Third IMO Greenhouse Gas Study 2014
GHG3%20Executive%20Summary%20and%20Report
GHG3%20Executive%20Summary%20and%20Report
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
64 <strong>Third</strong> <strong>IMO</strong> GHG <strong>Study</strong> <strong>2014</strong><br />
increased numbers of satellites providing greater spatial and temporal coverage, and increased<br />
experience in filtering and processing the raw data for use in modelling.<br />
• A quality advantage in this work is that our approach for the bottom-up activity-based inventory<br />
uses calculations for individual vessels. By maximizing vessel-specific activity characterization using<br />
AIS data sources, this work quantifies the variability among vessels within a type and size category.<br />
This eliminates the dominant uncertainties reported by the Second <strong>IMO</strong> GHG <strong>Study</strong> 2009 and most<br />
published inventories.<br />
• The AIS-informed bottom-up methodologies cannot directly distinguish between fuel type and voyage<br />
type, which requires additional analyses and some expert judgement. Our QA/QC on allocation<br />
of residual/distillate fuels (HFO/MDO) and international/domestic shipping provides transparent and<br />
reproducible methodologies, with the opportunity to adjust these if and when better information<br />
becomes available in the future.<br />
At the time that this report was written, there were too few data sets of on-board measurements of CO 2<br />
emissions for any statistically representative quality assurance investigation of the modelled CO 2 emission to<br />
be carried out. The closest that the quality assurance can therefore get to the end product of this study is the<br />
fuel consumption comparison (modelled estimate compared with operator data), carried out using noon report<br />
data. This is done for a sample of approximately 500 ships (approximately 1% of all vessels) representing over<br />
60,000 days of at-sea operation. This sample is described in detail in Annex 3. It should be noted that noon<br />
report data are not infallible; their reliability and the implications for the comparative analysis undertaken here<br />
are discussed in greater detail in Annex 3.<br />
To provide further assurance of the inputs and assumptions of the bottom-up method, specifically the activity<br />
estimate, the consortium also performed analysis with LRIT data (approximately 8,000 ships and 10% of the<br />
global fleet) and third-party literature study.<br />
Noon reports, LRIT data and the literature were used for the following components of quality assurance work:<br />
• The activity estimation quality was assured using:<br />
––<br />
spatial coverage analysis with information on the number of messages received in different geographical<br />
locations and contrasting the AIS coverage with coverage maps obtained from alternative sources (e.g.<br />
LRIT);<br />
––<br />
temporal coverage analysis to test whether the derived profiles of time spent in different modes of<br />
operation (e.g. in port, at sea) and at different speeds are representative;<br />
––<br />
comparison of the AIS-derived activity parameters speed and draught against noon report data;<br />
––<br />
description of coverage statistics for each year and each fleet (to evaluate AIS completeness and<br />
facilitate imputed algorithms to estimate CO 2 emissions from periods when observations are missing).<br />
• Fleet specifications and model assumption quality were assured using:<br />
––<br />
investigations into the robustness of the IHSF database;<br />
––<br />
comparative evaluation of prior work, independently produced and published by consortium<br />
members, including peer-reviewed reports and scientific articles;<br />
––<br />
consultation of third-party inventory and shipping literature (including the work of consortium partners)<br />
providing substantial fleet data.<br />
• Fuel consumption estimate quality was assured using:<br />
––<br />
comparison of calculated fuel consumption to operators’ data recorded in noon reports pooled from<br />
data independently collected by several consortium partners.<br />
It should be noted that noon report data are not infallible; their reliability and the implications for the<br />
comparative analysis undertaken here are discussed in greater detail in Annex 3, along with detailed QA/QC<br />
for the source data and other analyses.<br />
Spatial coverage of activity estimates QA/QC<br />
The AIS data coverage, in terms of both space and time, is not consistent year-on-year during the period<br />
studied (2007–2012). For the first three years (2007–2009), no satellite AIS data were available, only data from