EDGAR_v4.3: a global air pollutant emission inventory from 1970 to 2010
Tuesday, 16 December 2014: 11:35 AM
The Emission Database for Global Atmospheric Research (EDGAR) provides consistent gridded anthropogenic emissions of greenhouse gases, precursor gases and aerosols from 1970 to 2010. Since EDGAR’s first release in 1996 (EDGARv2), a continuous improvement and upgrade of the emission data resulted in a sequence of releases. Here we present EDGAR_v4.3 (2014), which features new information on emission factors and an extension to 2009-2010 data compared to EDGAR_v4.2. EDGAR_v4.2 was used in many inverse modeling studies in EU, US, Africa and Asia yielding regional refinement of emission factors and adjustments of technology penetration (e.g. coal mining, power plants) and proxy data for geospatial distribution (e.g. passenger car transport). We focus on SO2, NOx, CO, NMVOC, NH3, PM10, PM2.5, BC and OC emissions for the most recent year (2010), and compare them to two global inventories used in global modeling, as well as the regional inventories included in them. HTAP_v2 is a harmonized, global, gridded, emission database for 2010, developed for global and regional model tasks within the Task Force Hemispheric Transport Air pollution. It uses officially reported, gridded national inventories, complemented with science based data, partly gap-filled with EDGAR. However, since HTAP_v2 is relying on (sub-)national statistics, it may not be as consistent across countries and regions, as a globally calculated inventory using international statistics and global geospatial distributions. Another available global inventory is MACCity, covering the years 1980-2010. We compare EDGAR_v4.3 with HTAP_v2 and MACCity in order to explain differences from national estimates and address emission inventory uncertainties, indicating weaknesses and strengths of these databases. We present the geospatial distribution of emissions at 0.1x0.1 degree resolution, comparing the contribution of developing and emerging countries with industrialized regions, both as absolute and per capita data.