EUMETSAT Radio Occultation Observation With Metop-A ... - COSMIC

COSMIC 2012, Boulder
October 2012
EUMETSAT Radio Occultation Observation With Metop­A
and Metop­B in Orbit
A. von Engeln, C. Marquardt, Y. Andres
(EUMETSAT)
presented by H. Gleisner (ROM SAF/DMI)

Overview
1. Metop­A: Operational Activities
2. Metop­B: Early Results vs. ECMWF, Metop­A, COSMIC
3. Conclusion/Next Steps
COSMIC 2012, Boulder
October 2012
Slide

COSMIC 2012, Boulder
October 2012
Metop­A: Operational Activities
Slide

Metop­A: Operational Activities (1)
Robust provision of occultations with a timeliness of
about 1.2 hour, using 2 ground stations (Svalbard,
McMurdo)
COSMIC 2012, Boulder
October 2012
Robust provision of about 700 GRAS­A
occultations/day with very few outages
Slide

Metop­A: Operational Activities (2)
GRAS­A and COSMIC co­location statistics by latitude band (showing bias, std dev):
• co­locations within ≤300km, ≤3h for 2012, yields about 116 occultations/day
• grey shaded area indicates GRAS operational geometrical optics limitation
• 1K ≈
1.5% change in bending angles
COSMIC 2012, Boulder
October 2012
Slide

COSMIC 2012, Boulder
October 2012
Metop­B
Slide

Metop­B: Launch and Orbit Setup
COSMIC 2012, Boulder
October 2012
Roscosmos
• Metop­B successfully launched on Sept. 17, 2012
from Baikonur in Kazakhstan
• Metop­B flies in same orbit as Metop­A, thus:
• covers the same local solar time (09:30am)
• shifted by ~50min (~180 Degree) in orbit
• GRAS switch on / first data on Sept. 24:
• initially processed on back­up node
• after internal review on ~1 week of data
moved to operational node and made
available to Cal/Val partners (DWD, ECMWF,
Met Office, Meteo France, ROM SAF) on Oct.
1
• after internal review / feedback from Cal/Val
partners on 1 month of data it is just about to
be made available to everybody (with preoperational
status, dissemination team is
setting up the services right now)
Slide

Metop­B: Internal Metop­A / ­B Validation (1)
GRAS­A (left) and ­B (right) profile statistics against ECMWF (each showing bias, std
dev):
• show both very similar deviations against ECMWF
• grey shaded area indicates GRAS operational geometrical optics limitation
COSMIC 2012, Boulder
October 2012
Slide

Metop­B: Internal Metop­A / ­B Validation (2)
GRAS­A and ­B co­location statistics by latitude band (showing bias, std dev):
• co­locations within ≤300km, ≤3h, yields about 100 occultations/day
• lower deviations than found against ECMWF (thus likely model biases visible in
ECMWF)
• grey shaded area indicates GRAS operational geometrical optics limitation
COSMIC 2012, Boulder
October 2012
Slide

Metop­B: Internal Metop­A / ­B Validation (3)
GRAS­A (left) and GRAS­B (right) 2 weeks drift statistics against ECMWF at 50km (top 2 plots
global, bottom 2 plots by lat separation, bias and std dev shown):
• show both very similar deviations against ECMWF
• minor separation between setting/rising caused by used sequential filter POD processing
• Metop­B went through collision avoidance maneuver on Oct. 23 (grey shaded)
COSMIC 2012, Boulder
October 2012
Slide

Metop­B: Internal Metop vs. COSMIC Validation (1)
GRAS­A, GRAS­B, COSMIC number of daily occultations:
• very similar numbers for Metop­A, Metop­B (early dip is caused by monitoring setup)
• total of Metop­A and Metop­B similar to COSMIC total (at least for this month)
COSMIC 2012, Boulder
October 2012
Slide

Metop­B: Internal Metop vs. COSMIC Validation (2)
GRAS­A, GRAS­B, COSMIC global statistics against ECMWF:
• show all very similar deviations against ECMWF
• grey shaded area indicates GRAS operational geometrical optics limitation
COSMIC 2012, Boulder
October 2012
Slide

Metop­B: Internal Metop vs. COSMIC Validation (3)
Global GRAS­A, GRAS­B, COSMIC co­location statistics (showing bias, std dev):
• co­locations within ≤300km, ≤3h
• show all very similar agreement, smaller deviations than against ECMWF
• grey shaded area indicates GRAS operational geometrical optics limitation
COSMIC 2012, Boulder
October 2012
Slide

Metop­B: External Metop­B Validation
ECMWF, DWD, Met Office, Meteo France provided early data validation:
• show all very similar agreement between Metop­A and Metop­B
• ECMWF already provides monitoring of Metop­B (not yet assimilated, parallel trials
ongoing) at www.ecmwf.int
COSMIC 2012, Boulder
October 2012
Slide

Metop­B: In the RO Global Observing System
COSMIC 2012, Boulder
October 2012
• Metop­B to take over
operational service
early 2013
• Metop­A will not be
fully operational then
(e.g. only Svalbard
downlink)
• plan to fly Metop­A
and ­B in in current
orbits until ­C launch
(if ­A healthy)
• will partly fill the
possible data gap
after COSMIC­1
Slide

COSMIC 2012, Boulder
October 2012
Conclusion/Next Steps
Slide

Conclusion
• Metop­A GRAS operational data:
• very robust, low timeliness data stream, expected to last beyond initial life time
• Metop­B:
• launched on Sept. 17 2012, doubles available EUMETSAT data
• will partly fill the coming gap in RO observations
• internal / external validation shows excellent agreement
• trial dissemination to Cal/Val partners started very early: Oct. 01
• pre­operational dissemination started within about 1 month after switch on: expected
to start at or week after the COSMIC workshop
COSMIC 2012, Boulder
October 2012
Slide: 17

Next Steps
• Metop­B:
• tracking parameter fine tuning planned for several weeks in Dec, Jan (thus no data
available during some of these period ­ we try to limit dissemination outages)
• fully operational status (depends on full satellite declared operational) early 2013
• Operational processor update:
• Geometric optics and improved orbit processing planned shortly after Metop­B
commissioning
• wave optics planned for mid­2013
COSMIC 2012, Boulder
October 2012
Slide: 18