EDGAR-HTAP emission inventory (2000-2005)

Currently available global emission inventories differ in the compounds included, regions and emission sources covered, and spatial and temporal resolution. Atmospheric modellers, therefore, often compile their own emission inputs, drawing upon different pieces of the available inventories that are related to their specific project. This is the case with most of the HTAP multi-model

experiments described in the HTAP Interim Assessment Report and in Chapter 4 of this report, where atmospheric modelling groups were asked to provide their own best estimates for emissions for a given year. Although the variability introduced by the different emission estimates captures some of the real uncertainty in the magnitude of the emissions, the variability makes it difficult to compare modelling results.

79

In the HTAP Interim Assessment Report, the TF HTAP recommended further efforts for harmonization of emission inventories. The recommendations included reaching out to other organizations and research projects to facilitate the incorporation of other emission inventories with local knowledge into global emission inventories, especially in regions where emission factors and activity data are poorly known. The European Commission‘s Joint Research Centre initiated a project to implement this recommendation. The result of this project is the EDGAR-HTAP inventory, the development of which is described below.

The goal of the EDGAR-HTAP project is to compile a global emissions dataset for 2000 to 2005 using, to the extent possible, official or scientific inventories at the national or regional scale that are likely to be accepted by policy makers in each region of the world. The resulting inventory is a compilation as opposed to a consistent inventory calculated using activity data and emission factors.

In order to provide a complete global picture of emissions, this compilation of different official inventories needed to be represented in a harmonized way, and gap-filled with global emission data of EDGARv4.1, as recently made public at http://edgar.jrc.ec.europa.eu [EC/JRC/PBL, 2010]. The Intergovernmental Panel on Climate Change (IPCC) code—defined by the IPCC Guidelines (1996) and conforming with the United Nations Framework Convention on Climate Change (UNFCCC) Common Reporting Format, CRF—is used as the key for harmonizing the sectoral data. A

corresponding EDGAR-HTAP coding system is developed to allow aggregation at different levels of detail. It should be mentioned that different levels of detail are presented for the different countries in order to avoid subtraction of country-specific emission data. Moreover, the database system allows the user to calculate for given activity datasets the implied emission factor and so to compare the level of emission and abatement measures across regions. To provide the modelling community with consistent gridding of these emissions, the resulting emissions by country and sector are allocated to a 0.1º × 0.1º grid using the EDGARv4.1 database. The highest aggregation level, as represented in Table 3.1, is used for presenting the gridded maps for each year and each significant compound of that aggregated sector. The combination of policy and scientific inventories makes this inventory different from the work by Lamarque et al. [2010], as described in section 3.2.2.

Table 3.1. EDGAR-HTAP categorisation of emission sources (as applied for the gridded maps) Source emission category at highest level of aggregation IPCC sectors

htap_1_aircraft 1.A.3.a

htap_2_ships 1.A.3.d

htap_3_energy_combustion_&_fuel_production/distribution 1.A.1.a, 1.A.1.b, 1.A.1.c, 1.A.2, 1.B.1, 1.B.2

htap_4_industry_processes 2.A, 2.B, 2.C, 2.D, 2.G

htap_5_ground_transport 1.A.3.b, 1.A.3.c, 1.A.3.e

htap_6_residential 1.A.4, 1.A.5

htap_7_solvents 3.A, 3.B, 3.C, 3.D

htap_8_agriculture 4.A, 4.B, 4.C, 4.D (not 4E)

htap_9_agricultural_waste_burning 4.F, 4.G

htap_10_waste 6.A, 6.B, 6.C, 6.D

Pollutants included in the EDGAR-HTAP inventory are carbon monoxide (CO), nitrogen oxides (NOx), sulphur dioxide (SO2), non-methane volatile organic compounds (NMVOC), methane (CH4), ammonia (NH3), black carbon (BC), as well as particulate matter in coarse + fine (PM10) and fine (PM2.5) fractions. The EDGAR-HTAP inventory has been compiled using the following

inventories in descending priority order:

1. Inventories submitted by Parties to the UNFCCC. So-called Annex I countries

(industrialized countries and economies in transition) report annual emissions of greenhouse gases (CO2, CH4, N2O, F-gases) and indirect greenhouse gases pollutants (CO, NMVOC, NOx

80

and SO2). These data have to follow specific guidelines in reporting and inventory

construction and are subject to review by external inventory expert teams. For the EDGAR-HTAP inventory, UNFCCC emissions data were downloaded from the UNFCCC data portal (http://unfccc.int/ghg_data) in the CRF format. Non-Annex I parties submit emission inventory data to the UNFCCC periodically under different reporting requirements than Annex I parties. These data vary in quality and detail, especially regarding air pollutants, and do not provide complete time-series. For this reason, non-Annex I data were not used in the EDGAR-HTAP inventory. UNFCCC emissions follow IPCC source classifications.

2. Inventories submitted by Parties to the LRTAP convention. Countries that are parties to the LRTAP Convention provide annual emission inventory data for CO, NH3, NMVOC, NOx, SOx, particulate matter, heavy metals and persistent organic pollutants. These inventories are reported to the European Monitoring and Evaluation Programme‘s (EMEP) Centre on Emission Inventories and Projections (CEIP). Emissions as reported by Parties have been kindly provided by CEIP (Robert Wankmueller, personal communication, 2009). These data have been summed to IPCC source categories using the conversion table from the

Nomenclature for Reporting to CRF.

The UNFCCC and EMEP datasets have been combined for each country and source category, giving priority to the (internationally reviewed) UNFCCC data. Data gaps in time series are filled using interpolation or extrapolation from the last/first year available. Where a complete time series of a compound is missing (BC and OC), EDGARv4.1 is used for gap-filling.

International aviation and international navigation are excluded from this combined official inventory dataset and are taken from EDGARv4.1 bunker data. For the United States, the U.S.

EPA National Emission Inventory (NEI) data for the years 2002 and 2005 have been included in the EDGAR-HTAP, replacing the UNFCCC data described above [U.S. EPA, 2009].

3. Asian countries. For countries in Asia (except China and Japan) the REAS inventory [Ohara et al., 2007] has been included by country and detailed sector for CH4, CO, NOx, SO2, NH3, OC and BC emissions (J. Kurokawa, personal communication, 2009). Emissions have been allocated to IPCC source categories. For those sectors where REAS emissions were reported to be zero, gaps were filled using inter/extrapolation to complete time series or EDGARv4.1 data to complete missing compounds (i.e., NMVOC, PM10 and PM2.5). REAS data for agriculture were available only for the year 2000 and for this sector data from EDGARv4.1 have been used. Emissions in Japan have been taken from the official inventory submitted to the UNFCCC, completed with REAS data for NH3, OC and BC. For China, emissions were provided by IIASA (Z. Klimont, personal communication, 2009) taken from the GAINS-China dataset. Data were available for the years 2000 and 2005 and the years in between have been interpolated and allocated to the IPCC source categories. No gap-filling with EDGARv4.1 was necessary for China.

Other countries, missing sectors, international shipping and international aviation. Emissions from non-Annex I countries and countries outside Asia are taken from the EDGARv4.1 dataset (http://edgar.jrc.ec.europa.eu/overview.php. When inventories described under steps 1-3 did not provide emissions for a specific compound or sector, data were included from EDGARv4.1. This concerns mainly emissions of BC and OC for Annex I countries (to gap-fill EMEP data) and agricultural emissions, and NMVOC, PM10 and PM2.5 emissions for Asian (except China and Japan) countries (to gap-fill REAS data). Emissions from international shipping and from international aviation are taken from EDGARv4.1. For the international shipping and aviation, the EDGARv4.1 data were calculated using IEA/OECD [2007] fuel consumption and methodology; emission factors were taken from Dalsøren et al. [2009; 2007] and Eyring et al. [2005] for shipping and Eyers et al.

[2004] for aviation. An overview of the coverage of species by each official dataset and the geo-coverage of the various emission datasets is presented in Table 3.2.

81

Table 3.2. Overview of the coverage of compounds and regions in the official emission datasets.

The resulting emissions by country and IPCC source category are allocated to a 0.1° × 0.1°

grid, as defined in the EDGARv4.1 database. The 0.1º × 0.1º grid follows a reference country and sea grid definition based on GPWv3 [CIESIN, 2005]. Allocation of emissions is based on geographical data such as location of energy and manufacturing facilities, road networks, shipping routes, human and animal population density and agricultural land use. The resulting emissions estimates are presented in the context of historical trends and future scenarios in section 3.4. All grid maps for all years and all compounds at sectoral level are made available on the HTAP wiki site (http://htap.icg.fz-juelich.de/data/), together with other related documentation tables.

Figures 3.1 and 3.2 summarize the major spatial features of the EDGAR-HTAP inventory.

Figure 3.1 presents total emissions in the year 2005 of the ten major species included in the EDGAR-HTAP inventory (NOx, SO2, CH4, NH3, NMVOC, CO, PM10, PM2.5, BC and OC). Note that the emissions do not include biogenic emissions from land use change and forestry or savannah burning.

Figure 3.2 presents a second series of maps for each of these species showing some of the key contributing source types. These source types have been chosen so that a perspective on the global distributions of all major source types can be obtained. Source types included in Figure 3.2 are:

international and domestic aviation, shipping in international and inland waterways, waste disposal, agriculture, industrial solvent use, agricultural waste burning, energy combustion and fuel production, industrial process emissions, residential, and ground transportation by road and rail.

82

NOx totals (excl. Savannah burning) in ton NO2/grid cell SO2 totals (excl. Savannah burning) in ton SO2/grid cell

CH4 totals (no land-use change & forestry)in ton CH4/grid cell NH3 totals (no land use change & forestry) in ton NH3/grid cell

VOC totals in ton NMVOC species/ grid cell CO totals (no Savannah burning) in ton CO/grid cell

PM10 totals (no Savannah burning) in ton species/grid cell PM2.5 totals (no Savannah burning) in ton species/ grid cell

BC totals (no Savannah burning) in ton C/ grid cell OC totals (no Savannah burning) in ton species/ grid cell Figure 3.1. Gridded maps of anthropogenic emission totals in 2005 for each compound in the

EDGAR-HTAP emissions database on a 0.1° × 0.1° grid (in tons of species/grid cell). (Please note: biomass is included but neither biogenic emissions from land use change and forestry nor savannah burning). [Graphic courtesy of European Commission Joint Research Centre.]

83

ton NO2 emissions from aviation (international+domestic) /cell ton SO2 shipping emissions (international +inland waterways) /cell

ton CH4 emissions for the waste sector per cell ton NH3 emissions for agriculture (no savannah/waste burning) /cell

ton NMVOC emission from solvent industry per cell ton CO emission of agricultural waste burning (no savannah fire) /cell

ton PM10 emission of energy combustion & fuel production /cell ton PM2.5 process emissions from industry per cell

ton BC emissions from residential sector per cell ton OC emissions from ground transport (road – rail) per cell Figure 3. 2. Gridded maps of sector-specific anthropogenic emissions in 2005 in the

EDGAR-HTAP emissions database on a 0.1° × 0.1° grid (in tons of species/grid cell). [Graphic courtesy of European Commission Joint Research Centre.]

84