies the following rule:<br />

62 Rule<br />

WP2-206<br />

· Rule 2.2.2.6a: If a microwave wind speed measurement contemporaneous with a microwave SST measurement is<br />

CFD.0170<br />

2.2.2.6a<br />

greater than 12 ms -1 , bit 3 of the L2P confidence data variable MW_SST_flags should be set.<br />

62 Rule<br />

WP2-207<br />

· Rule 2.2.2.6b: If a contemporaneous microwave wind speed measurement is unavailable, the nearest NWP analysis<br />

CFD.0180<br />

2.2.2.6b<br />

in space and time toi the SST measurement should be used to evaluate Rule 2.2.2.6a.<br />

62 WP-ID2.2.2.7 Derive a Microwave Proximity Confidence Value (MPCF)<br />

62 WP2-208 MPCV data are analogous to Infrared Proximity Confidence Values (IPCV) and provide a method to better understand the<br />

CFD0210<br />

character of microwave SST and their relationship to the SST1m. Microwave SST measurements are known to be of a<br />

reduced quality when they are in close proximity to clearly flagged rainfall, due to side lobe contamination in the proximity<br />

of land and sea ice. Also, at wind speeds < 2 ms -1 , a cool skin effect may also exist whereas at high wind speeds the<br />

emissivity model used to derive SST is poorly defined. These factors may be used to derive a Proximity_Confidence<br />

© 2004 SOC Page 154 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

value for microwave SST retrievals that can then be used to derive a MPCV for use within the GHRSST-PP MDB and<br />

ultimately the assignment of SSES for MW SST data. The initial MPCV scheme proposed for the GDS is based on the<br />

following criteria:<br />

62 WP2-2090 1. Physical distance of an SST measurement from the nearest rain flag. CFD0200<br />

62 WP2-210 2. Physical distance of an SST measurement from nearest land. CFD0200<br />

62 WP2-211 3. Physical distance of an SST measurement from nearest sea ice (held in bit 4 of MW_SST_flags). CFD.0190<br />

62 WP2-212 4. Conditions when the surface wind speed is > 12 ms -1 and is likely to modify the emissivity conditions of the sea surface<br />

CFD.0170<br />

in the 6-11GHz frequency range.<br />

62 WP2-213 5. Low wind speed conditions < 2ms -1 when a cool skin/warm layer effect may be present.<br />

62 WP2-214 6. For the special case of TMI SST, conditions when the SST is < 285K. CFD0160<br />

63 WP2-215 The proposed scheme is an initial configuration for MPCV and further research is required within the GHRSST-PP to<br />

NASR<br />

establish an optimal scheme, which should be undertaken at RDAC once sufficient data are available within the GHRSST-<br />

PP MDB.<br />

63 The MPCV scale is shown in Table 2.2.2.7.1.<br />

63 Table<br />

WP2-216 Microwave proximity confidence value (MPCV) scale definitions. CFD.0210<br />

2.2.2.7.1<br />

63 WP2-217 If any of the test bitflags are set for the MW retrieval, the data will be flagged as questionable. CFD.0210<br />

63 WP2-218 In addition two surface wind speed threshold values are used to delineate the wind speed conditions for which the best<br />

SST retrievals and relationship to the in situ observations is expected using the threshold parameters MPCV_wind1 and<br />

CFD.0210<br />

CFD.0220<br />

MPCV_wind2.<br />

63 WP2-219 The values for MPCV_windn shown in Figure 2.2.2.7.1 are expected to vary for a given sensor and the exact value of<br />

CFD.0220<br />

these parameters is a subject of on-going research and development within the GHRSST-PP.<br />

© 2004 SOC Page 155 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

63 WP2-220 Finally, the physical distance in km specified in the L2P configuration file variable MPCV_proximity is used to identify<br />

pixels that are far away from contaminating effects of sea ice (flagged in WP-ID2.1.3).<br />

SR entry<br />

CFD.0190<br />

CFD.0220<br />

CFD.0210<br />

63 WP2-221 All threshold values will be stored in a L2P configuration file specific to each RDAC or GDAC and maintained at the<br />

GHRSST-PP international project office that can be modified based on an assessment of sensor specific values.<br />

CFD.0220<br />

EXP.L2P.0050<br />

63 WP2-222 This will only be possible once the GHRSST-PP MDB has sufficient data available for analysis. EXP.L2P.0050<br />

63 WP2-223 The filename of this configuration file should be written into each L2P file using the optional global netCDF attribute<br />

format<br />

“References” as described in Appendix A1.2.1.<br />

63 WP2-224 It is expected that the methodology used to derive MPCV will be modified based on experience in the GDSv2. EXP.L2P.0040<br />

64 Figure<br />

WP2-225<br />

Schematic diagram showing the initial scheme (GDS-v1.0) proposed to derive MPCV values. Note that the TMI is<br />

see Fig. 3-2 of<br />

2.2.2.7.1<br />

subject to an additional lookup based on the actual SST value which is not shown.<br />

SRD<br />

64 WP-ID2.2.2.8 Assign L2P Sensor Specific Error Statistics<br />

64 WP2-226 This processing step is used to assign the L2P confidence data record variable bias and sd defined in Table A1.2.2 in<br />

see below<br />

Appendix A1. The GDS specifies the following rules:<br />

64 Rule<br />

WP2-227<br />

Rule 2.2.2.8a: If pixel bias error and sd error statistics are provided by a data provider, these values may be assigned to<br />

L2P.6010<br />

2.2.2.8a<br />

the L2P bias and sd variables respectively. In this case, the L2_native_bias and L2_native_sd flags of the<br />

corresponding L2P confidence data record should also be set as these error estimates are independent of GHRSST-PP<br />

methods.<br />

64 WP2-228 If native bias and sd error values are not available or an RDAC chooses to ignore the error statistics provided by a data<br />

provider these must be assigned based on the L2P Proximity_Confidence (MPCV) variable using sensor specific error<br />

L2P.6020<br />

CFD.0020<br />

statistics (SSES) as described in WP-ID3. In this case the L2_native_bias and L2_native_sd. flags of the corresponding<br />

L2P confidence data record should be set to zero.<br />

© 2004 SOC Page 156 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

64 The GDS specifies the following rule:<br />

Rule<br />

WP2-229<br />

Rule 2.2.2.2b: If pixel bias error and sd error statistics are not available for SST measurements, these values should be<br />

L2P.6020<br />

2.2.2.2b<br />

assigned the most recent SSES bias and sd values for each data stream and pixel Proximity_Confidence value. In<br />

addition, the L2_native_bias and L2_native_sd. flags of the corresponding L2P confidence data record should be set to<br />

zero.<br />

WP2-230 It is expected that the MPCV and SSES schemes will develop considerably throughout the GHRSST-PP. EXP.SYS.0100<br />

WP-ID2.3 Evaluate L2P data<br />

64 WP2-231 Each RDAC and GDAC centre will evaluate whether or not a L2P data set should be admitted to the GDS processor based<br />

L2P.7000<br />

64 WP-ID2.3.1 Reject L2P data file<br />

on objective criteria Including completeness, timeliness and quality.<br />

64 WP2-232 If a L2P data file has been rejected by an RDAC/GDAC, an error message must be raised and logged at the GHRSST-PP<br />

CTR.0550<br />

ERRLOG system as described in Appendix A7.2.<br />

64 WP2-233 No MMR_FR metadata are prepared for rejected L2P data streams. CTR.0550<br />

64 WP-ID2.3.2 Accept L2P data file<br />

64 WP2-234 If a L2P data file is accepted for further use in the GDS, then a L2P data file should be generated as described in WP-<br />

L2P.7000<br />

ID2.4.<br />

64 WP2-235 A corresponding MMR_FR metadata record should be generated as specified in WP-ID2.5. CTR.0530<br />

64 WP-ID2.4 Format L2P data file<br />

64 WP2-236 Each L2P data set should be prepared, formatted and archived as a netCDF file following the data format described in<br />

L2P.0060<br />

Appendix A1.2 and A1.4.<br />

64 WP-ID2.5 Generate and register L2P MMR_FR metadata record<br />

64 WP2-237 For each L2P data a corresponding MMR_FR must be prepared by the responsible RDAC/GDAC processor and registered CTR.0530<br />

© 2004 SOC Page 157 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

at the MMR as soon as the L2P data become available to the user community.<br />

65 WP-ID2.5.1 Format L2P MMR_FR metadata record<br />

65 WP2-238 A MMR_FR metadata record should be prepared for each L2P data file according to the specification provided in Appendix<br />

CTR.0530<br />

A6.2.<br />

65 WP-ID2.5.2 Register L2P MMR_FR metadata record with MMR system<br />

65 WP2-239 MMR_FR metadata records should be delivered to the MMR system as soon as possible after the L2P data set has been<br />

CTR.0580<br />

evaluated following the procedure described in Appendix A6.4.<br />

65 WP2-240 RDAC should aim to register MMR_FR records within 60 minutes of L2P data production in order that users are informed<br />

CTR.0580<br />

in a timely manner.<br />

65 WP2-241 It is recognised that, if the MMR_FR record is submitted by e-mail, delays in registration are beyond the control of the<br />

CTR.0110<br />

RDAC. At the 4 th GHRSST-PP Science Team workshop, it was proposed to enable MMR data records to be delivered to<br />

the MMR via ftp push from the RDAC to circumvent this problem.<br />

67 WP-ID3 The GHRSST-PP Matchup <strong>Data</strong> base (MDB) and derivation of Sensor Specific Error Statistics (SSES)<br />

Description:<br />

67 WP3-001 This WP describes the GHRSST-PP match up database system (MDB) and the MDB data records that should be<br />

NASR<br />

produced at each RDAC.<br />

67 WP3-002 The GDS requires that error estimates are assigned to every satellite SST measurement during the derivation of L2P data<br />

NASR<br />

products. As an in situ observation is not available for every SST measurement, Sensor Specific Error Statistics (SSES)<br />

are derived by a statistical analysis of MDB data records for a given sensor type and time period. SSES consist of a<br />

statistical mean bias and standard deviation (sd.) error estimate that is correlated to the mean L2P Proximity_confidence<br />

value for a given sensor. These values are then assigned to all L2P measurements that have the same<br />

Proximity_confidence value.<br />

© 2004 SOC Page 158 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

67 WP3-003 In addition, The MDB system also provides a resource that can be used to validate GHRSST-PP data products. NASR<br />

67 WP3-004 In the case of microwave SST, the derivation of SSES also requires that error components are assigned to each pixel<br />

NASR<br />

based on static look-up-tables of mean error due to sidelobe contamination, proximity to rain and wind speed. These<br />

tables will be provided by Remote Sensing <strong>System</strong>s (RSS) for the TMI and AMSR-E (and in the future, Windsat)<br />

microwave instruments.<br />

67 WP3-005 The main source of in situ observations used in this system will be those available from the GTS supplied through<br />

NASR<br />

international in situ data centres (e.g., MED, CORIOLIS, FNMOC) at which QC procedures are used to filter out poor<br />

quality data.<br />

67 WP3-006 These data centres only provide in situ observations in a delayed mode (typically 24-72 hours, although delays may be<br />

MDB.0010<br />

longer than this) and therefore RDAC and GDAC can only provide MDB records in a delayed mode.<br />

67 WP3-007 RDAC are expected to arrange access to regional in situ data sources that are not available via the GTS system but can<br />

MDB.0120<br />

be included in the MDB system.<br />

67 WP3-008 The GHRSST-PP reanalysis project will be able to make use of other in situ observations that are unavailable in a NRT<br />

NASR<br />

mode.<br />

inputs<br />

68 WP3-009 • L2P data files MDB.0040<br />

68 WP3-010 • In situ ocean-atmosphere measurements from buoys, drifters and ships MDB.0020<br />

68 WP3-011 • SSES configuration file NASR<br />

68 WP3-012 • Static Look-up-Table (LUT) of microwave SST sidelobe contamination error NASR<br />

68 WP3-013 • Static LUT of microwave SST rain contamination errors NASR<br />

68 WP3-014 • Static LUT of microwave SST wind speed errors NASR<br />

outputs<br />

© 2004 SOC Page 159 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

68 WP3-015 • MDB records delivered to the GHRSST-PP MDB CTR.0580<br />

68 WP3-016 • SSES for each sensor derived on a on a regular basis NASR<br />

Acceptance tests<br />

68 WP3-017 • MDB records are formatted correctly and are delivered to the GHRSST-PP MDB system ATPD<br />

68 WP3-018 • SSES analysis is performed on a weekly basis and SSES are published to all RDAC and via the GHRSST-PP UIS ATPD<br />

Metric for performance assessment:<br />

68 WP3-019 • MDB records are produced in a timely manner using a target of within 24 hours from when the in situ data are made<br />

available by a data centre.<br />

ATPD<br />

MDB.0010<br />

68 WP3-020 • MDB records are provided and ingested at the MDB in near real time. MDB rejection rate < 2% CTR.0130<br />

68 WP3-021 The MDB will be physically realised as a complete relational database system that can be analysed using structured query<br />

NASR<br />

language (SQL) routines.<br />

68 WP3-022 These data provide a collective resource that is shared and populated by all RDAC and GDAC within the GHRSST-PP to<br />

NASR<br />

generate an independent assessment of the absolute error of all satellite SST data streams.<br />

68 WP3-023 The MDB forms the core data resource for both the derivation of Sensor Specific error statistics (SSES) that are assigned<br />

NASR<br />

to all L2P data in WP-ID2 and for the on-going validation of satellite data streams described in WP-ID5.<br />

68 WP3-024 Both of these rely on the regular population of the GHRSST-PP MDB using near real time satellite and in situ observations. MDB.0020<br />

69 WP3-025 The generation of individual GHRSST-PP MDB records and their storage in the GHRSST-PP MDB database system CTR.0580<br />

69 WP3-026 GHRSST-PP L2P data files for each individual satellite sensor are matched in space and in time with in situ observations<br />

MDB.0050<br />

(WP-ID3.1.1) if available.<br />

69 WP3-027 Time and space constraints that define near contemporaneous match-up criteria are stored in an MDB configuration file<br />

MDB.0230<br />

which will be maintained by the GHRSST-PP International Project Office<br />

69 WP3-028 If in situ data can be matched with satellite data then a MDB data record is prepared and formatted according to the MDB.0210<br />

© 2004 SOC Page 160 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

specification set out in Appendix A4.1 (WP-ID3.1.2).<br />

69 WP3-029 in situ observations may be matched to SST measurements derived from different satellite sensors and are thus used<br />

MDB.0250<br />

more than once in the system providing consistency across GHRSST-PP data products<br />

69 WP3-030 only 1 instance of each in situ observation (the nearest in space and time to the satellite data) is used to define a matchup<br />

MDB.0260<br />

record<br />

70 WP3-031 SSES are published on the GHRSST-PP web site as part of the GHRSST-PP <strong>User</strong> Information Service (UIS) and all<br />

NASR<br />

RDAC/GDAC are informed of the latest issue of SSES for use in the generation of GHRSST-PP L2P data products<br />

WP-ID3.1 Generation of Match Up <strong>Data</strong>base (MDB) records<br />

70 WP3-032 The GHRSST-PP MDB system is a relational database containing all GHRSST-PP MDB data records. The GHRSST-PP<br />

NASR<br />

MDB will be implemented and maintained at the JPL PO.DAAC using the MySQL /database system<br />

70 WP3-033 MDB data records should be produced by RDAC and ingested at the MDB in near real time (1-7 days[1]) in order for SSES<br />

MDB.0030<br />

to be generated in a timely manner. [1] Note that the MDB will be used extensively by the GHRSST-PP reanalysis project<br />

so that the time constraints specified here should be interpreted in the broadest possible sense; if a match-up data record<br />

has been established, it should be delivered to the MDB system regardless of the timeliness of the data.<br />

70 WP3-034 Timeliness of MDB data is related to the availability of in situ data. Time penalties are incurred because data centres<br />

MDB.0030<br />

careful quality control and validate in situ measurements<br />

70 WP3-035 Collaboration with operational data centres that are well practiced in the QC of in situ observations must be established<br />

and maintained by RDAC and GDAC especially in the case of data that are made available on the GTS<br />

70 WP3-036 The GHRSST-PP expects to work closely with the CORIOLIS data centre (France), the Japanese Meteorological Agency<br />

MDB.0130<br />

MDB.0110<br />

MDB.0110<br />

(JMA), and the MEDS data centre (Canada) amongst others as described in Appendix A3.4.5 and Appendix A3.4.6<br />

70 WP3-037 the MDB is open to all in situ observations that are deemed of sufficient quality by the RDAC and GDAC teams MDB.0120<br />

MDB.0110<br />

© 2004 SOC Page 161 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

WP-ID3.1.1 Derivation of MDB data record<br />

70 WP3-038 The data inputs to GHRSST-PP MDB records are L2P data streams and in situ observations. MDB.0020<br />

70 WP3-039 For an MDB data record to be accurate and legitimate the comparison must be between like measurements although in the<br />

MDB.0270<br />

case of SSTskin observations, in situ data are extremely scarce<br />

70 WP3-040 Instead, emphasis is placed on the use of in situ SST1m data derived from drifting/moored buoy and ship observations. MDB.0270<br />

71 WP3-041 in situ data drawn down directly from the GTS must be QC’d locally before use MDB.0130<br />

71 WP3-042 Other data streams are only available to national institutes where they must be pre-processed and matched to satellite<br />

MDB.0120<br />

data on a case by case basis.<br />

71 WP3-043 each RDAC and GDAC centre must be responsible for implementing basic in situ QC procedures using local<br />

MDB.0130<br />

ocean/atmosphere knowledge and experience<br />

71 WP3-044 limits of coincidence must be applied to constrain the contributions to the error budget of the SSES/validation procedure<br />

MDB.0230<br />

from these sources to acceptable levels (say


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

MDB_space_constraint.<br />

71 WP3-049 [1] Based on discussions at the 4th GHRSST-PP Science Team Workshop, Los Angeles, USA, September 2003,<br />

MDB.0240<br />

MDB_time_constraint = 25 km radius of the pixel centre point and MDB_space_constraint should be within ±6 ours of the<br />

satellite overpass. These constraints are considered the absolute maximum and all efforts should be made to reduce<br />

these to the smallest deviations from zero as possible.<br />

71 WP3-050 In situ data that cannot be matched to better than these criteria should be ignored MDB.0240<br />

71 WP3-051 RDAC and GDAC should aim to use data with the highest spatial and temporal coincidence MDB.0210<br />

71 Rule<br />

WP3-052<br />

Because of the day-night asymmetry in the growth and decay of the diurnal thermocline, more stringent conditions for<br />

MDB.0240<br />

3.1.2a<br />

acceptable time intervals may be required during the day than at night<br />

71 Rule<br />

WP3-053<br />

For these and ambiguous cases, such as close to the terminator, the smallest spatial separation and shortest possible time<br />

MDB.0240<br />

3.1.2a<br />

interval should be sought.<br />

72 Rule<br />

WP3-054<br />

Only one instance of each in situ observation for each individual satellite sensor should occur within the MDB i.e. the<br />

MDB.0250<br />

3.1.2b<br />

satellite data having the smallest deviation in space and time from the in situ observation should be used within the MDB<br />

MDB.0210<br />

and all other duplicates should be removed<br />

72 Rule<br />

WP3-055<br />

L2P data adjacent to the validation measurement must be extracted to provide information on the spatial variability in the<br />

MDB.0320<br />

3.1.2c<br />

vicinity of the validation point<br />

72 Rule<br />

WP3-056<br />

Appropriate ‘missing data’ values will be used if any of the elements of the pixel array extend beyond the edges of the<br />

MDB.0320<br />

3.1.2c<br />

instrument swath.<br />

72 Rule<br />

WP3-057 A 5 x 5 array of L2P data should be extracted centred on the validation pixel. MDB.0320<br />

3.1.2c<br />

72 Rule<br />

WP3-058 All L2P confidence data each pixel stored in a MDB record must accompany the extracted data MDB.0310<br />

3.1.2d<br />

© 2004 SOC Page 163 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

72 Rule<br />

WP3-059<br />

A minimum delay of 24 hours is to be expected when using data from operational in situ data centres due to the latency<br />

MDB.0030<br />

3.1.2e<br />

incurred by QC procedures<br />

72 Rule<br />

WP3-060 all MDB records are admissible irrespective of their timeliness for use by the GHRSST-PP reanalysis project MDB.0030<br />

3.1.2e<br />

72 Rule<br />

WP3-061<br />

Additional geophysical fields will be required to provide insight into any uncorrected perturbations to the SSTs which might<br />

MDB.0330<br />

3.1.2f<br />

result in systematic errors. All available in situ parameters should be included in the MDB record using the spare<br />

WP-ID3.1.2 Format of MDB record<br />

72 WP3-062 Each in situ SST matched to a near contemporaneous satellite SST will be formatted as an XML text file following the<br />

Format<br />

format described in Appendix A4.<br />

WP-ID3.1.3 Insert MDB record into the GHRSST-PP MDB system<br />

72 WP3-063 MDB data records should be sent to the GHRSST-PP MDB either as an e-mail message or via ftp put CTR.0580<br />

72 WP3-064 An error message will be returned to the data provider if the MDB record is incorrectly formatted. NASR<br />

WP-ID3.2: Derivation of SSES using MDB records<br />

78<br />

72-<br />

79-<br />

WP-ID3.3: Publish SSES on GHRSST-PP web site and inform RDAC/GDAC<br />

80<br />

WP-ID4: Generation of L4 Analysed data products (L4)<br />

Description<br />

81 WP4-001 The GDS will produce an analysed estimate of SSTfnd every 24 hours using analysis procedures that depend on L2P data<br />

SYS.0080<br />

products as input.<br />

© 2004 SOC Page 164 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

81 WP4-002 GDS L4SSTfnd data products provide complete coverage SST data fields that are free of gaps for each APPW NASR<br />

81 WP4-003 the GHRSST-PP Science Team have specified that a diversity of approach is required to establish the optimal analysis<br />

DES.SYS.0040<br />

method that will maximise the impact of complementary satellite observations<br />

81 WP4-004 Standard GHRSST-PP L4SSTfnd products will be generated on a grid scale of 1/12° every 24 hours. NASR<br />

81 WP4-005 Ultra-high resolution (UHR) L4SSTfndUHR data products will be generated for regional areas at a grid scale of 1/48°<br />

L4A.0020<br />

latitude x longitude (~2km) every 24 hours<br />

inputs<br />

81 WP4-006 • L2P data files L4A.0030<br />

81 WP4-007 • L4 data processor configuration file L4A.0030<br />

outputs<br />

81 WP4-008 • Global coverage L4SSTfnd data product (L4FND) and associated confidence and error statistics for each APPW NASR<br />

81 WP4-009 • Regional coverage L4SSTfndUHR data product (L4FNDUHR) and associated confidence and error statistics for each<br />

APPW<br />

L4A.0100<br />

L4A.0051<br />

81 WP4-010 • MMR_FR metadata record for each L4 product NASR<br />

81 WP4-011 • Extracted HR-DDS data granules and HR-DDS MMR_FR metadata record for each data granule DDS.0040<br />

INT.DDS.0030<br />

acceptance tests<br />

81 WP4-012 • Identical L4 output data products are produced at each RDAC and GDAC when the GDS test data set us used compared<br />

to the reference processor<br />

ATPD<br />

EXP.L2P.0070<br />

81 WP4-013 • L4 MMR_FR metadata entries are timely and accurate ATPD<br />

81 WP4-014 • HR-DDS L4 data products are extracted correctly ATPD<br />

81 WP4-015 • L4 HR-DDS-MMR –FR metadata is correct and accepted by the HR-DDS MMR system ATPD<br />

© 2004 SOC Page 165 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

Metric for performance assessment<br />

82 WP4-016 • L4 products and HR-DDS L4 granules are produced in a timely manner ATPD<br />

82 WP4-017 • L4 MMR_FR metadata records provided and ingested at MMR and HR-DDSS MMR in real time. MMR and HR-DDS<br />

MMR rejection rate < 2%<br />

PER.DDS.0020<br />

CTR.0130<br />

82 WP4-018 • OPLOG L4_DPSR records correct and timely ATPD<br />

82 WP4-019 L4 SST data products provide an estimate of the foundation SST (SSTfnd). L4A.0054<br />

82 WP4-020 At 24 hour intervals, called Analysed Product Processing Windows (APPW) defined in Table 2.3.1, global coverage L2P<br />

L4A.0020<br />

data products are used in an analysis procedure.<br />

82 WP4-021 Two L4 data products are specified by the GDS: an estimate of SSTfnd at a grid resolution of 1/12° having global coverage<br />

NASR<br />

and an Ultra High Resolution SSTfnd product at a grid resolution of 1/48° having regional coverage.<br />

82 WP4-022 Each data product will include an estimate of the diurnal variability based on a parameterisation that can be driven using<br />

ancillary solar radiation and wind speed data<br />

LA4.0040<br />

L4A.0055<br />

82 WP4-023 and will be formatted to a common GDS L4 data format (described in Appendix A1) format<br />

82 WP4-024 will be made available to the user community in real time. DIS.0020<br />

82 WP4-025 The processor takes L2P data files as input together with a L4 configuration file containing threshold and data processor<br />

configuration settings.<br />

L4A.0030<br />

L4A.0035<br />

82 WP4-026 For each APPW L2P data are first collated for each APPW and grid cell. L4A.0060<br />

82 WP4-027 L4SSTfnd is calculated using analysis procedures for each output grid cell (WP-ID4.2 in Figure 4.1). L4A.0080<br />

82 WP4-028 An estimate of the diurnal variability within a 24 hour period is then derived for each grid cell (WP-ID4.3 in Figure 4.1). L4A.0055<br />

82 WP4-029 Each L4 data product is then formatted as a L4 data file (WP-ID4.4 in Figure 4.1) NASR<br />

82 WP4-030 a separate MMR_FR is prepared and submitted to the MMR (WP-ID4.5 in Figure 4.1). NASR<br />

82 WP4-031 HR-DDS granules are extracted from each L4 data product (WP-ID4.6 in Figure 4.1) NASR<br />

© 2004 SOC Page 166 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

82 WP4-032 a HR-DDS MMR_FR prepared and submitted to the HR-DDS MMR system (WP-ID4.7). NASR<br />

82 WP4-033 There are several schemes that have been developed by operational agencies and research establishments (e.g.,<br />

Reynolds and Smith) that could be used but in each case, the specific method must be adapted to the high resolution<br />

L4A.0080<br />

L4A.0095<br />

space scales that the GHRSST-PP is attempting to resolve<br />

82 WP4-034 it is expected that L4 data products will evolve as RDAC gain more experience with high resolution SST measurements. L4A.0090<br />

WP-ID4.1 Collation of Available L2P <strong>Data</strong><br />

83 WP4-035 all L2P data eligible for analysis must first be collated. L4A.0030<br />

83 WP4-036 L2P data must then be mapped into the analysis grid. L4A.0050<br />

84 WP4-037 L2P data must first be re-sampled onto the L4FND standard output grid which is defined in Appendix A1.1. NASR<br />

84 WP4-038 A number of input sensors (infrared imagers) have finer spatial resolution than the output grid specification, as shown in<br />

L4A.0060<br />

Figure 4.1.1. For these cases the rule will be to average the values of all pixels which overlap the product cell entirely and<br />

which have a L2P confidence record Proximity_Confidence value equal to the highest encountered within the cell, to<br />

produce a single value<br />

84 Rule 4.1a WP4-039 In the case of a smaller L2P input pixel than the L2FND grid cell size, L4FND data product cell values are derived from an<br />

L4A.0060<br />

average of the L2P pixel which completely overlap the product cell and which have a L2P confidence record<br />

Proximity_Confidence value equal to the highest encountered within the cell, to produce a single value.<br />

84 Rule 4.1b WP4-040 For input pixels that straddle the boundary between output grid cells, a weighting function may be applied to the input<br />

L4A.0060<br />

values according to the degree of coverage of the output grid cell and according to the SSES.<br />

84 Rule 4.1c WP4-041 The SSES value for a re-gridded data set will take the value associated with the Proximity_Confidence value assigned by<br />

L4A.0060<br />

Rule 4.1a.<br />

84 WP4-042 When pixels in the input data stream are much larger than those of the SSTFND output grid (e.g., microwave instruments),<br />

L4A.0060<br />

as illustrated in Figure 4.1.2, the output grid cell takes the value and confidence data record of the input pixel in which it<br />

© 2004 SOC Page 167 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

lies. Should it straddle more than one input pixel, it will take a weighted average of all those input pixels with which it<br />

overlaps and which have the highest L2P confidence record Proximity_Confidence value amongst the overlapping pixels.<br />

Weights will be assigned based on the coverage an input pixel has over the output grid.<br />

85 Rule 4.1d WP4-043 In the case of a larger pixel than the L4FND grid cell size, the output grid cell takes the value of the L2P pixel in which it<br />

L4A.0060<br />

lies.<br />

85 Rule 4.1e WP4-044 In the case where the output grid cell is covered by more than one L2P pixel, it will take the weighted average of all those<br />

L4A.0060<br />

L2P pixels with which it overlaps and which have the highest L2P confidence record Proximity_Confidence value<br />

amongst the overlapping pixels. Weights will be assigned based on the coverage an input pixel has over the output grid.<br />

85 WP4-045 There are other optimal methods for re-gridding data that may be investigated by the GHRSST-PP during the<br />

implementation of the GDS-v1.0 and may provide improvements over the simplistic method described here.<br />

85 WP4-046 in most optimal analysis systems, the analysis itself will optimise the re-gridding of data although re-gridding prior to<br />

L4A.0090<br />

L4A.0095<br />

L4A.0080<br />

interpolation provides a more straightforward processing system<br />

WP-ID4.2 Generate L4 SSTfnd data products (L4SSTfnd)<br />

85 WP4-047 Bias correction of all input data to the analysis procedure is critical to obtaining a valid output (see for example Reynolds et<br />

L4A.0080<br />

al, 2002).<br />

85 WP4-048 The GDS will use the L2P SSES derived bias as a measure of the overall uncertainty associated with each input data<br />

L4A.0080<br />

stream.<br />

85 WP4-049 bias due to diurnal stratification and cool skin effects must also be accounted for using additional data. L4A.0054<br />

85 WP4-050 Some satellite sensors provide a direct estimate of the SSTskin (AATSR) that must be adjusted to the SSTfnd. These data<br />

L4A.0054<br />

can then be used in an OI procedure to derive complete global fields.<br />

WP4-051 it is expected that the bias correction strategy will evolve and an upgrade of this component of the GDS is expected. L4A.0090<br />

86 WP4-052 L4FNDUHR data products will be generated for regions shown in Table 4.2.1. Other areas are expected in the future (e.g., L4A.0010<br />

© 2004 SOC Page 168 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

86 Table<br />

4.2.1<br />

European coastal seas) and will be described in Table 4.2.1 when operations begin.<br />

WP4-053 Name : Mediterranean, Regional project : <strong>Medspiration</strong> (EU-RDAC), Area : 30°N to 46°N and from 6°W to 36.5°E at 2<br />

km spatial resolution<br />

SYS.0081<br />

86 WP4-054 The GDS analysis scheme used to generate L4FND data products must: NASR<br />

86 WP4-055 • Provide a daily global coverage combined SSTfnd product that builds on the synergy between complementary SST data<br />

NASR<br />

streams having a grid cell size of 1/12° latitude x longitude (GDS-v1.0),<br />

86 WP4-056 • Provide error estimates for each analysed grid cell, L4A.0051<br />

86 WP4-057 • Account for differences in spatial and temporal sampling characteristics of each data stream L4A.0052<br />

86 WP4-058 • Account for gaps in coverage due to the presence of cloud, rain or lack of input L2P data L4A.0053<br />

86 WP4-059 • Account for SST diurnal variability both in space and time L4A.0054<br />

86 WP4-060 • Provide a measure of diurnal variability within the data product time domain to accompany the SSTfnd estimate that can<br />

also provide a best estimate of the skin temperature of the ocean (SSTskin) for each grid cell<br />

86 WP4-061 the analysis scheme described by Reynolds et al. (2002) may be modified to cater for high resolution spatial scales of<br />

L4A.0040<br />

L4A.0055<br />

L4A.0095<br />

SST.<br />

86 WP4-062 Appropriate spatial de-correlation length scales must be derived by the GHRSST-PP RDAC projects and close interaction<br />

L4A.0095<br />

with the GHRSST-PP Science Team.<br />

86 WP4-063 All parameters associated with the operation of the L4SSTfnd analysis scheme should be stored in a L4SSTfnd<br />

L4A.0035<br />

configuration file allowing for their easy editing and adjustment<br />

86 WP4-064 it is expected that alternative methods to the scheme proposed by Reynolds et al (2002) will be explored and compared<br />

L4A.0090<br />

during the GHRSST-PP within the framework of the GODAE inter-comparison project.<br />

WP-ID4.3 Generate L4 diurnal variation data fields<br />

86 WP4-065 Diurnal variation data fields form part of each L4 data product and are designed to enable a user to estimate the magnitude L4A.0055<br />

© 2004 SOC Page 169 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

and phase of diurnal variability at temporal intervals of 6 hours together with the SSTfnd.<br />

86 WP4-066 The diurnal variation will be referenced to the SSTskin. L4A.0055<br />

86 WP4-067 as the GHRSST-PP matures, a more comprehensive strategy will be adopted to generate estimates of SSTskin from<br />

L4a.0040<br />

multiple sensors<br />

87 WP4-068 a parameterisation of diurnal variability should be made available to the user community that allows the phase and<br />

L4A.0100<br />

magnitude of diurnal variation to be estimated at hourly intervals for each SST grid cell.<br />

87 WP4-069 The Diurnal signal will be referenced to the SSTsub-skin temperature so that warm layer effects are included L4A.0040<br />

87 WP4-070 an additional parameterisation for cool skin effects will be required to obtain the SSTskin. L4A.0040<br />

87 WP4-071 The initial specification is to use the diurnal variation scheme proposed by Alice Stuart-Menteth L4A.0040<br />

L4A.0070<br />

L4A.0055<br />

87 WP4-072 As an initial specification the bias correction strategy described in Donlon et al (2002) may be used to bias adjust SSTsubskin<br />

observations to SSTskin<br />

87 WP4-073 The appropriate coefficients for estimating diurnal variation are incorporated as part of the GHRSST-PP SSTfnd data<br />

format<br />

product using the Skin_parameterisation L4 SST grid cell ancillary data record described in Appendix A1.5, Table A1.5.2.<br />

87 WP4-074 In addition the L4 SST grid cell ancillary data record Skin_parameterisation_source should be updated according to<br />

format<br />

Table 4.3.1.<br />

87 Table<br />

WP4-075 Code values to be used to update the L4FND ancillary data record variable Skin_parameterisation_source (Table A1.5.2) format<br />

4.3.1<br />

87 Table<br />

WP4-076 0 : No parameterisation specified for this grid cell format<br />

4.3.1<br />

87 Table<br />

WP4-077 1 : Stuart-Menteth parameterisation format<br />

4.3.1<br />

© 2004 SOC Page 170 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

92 WP4-078 Figure 4.3.4 provides a schematic diagram showing one implementation approach for the Stuart-Menteth scheme within<br />

SR entry<br />

L4A.0070<br />

the GHRSST-PP L4 analysis system.<br />

93 WP4-079 It is clear that further work is required within the GHRSST-PP to develop and validate the Stuart-Menteth scheme<br />

L4A.0070<br />

discussed here.<br />

WP-ID4.4 Format L4 data product<br />

93 WP4-080 Each L4 data set will be formatted and archived in netCDF format following the data format described in Appendix A1.2,<br />

format<br />

A1.3.<br />

94 Rule<br />

WP4-081<br />

The L4FND grid cell ancillary data field Normalised_analysis_error should be updated with the normalised error of the<br />

format<br />

4.4.1<br />

analysis procedure.<br />

94 Rule<br />

WP4-082 The L4FND grid cell ancillary data field Bias should be updated with the analysis procedure bias error estimate. format<br />

4.4.2<br />

94 Rule<br />

WP4-083<br />

The L4FND grid cell ancillary data field Skin_parameterisation should be updated with a suite of coefficients that allow a<br />

format<br />

4.4.3<br />

user to compute an estimate of the SSTskin. The skin parameterisation scheme to be used is specified in the bitfield<br />

variable Skin_parameterisation_source according to Table 4.3.1.<br />

94 Rule<br />

WP4-084 If the grid cell is located over land, the L4FND ancillary data record bitfield Land_mask should be set to 1. format<br />

4.4.4<br />

94 Rule<br />

WP4-085<br />

The L4FND grid cell ancillary data field FractionalSeaIce variable should be set to the most appropriate value derived<br />

format<br />

4.4.5<br />

from the L2P input data streams appropriate to this output grid cell. A value of O indicates open ocean and 100 indicates<br />

100% sea ice coverage. In addition, the L4FND grid cell ancillary data field FractionalSeaIce_source should be set<br />

according to the source of data used to define the value of FractionalSeaIce as defined in Table 2.1.1.3.<br />

WP-ID4.5 Generate and deliver L4 MMR_FR metadata records<br />

94 WP4-086 a MMR_FR metadata record must be prepared and successfully submitted to the GDS MMR by the responsible Out of scope<br />

© 2004 SOC Page 171 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

processing centre for all L4 data products.<br />

WP-ID4.5.1 Format L4 MMR_FR metadata record<br />

94 WP4-087 A MMR_FR metadata record should be prepared for each L4 data file according to the specification provided in Appendix<br />

Out of scope<br />

A6.2.<br />

WP-ID4.5.2 Register L4 MMR_FR metadata record with MMR system<br />

94 WP4-088 L4 MMR_FR metadata records should be delivered to the MMR system as soon as possible after the L4 data set has been<br />

Out of scope<br />

produced following the procedure described in Appendix A6.4.<br />

94 WP4-089 WP-ID4.7 Extraction of L4 match up database file records (MDB_FR) NASR<br />

94 WP4-090 L4* MDB_FR should be extracted according to the procedures described in WP-ID3. Out of scope<br />

159 Appendix A5. GHRSST-PP High Resolution Diagnostic <strong>Data</strong> Set (HR-DDS) data granule data product format<br />

specification<br />

159 AP5-001 The GHRSST-PP High-resolution Diagnostic <strong>Data</strong> Set (HR-DDS) system provides a distributed data resource and a<br />

NASR<br />

framework to analyse and inter-compare L2P and L4 data products in near real time, together with other data products<br />

including NWP analyses, operational ocean and atmospheric model outputs and in situ observations<br />

159 AP5-002 In order to reduce the overhead of data storage and transport, the HR-DDS consists of about150 globally distributed 2º x<br />

NASR<br />

2º latitude x longitude sites<br />

159 AP5-003 Figure A5.1 shows a map of primary[1] HR-DDS sites (v2.3) for which all L2P and L4* data streams[2] will be extracted<br />

and archived as data granules. Each HR-DDS site is strategically positioned in order to address a particular issue; for<br />

implies<br />

CFI.DDS.0020<br />

example, a particularly dynamic or “quiet” oceanographic or atmospheric region, location of additional in situ infrastructure,<br />

areas known to be influenced by atmospheric aerosol or persistent cloud cover and areas already having a significant<br />

scientific interest. Each site shown in Figure A5.1 is defined in Section A5.3.<br />

159 AP5-004 The HR-DDS system provides a real time data resource the main learning tool for GHRSST-PP project scientists to NASR<br />

© 2004 SOC Page 172 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

develop and refine the GDS, to investigate differences between complementary data streams, to investigate regional and<br />

time variant bias statistics, to monitor the quality of input and output data streams and as a core component of the<br />

GHRSST-PP reanalysis (RAN) project<br />

159 AP5-005 The HR-DDS constitutes the virtual laboratory of the GHRSST-PP allowing, for example, a full implementation of<br />

NASR<br />

Integrated Global Observing Strategy (IGOS) measurement principles (see http://www.igospartners.org/).<br />

159 AP5-006 This Appendix provides a summary overview of the GHRSST-PP HR-DDS system. A complete Scientific and technical<br />

NASR<br />

description of the HR-DDS system can be found in GHRSST/13 available at http://www.ghgrsst-pp.org<br />

A5.1 Summary of the HR-DDS system configuration and operation<br />

159 AP5-007 HR-DDS granules are produced by RDAC and GDAC centres in real time as netCDF data files according to the<br />

DDS.0120<br />

specifications provided in Appendix A1<br />

159 AP5-008 HR-DDS granules are archived locally together with a corresponding HR-DDS metadata record DDS.0110<br />

159 AP5-009 HR-DDS metadata records are produced according to the specifications described in Appendix A6 and are identical to<br />

format<br />

other GDS metadata records<br />

159 AP5-010 HR-DDS archive centres are referred to as HR-DDS nodes and in general, HR-DDS nodes should be part of RDAC and<br />

DDS.0110<br />

GDAC<br />

160 AP5-011 Figure A5.2 provides a schematic diagram of the GHRSST-PP HR-DDS data system which is based on a number of<br />

NASR<br />

distributed HR-DDS nodes that are interconnected using the Distributed Oceanographic <strong>Data</strong> <strong>System</strong> (DODS, see<br />

http://www.dods.org).<br />

160 AP5-012 The DODS architecture uses a client/server model and the http protocol to provide a framework that simplifies . . . . . . NASR<br />

160 AP5-013 In addition to local data access via DODS/OPeNDAP servers at each HR-DDS node, HR-DDS nodal data archives are<br />

NASR<br />

virtually combined as a single data archive by a HR-DDS virtual server. The HR-DDS virtual server uses GHRSST-PP<br />

Master Metadata Repository (MMR) HR-DDS metadata records to catalogue HR-DDS data granules archived at all<br />

© 2004 SOC Page 173 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

distributed HR-DDS nodes. Using a HR-DDS interface, users at the HR-DDS virtual server may access all HR-DDS<br />

granules (maintained at distributed locations, served by local DODS servers) using DODS aware client applications as if<br />

they were a single data archive residing on a local computer.<br />

160 AP5-014 In addition to DODS access, HR-DDS granules may also be accessed using secure ftp (sftp) transactions or through a Live<br />

NASR<br />

Access Server (LAS, http://ferret.pmel.noaa.gov/Ferret/LAS/). LAS is a highly configurable Web server designed to<br />

provide flexible access to geo-referenced scientific data and can present distributed data sets as a unified virtual data base<br />

through the use of DODS networking<br />

160 AP5-015 Using a web browser interface LAS enables a user to visualize data with on-the-fly graphics; request custom subsets of<br />

NASR<br />

variables in a choice of file formats ;access background reference material about the data (metadata) ;compare<br />

(difference) variables from distributed locations<br />

161 AP5-016 LAS enables the data provider to unify access to multiple types of data in a single interface ;create thematic data servers<br />

NASR<br />

from distributed data sources ; offer derived products on the fly ; remedy metadata inadequacies (poorly self-describing<br />

data) ; offer unique products (e.g. visualization styles specialized for the data)<br />

161 AP5-017 LAS will be used extensively in the GODAE project and will allow data to be extracted from many operational ocean model<br />

NASR<br />

systems for inclusion and comparison with HR-DDS SST data. In addition, the LAS system is in active development to<br />

provide advanced features such as temporal aggregation of data and advanced display options in collaboration with the<br />

GODAE <strong>Data</strong> Sharing Pilot Project.<br />

161 AP5-018 Each HR-DDS node is thus responsible for NASR<br />

161 AP5-019 (a) Installing and maintaining a data archive of HR-DDS granules and associated MMR DSR and MMR-FS metadata<br />

DDS.0110<br />

records (see Appendix A6)<br />

161 AP5-020 (b) Operating a DODS server to serve HR-DDS data DDS.0140<br />

EXT.DDS.0030<br />

© 2004 SOC Page 174 of 193


<strong>Medspiration</strong> MED-SOC-RS-001_1<br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong> Issue F<br />

Page Table, fig. Reference ID Text from reference documents<br />

SR entry<br />

161 AP5-021 (c) Operating an sftp server for HR-DDS granule access EXT.DDS.0040<br />

161 AP5-022 (d) Operating other optional data interface software (e.g., LAS) tbd<br />

A5.2 HR-DDS data granule file format<br />

161 AP5-023 HR-DDS granules are formatted as netCDF data files according to the format description laid out for L2P data file in<br />

DDS.0120<br />

Appendix A1<br />

161 AP5-024 The main difference between L2P data and HR-DDS data files is that HR-DDS data have been re-sampled using a nearest<br />

neighbour scheme to a uniform resolution grid specified in Table A1.1.3<br />

DDS.0030<br />

DDS.0090<br />

161 AP5-025 All L2P confidence data are re-sampled in the same manner to provide a multiple image netCF file. DDS.0030<br />

A5.3 HR-DDS metadata record format and registration<br />

161 AP5-026 HR-DDS granule metadata records are identical in content and format to GDS metadata records. A full description of<br />

format<br />

MMR-DSR and MMR_FR is given in Appendix A6.<br />

161 AP5-027 A MMR DSR will be registered at the GHRSST-PP MMR for each HR-DDS site that is described in Appendix A5.4. DDS.0060<br />

EXT.DDS.0050<br />

161 AP5-028 A MMR_FR will be generated for each HR-DDS granule and registered at the MMR following the procedure described in<br />

Appendix A6.4.<br />

DDS.0070<br />

EXT.DDS.0060<br />

A5.4 Location of HR-DDS sites<br />

162-<br />

AP5-029<br />

Tables A5.4.1, A5.4.2, A5.4.3 and A5.4.4 describe the locations and formal names of each HR-DDS (v2) site for open<br />

NASR<br />

167<br />

ocean, moored buoys, SIMBIOS and other sites respectively<br />

© 2004 SOC Page 175 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Table A2. Trace between <strong>System</strong> <strong>Requirements</strong> and the relevant sections of the user<br />

requirement documents to which relate.<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

3.1.1 SYS.0010 sow-026 urd-084 GDS-016 GDS-012 GDS-013<br />

3.1.1 SYS.0015 sow-059 sow-060 sow-064 sow-065 urd-83 urd-131<br />

GDS-018 GDS-042<br />

3.1.1 SYS.0020 sow-007 urd-106 urd-133 GDS-015<br />

3.1.1 SYS.0021 urd-133 GDS-017<br />

3.1.2 SYS.0030 sow-027 sow-028 sow-029 sow-030 sow-031 sow-032<br />

sow-033 sow-034 sow-035 urd-107 sow-062<br />

3.1.2 SYS.0040 (urd-109)<br />

3.1.3 SYS.0050 sow-040 urd-083 urd-084 urd-100 urd-111 GDS-19<br />

3.1.3 SYS.0051 urd-085<br />

3.1.3 SYS.0052 sow-042<br />

3.1.3 SYS.0055 sow-048<br />

3.1.3 SYS.0060 sow-057 urd-092 urd-115 ap4-002 GDS-013<br />

3.1.3 SYS.0061 sow-058 urd-115<br />

3.1.3 SYS.0070 sow-053 urd-091 urd-104 urd-114 GDS-013<br />

3.1.3 SYS.0071 sow-054 sow-055 soe-056 urd-114<br />

3.1.3 SYS.0080 sow-041 sow-050 urd-093 urd-100 urd-113 GDS-20<br />

WP4-001 GDS-022<br />

3.1.3 SYS.0081 urd-096 WP4-053 GDS-022<br />

3.1.3 SYS.0082 sow-052 sow-68 GDS-022<br />

3.1.3 SYS.0083 sow-051<br />

3.1.4 SYS.0090 urd-112 GDS-013<br />

© 2004 SOC Page 176 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

3.1.4 SYS.0100 urd-111 GDS-075 GDS-013<br />

3.1.4 SYS.0110 sow-046 sow-047 urd-132<br />

3.1.4 SYS.0120<br />

3.1.5 SYS.2010 GDS-056 GDS-057<br />

3.1.5 SYS.2020 sow-047<br />

3.1.5 SYS.2030 GDS-088<br />

3.1.5 SYS.2040 urd-112<br />

3.1.5 SYS.2050 GDS-089 GDS-064 sow-008<br />

3.1.5 SYS.2060 urd-129 GDS-050 ap4-002<br />

3.1.5 SYS.2070<br />

3.1.5 SYS.2080 GDS-051 GDS-091<br />

3.1.5 SYS.2081 GDS-095<br />

3.1.5 SYS.2082 GDS-095<br />

3.1.5 SYS.2090<br />

3.2. L2P.0010 sow-043 GDS-047 WP1-032 GDS-056 WP1-055 WP1-057<br />

3.2. L2P.0020 sow-043 GDS-047 WP2-010 GDS-057<br />

3.2. L2P.0030 GDS-048 WP2-048<br />

3.2. L2P.0040 sow-043 GDS-043 GDS-048 WP2-050 WP2-041 GDS-070<br />

3.2. L2P.0050 sow-043 GDS-043 GDS-049 WP2-054 WP2-041 GDS-068<br />

GDS-073<br />

L2P.0055<br />

WP2-005<br />

3.2. L2P.0060 WP2-236 WP2-055 WP2-041 GDS.060<br />

3.2. L2P.0070 sow-043 GDS-063<br />

© 2004 SOC Page 177 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

3.2. L2P.0080 WP2-042<br />

3.2. L2P.0090 WP2-014 WP2-043<br />

3.2.6 L2P.6010 WP2-175<br />

3.2.6 L2P.6020 WP2-176 WP2-177<br />

3.2.7 L2P.7000 WP2-231 WP-234<br />

3.2.7 L2P.7010<br />

3.2.7 L2P.7020<br />

3.2.7 L2P.7030<br />

3.2.7 L2P.7040<br />

3.2.2 ING.0010 WP1-024<br />

3.2.2 ING.0020 WP-007<br />

3.2.2 ING.0030 sow-048 WP1-008 WP1-017 WP1-025<br />

3.2.2 ING.0040 WP1-028 WP1-066<br />

3.2.2 ING.0050 WP1-044<br />

3.2.2 ING.0060 WP1-026<br />

3.2.2 ING.0070 WP2-060 WP2-075<br />

3.2.2 ING.0080 WP2-076<br />

3.2.2 ING.0090 WP2-077<br />

3.2.2 ING.0100 WP1-009 WP1-015 WP1-059<br />

3.2.2 ING.0110 WP1-056<br />

3.2.2 ING.0120 WP1-062<br />

3.2.2 ING.0130 sow-007 urd-085 urd-096<br />

© 2004 SOC Page 178 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

3.2.1 ING.1010<br />

3.2.1 ING.1020<br />

3.2.3 QLC.0010 WP1-009 WP1-016 WP2-002 WP2-003<br />

3.2.3 QLC.0020 WP2-011 WP2-065<br />

3.2.3 QLC.0030<br />

3.2.3 QLC.0035 WP2-016 WP2-096<br />

3.2.3 QLC.0040 WP2-071 etc.<br />

3.2.3 QLC.0045 WP2-145 WP1-058<br />

3.2.3 QLC.0050 WP2-149<br />

3.2.3 QLC.0060 WP2-153<br />

3.2.3 QLC.0065 WP2-197<br />

3.2.3 QLC.0070 WP2-071 WP2-059 WP2-187<br />

3.2.3 QLC.0080 WP2-072 WP2-074 WP2-187<br />

3.2.3 QLC.0090 WP2-102 WP2-099<br />

3.2.3 QLC.0100 WP2-118<br />

3.2.3 QLC.0110 WP2-131<br />

QLC.0120 WP2-119 WP2-132 WP2-103<br />

3.2.4 MRG.0010 WP2.-004<br />

3.2.4 MRG.0020 WP2-058<br />

3.2.4 MRG.0030 WP2-061<br />

3.2.4 MRG.0040 WP2-062<br />

3.2.4 MRG.0050 WP2-063<br />

© 2004 SOC Page 179 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

3.2.4 MRG.0060 WP2-064<br />

3.2.4 MRG.0070 WP2-080<br />

3.2.4 MRG.0080 WP2-081<br />

3.2.4 MRG.0090 WP2-082<br />

3.2.4 MRG.0110 WP2-083 WP2-085<br />

3.2.4 MRG.0120 WP2-084<br />

3.2.4 MRG.0130 WP2-093 WP2-090 WP2-091<br />

3.2.4 MRG.0150 WP2-095<br />

3.2.4 MRG.0160 WP2-096<br />

3.2.4 MRG.0170 WP2-097<br />

3.2.4 MRG.0180 WP2-211<br />

3.2.4 MRG.0190 WP2-212<br />

3.2.4 MRG.0200 WP2-213 WP2-215<br />

3.2.4 MRG.0210 WP2-214<br />

3.2.4 MRG.0220 WP2-122<br />

3.2.4 MRG.0230 WP2.123<br />

3.2.4 MRG.0240 WP2-124 WP2-126<br />

3.2.4 MRG.0250 WP2-125<br />

3.2.4 MRG.0260<br />

3.2.5 CFD.0010 GDS-070<br />

3.2.5 CFD.0020 WP2-176 WP2-228 WP2-143<br />

3.2.5 CFD.0030 WP2-012<br />

3.2.5 CFD.0050 WP2-139 WP2-133 WP2-179<br />

© 2004 SOC Page 180 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

3.2.5.1 CFD.0100 WP2-142<br />

3.2.5.1 CFD.0110 WP2-194<br />

3.2.5.1 CFD.0120 WP2-195<br />

3.2.5.1 CFD.0140 WP2-201 WP-199<br />

3.2.5.1 CFD.0150 WP2-202 WP-200<br />

3.2.5.1 CFD.0160 WP2-204<br />

3.2.5.1 CFD.0170 WP2-206<br />

3.2.5.1 CFD.0180 WP2-207<br />

3.2.5.1 CFD.0190 WP2-211 WP2-220<br />

3.2.5.1 CFD.0200 WP2-209 WP2-210<br />

3.2.5.1 CFD.0210 WP-208 WP2-216<br />

3.2.5.1 CFD.0220 WP2-220 WP2-181 WP2-182<br />

3.2.5.2 CFD.0300 WP2-158 WP2-154 WP2-156<br />

3.2.5.2 CFD.0310 WP2-159 WP2-156<br />

3.2.5.2 CFD.0320 WP2-160 WP2-161 WP2-162 WP2-163 WP2-167 WP2-134<br />

WP2-137<br />

3.2.5.2 CFD.0330 WP2-167<br />

3.2.5.2 CFD.0340 WP2-166<br />

3.2.5.2 CFD.0350 WP2-168<br />

3.2.5.2 CFD.0360 WP2-169<br />

3.2.5.2 CFD.0370 WP2-170<br />

3.3.1.1 ARC.0010 sow-026 GDS-059<br />

3.3.1.1 ARC.0020<br />

© 2004 SOC Page 181 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

ARC.0030<br />

ARC.0040<br />

sow-026<br />

3.3.1.1 ARC.0110<br />

3.3.1.1 ARC.0120<br />

3.3.1.2 ARC.0210<br />

3.3.1.3 ARC.0310<br />

3.3.1.3 ARC.0320<br />

3.3.1.3 ARC.0330<br />

3.3.1.4 ARC.0410<br />

3.3.2 L4A.0010 sow-050 WP4-052<br />

L4A.0015<br />

3.3.2 L4A.0020 GDS-045 WP4-005 WP4-020<br />

L4A.0025<br />

3.3.2 L4A.0030 WP4-035 WP4-006 WP4-007 WP4-025<br />

L4A.0035 WP4-025 WP4-063<br />

3.3.2 L4A.0040 WP4-022 WP4-060 WP4-069 WP4-070 WP4-071<br />

3.3.2 L4A.0050 GDS-035 WP4-036<br />

3.3.2 L4A.0051 GDS-081 WP4-009<br />

3.3.2 L4A.0052 GDS-082<br />

3.3.2 L4A.0053 GDS-083<br />

3.3.2 L4A.0054 GDS-084 WP4-019 WP4-049 WP4-050<br />

3.3.2 L4A.0055 GDS-085 GDS-086 WP4-022 WP4-028 WP4-065<br />

3.3.2 L4A.0060 WP4-026 WP4-038 to WP4-044<br />

© 2004 SOC Page 182 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

3.3.2 L4A.0070 WP4-071 WP4-078 WP4-079<br />

3.3.2 L4A.0080 WP4-027 WP4-033 WP4-046<br />

3.3.2 L4A.0090 WP4-034 WP4-045 WP4-051 WP4-064<br />

L4A.0095 WP4-033 WP4-045 WP4-061 WP4-062<br />

3.3.2 L4A.0100 WP4-009<br />

3.3.3 MDB.0010 WP3-006 WP3-019<br />

3.3.3 MDB.0020 WP3-010 WP3-024 WP3-038<br />

3.3.3 MDB.0030 WP3-033 WP3-034 WP3-060 WP3.059<br />

3.3.3 MDB.0040 WP3-009 WP3-011<br />

3.3.3 MDB.0050 WP3-026<br />

3.3.3.1 MDB0110 WP3-035 WP3-036 sow-037<br />

3.3.3.1 MDB.0120 WP3-037 WP3-007 WP3-042<br />

3.3.3.1 MDB.0130 WP3-035 WP3-043 WP3-041<br />

3.3.3.2 MDB.0210 WP3-054 WP3-051 WP3-028<br />

3.3.3.2 MDB.0220 WP3.047<br />

3.3.3.2 MDB.0230 WP3-027 WP3-048 WP3-044 WP3-045<br />

3.3.3.2 MDB.0240 WP3-049 WP3-046 WP3-052 WP3-053<br />

3.3.3.2 MDB.0250 WP3-029 WP3-054<br />

3.3.3.2 MDB.0260 WP3-030<br />

3.3.3.2 MDB.0270 WP3-039 WP3-040<br />

3.3.3.3 MDB.0310 WP3-058<br />

3.3.3.3 MDB.0320 WP3-057 WP-055 WP3-056<br />

3.3.3.3 MDB.0330 WP3-061<br />

© 2004 SOC Page 183 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

3.3.4 DIS.0010 sow-040 sow-041<br />

3.3.4 DIS.0020 sow-060 WP4-024<br />

3.3.4 DIS.0030 sow-060<br />

3.3.4.1 DIS.0110 sow-061 urd-083<br />

3.3.4.1 DIS.0120 sow-061 urd-060 urd-083<br />

3.3.4.2 DIS.0210<br />

3.3.4.3 DIS.0310<br />

3.3.5.1 CTR.0110 WP1-026 sow-066 WP2-241<br />

3.3.5.1 CTR.0120 WP1-026 WP1-048 WP1-054<br />

3.3.5.1 CTR.0130 WP2-035 WP3-020 WP4-017<br />

3.3.5.1 CTR.0140 WP1-050<br />

3.3.5.1 CTR.0150 WP1-050<br />

3.3.5.1 CTR.0160<br />

3.3.5.1 CTR.0170 WP1-050<br />

3.3.5.1 CTR.0180 WP1-049<br />

3.3.5.1 CTR.0190<br />

3.3.5.2 CTR.0210<br />

3.3.5.2 CTR.0220<br />

3.3.5.3 CTR.0230<br />

3.3.5.3 CTR.0240<br />

3.3.5.3 CTR.0250<br />

3.3.5.4 CTR.0260<br />

© 2004 SOC Page 184 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

3.3.5.5 CTR.0310<br />

3.3.5.5 CTR.0320<br />

3.3.5.6 CTR.0410<br />

3.3.5.6 CTR.0420<br />

3.3.5.6 CTR.0430<br />

3.3.5.7 CTR.0510<br />

3.3.5.7 CTR.0520<br />

3.3.5.8 CTR.0530 GDS-054 WP1-045 WP1-046 WP2-006 WP2-235<br />

3.3.5.8 CTR.0540<br />

3.3.5.9 CTR.0550 WP2-007 WP2-232 WP2-233<br />

3.3.5.9 CTR.0560<br />

3.3.5.10 CTR.0570 URD-060<br />

3.3.5.10 CTR.0580 WP2-007 WP3-025 WP3-015 WP3-063<br />

3.4.2 DDS.0010 sow-053 sow-056<br />

3.4.2 DDS.0020 urd-065 sow-054<br />

3.4.2 DDS.0030 GDS-093 GDS-036 GDS-051 AP5-024 WP2-015<br />

3.4.2 DDS.0040 GDS-053 WP4-011<br />

3.4.2 DDS.0050<br />

3.4.2 DDS.0060 AP5-027<br />

3.4.2 DDS.0070 AP5-028<br />

3.4.2 DDS.0090 GDS-036 AP5-024<br />

3.4.2 DDS.0091 GDS-094<br />

3.4.3 DDS.0110 sow-055<br />

© 2004 SOC Page 185 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

3.4.3 DDS.0120 AP5-007 AP5-023<br />

3.4.4 DDS.0140 AP5-020<br />

3.4.5 DDS.0170<br />

3.4.5 DDS.0180<br />

3.4.5 DDS.0190 AP7-001<br />

3.4. DDS.1000<br />

3.4.1 DDS.1010 sow-028<br />

3.4.1 DDS.1030<br />

3.4.1 DDS.1040<br />

3.4.1 DDS.1050<br />

3.4.1 DDS.1060<br />

4.1.1 INT.L2P.0010<br />

4.1.1 INT.L2P.0020<br />

4.1.1 INT.L2P.0030<br />

4.1.1 INT.L2P.0040<br />

4.1.1 INT.L2P.0050<br />

4.1.1 INT.L2P.0060<br />

4.1.1 INT.L2P.0070<br />

4.1.1 INT.L2P.0080<br />

4.1.1 INT.L2P.0090<br />

4.1.1 INT.L2P.0100<br />

4.1.2 INT.ARC.0010<br />

© 2004 SOC Page 186 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

4.1.2 INT.ARC.0020<br />

4.1.2 INT.ARC.0030<br />

4.1.6 INT.ARC.0510<br />

4.1.6 INT.ARC.0520<br />

4.1.6 INT.ARC.0530<br />

4.1.6 INT.ARC.0540<br />

4.1.6 INT.ARC.0550<br />

4.1.3 INT.DIS.0010<br />

4.1.3 INT.DIS.0020<br />

4.1.3 INT.DIS.0030<br />

4.1.4 INT.MDB.0010<br />

4.1.5 INT.L4A.0010<br />

4.1.5 INT.L4A.0020<br />

4.1.5 INT.L4A.0030<br />

4.1.5 INT.L4A.0040<br />

4.1.7 INT.DDS.0010 GDS-051<br />

4.1.7 INT.DDS.0020 GDS-053<br />

4.1.7 INT.DDS.0030 WP4-011<br />

4.1.7 INT.DDS.0040<br />

4.1.7 INT.DDS.0050<br />

© 2004 SOC Page 187 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

4.1.7 INT.DDS.0060<br />

4.1.7 INT.DDS.0070<br />

4.1.7 INT.DDS.0080<br />

4.1.7 INT.DDS.0090<br />

4.1.7 INT.DDS.0100<br />

4.1.7 INT.DDS.0110<br />

4.1.7 INT.DDS.0120<br />

4.1.7 INT.DDS.0130<br />

4.1.8 INT.CTR.0020<br />

4.1.8 INT.CTR.0030<br />

4.1.8 INT.CTR.0040<br />

4.1.8 INT.CTR.0050<br />

4.1.8 INT.CTR.0060<br />

4.2.1 EXT.ING.0010 WP1-024<br />

4.2.1 EXT.ING.0020 WP1-024<br />

4.2.1 EXT.ING.0030 WP1-027<br />

4.2.1 EXT.ING.0040<br />

4.2.1 EXT.L2P.0010<br />

4.2.1 EXT.SYS.0010<br />

4.2.1 EXT.SYS.0020<br />

4.2.2 EXT.MDB.0010 sow-037<br />

© 2004 SOC Page 188 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

4.2.2 EXT.MDB.0020<br />

4.2.2 EXT.MDB.0030<br />

4.2.3 EXT.CTR.0210<br />

4.2.3 EXT.CTR.0220<br />

4.2.3 EXT.CTR.0230<br />

4.2.4 EXT.CTR.0310<br />

4.2.4 EXT.CTR.0320<br />

4.2.4 EXT.CTR.0330<br />

4.2.4 EXT.CTR.0340<br />

4.2.5 EXT.CTR.0410<br />

4.2.5 EXT.CTR.0420<br />

4.2.5 EXT.CTR.0430<br />

4.2.6 EXT.CTR.0510<br />

4.2.6 EXT.CTR.0520<br />

4.2.6 EXT.CTR.0530<br />

4.2.7 EXT.DIS.0010<br />

4.2.7 EXT.DIS.0020<br />

4.2.7 EXT.DIS.0030<br />

4.2.7 EXT.DIS.0040<br />

4.2.8 EXT.DDS.0010<br />

4.2.8 EXT.DDS.0020<br />

© 2004 SOC Page 189 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

4.2.8 EXT.DDS.0030 AP5-020<br />

4.2.8 EXT.DDS.0040 AP5-021<br />

4.2.8 EXT.DDS.0050 AP5-027<br />

4.2.8 EXT.DDS.0060 AP5-028<br />

5.1. PER.SYS.0010 GDS-049<br />

5.1. PER.SYS.0011<br />

5.1. PER.SYS.0015<br />

5.1. PER.SYS.0020<br />

5.1. PER.SYS.0030<br />

5.1. PER.SYS.0040<br />

5.1. PER.SYS.0050<br />

5.1. PER.SYS.0060<br />

5.1. PER.SYS.0070<br />

5.1. PER.SYS.0080 URD-099<br />

5.1. PER.DDS.0010<br />

5.1. PER.DDS.0020 WP4-017<br />

5.1. PER.DDS.0030<br />

5.1. PER.DDS.0050<br />

5.1. PER.DDS.0060<br />

5.1. PER.DDS.0070<br />

6.1. EXP.SYS.0010 WP2-129 WP2-230<br />

© 2004 SOC Page 190 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

6.1. EXP.SYS.0020 urd-105<br />

6.1. EXP.SYS.0030 urd-073 sow-041 urd-093<br />

6.2. EXP.L2P.0010 sow-070 sow-071<br />

6.2. EXP.L2P.0020<br />

6.2. EXP.L2P.0030 WP2-173<br />

6.2. EXP.L2P.0040 WP2-224<br />

6.2. EXP.L2P.0050 WP2-221 WP2-222<br />

6.2. EXP.L2P.0060<br />

6.2. EXP.L2P.0070 WP4-012<br />

6.3. EXP.DDS.0010 sow-067 sow-71<br />

6.3. EXP.DDS.0020 sow-067<br />

6.3. EXP.DDS.0030 sow-067 sow-70<br />

6.3. EXP.DDS.0040 sow-067<br />

6.3. EXP.DDS.0050<br />

6.3. EXP.DDS.0060<br />

7.. SEC.SYS.0010<br />

7.. SEC.DDS.0010<br />

8.. DES.SYS.0010 sow-070<br />

8.. DES.SYS.0020 sow-069<br />

8.. DES.SYS.0030 sow-066 WP1-051<br />

© 2004 SOC Page 191 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

8.. DES.SYS.0040 sow-067 GDS-017 WP4-003<br />

8.. DES.SYS.0050 sow-071<br />

8.. DES.SYS.0060 sow-70<br />

9.. REL.SYS.0010<br />

9.. REL.ARC.0010<br />

10.. TST.SYS.0020<br />

10.. TST.SYS.0030<br />

10.. TST.SYS.0040<br />

10.. TST.SYS.0050<br />

10.. TST.SYS.0060<br />

10.. TST.SYS.0070<br />

10.. TST.SYS.0080<br />

10.. TST.SYS.0090<br />

10.. TST.SYS.0100<br />

12.. CFI.SYS.0010 sow-080 GDS-023 WP2-170<br />

12.. CFI.SYS.0020 sow-081<br />

12.. CFI.SYS.0030 sow-082<br />

12.. CFI.SYS.0040 sow-083<br />

12.. CFI.L2P.0010 WP1-053<br />

12.. CFI.L2P.0020 sow-086<br />

12.. CFI.DDS.0010<br />

© 2004 SOC Page 192 of 193


<strong>Medspiration</strong><br />

<strong>System</strong> <strong>Requirements</strong> <strong>Document</strong><br />

MED-SOC-RS-001_1<br />

Issue F<br />

Section SR ID ref. References in the main user requirements matrix (Table A1)<br />

12.. CFI.DDS.0020<br />

12.. CFI.DDS.0030 WP5-003<br />

12.. CFI.DDS.0040<br />

© 2004 SOC Page 193 of 193

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!