Global EDGAR greenhouse gas emissions and national emissions reporting under the UN Climate Convention: availability, structure, definitions and role of uncertainties

JRC and PBL have compiled a comprehensive EDGAR v4 dataset for the ‘six’ greenhouse gases included in the Kyoto Protocol (CO2, CH4, N2O, HFCs, PFCs and SF6), which were constructed using consistently the 2006 IPCC methodology and combining activity data (international statistics) from publicly available sources and for the first time - to the extent possible - emission factors as recommended by the IPCC 2006 guidelines for GHG emission inventories. This dataset, that covers all countries, provides independent estimates for all anthropogenic sources from 1970 onwards that are consistent over time and comparable between countries. By using high resolution global grid maps per source category, we also compiled datasets with annual emissions on a 0.1x0.1 degree grid, as input for atmospheric models. Of the 220 UN nations only 43 industrialised countries (‘Annex I’) annually report their national GHG emissions in large detail from 1990 up to (presently) 2008, while most developing countries (‘non-Annex I’) for the UN Climate Convention (UNFCCC) and the Kyoto Protocol only report a summary table with emissions for one or more years (many only for 1994). All emissions in EDGAR 4 are detailed at country level using the same standard IPCC inventory source categories as used by industrialised countries in their report to the Climate Convention. Thus we provide full and up-to-date inventories per country, also for developing countries that go beyond the mostly very aggregated UNFCCC reports of the developing countries. Moreover, the time series back in time to 1970 provides for the UNFCCC trends a historic perspective. As part of our objective to contribute to more reliable inventories by providing a reference emissions database for emission scenarios, inventory comparisons and for atmospheric modellers, we strive to transparently document all data sources used and assumptions made where data was missing, in particular for assumptions made on the shares of technologies where relevant. We briefly present key elements and the level of detail of the EDGAR v4 datasets, including the results of an uncertainty assessment by main source and region and a comparison with UNFCCC data. Moreover, we will highlight the main differences with version 4.0, which was released in 2009. The different roles of uncertainty estimates in scientific inventories such as EDGAR and in official national GHG inventories will also be discussed, as well as some non-territorial, non-annual, non-actual emission aspects of the inventory that is pivotal information for using and interpreting these emissions inventories by the Earth System and Atmospheric Modelling communities. A special challenge is compiling updated global inventories for more recent years, e.g. through 2008, with limited capacity in a reasonably accurate way that still captures relevant emission trends, of which we also present the methods used and show key results.