MNP logo
Logo partner leeg
EDGAR
Home Contact PrintSitemap
Search
    > Documentation > Key differences with EDGAR2
Partners EDGAR  
Key differences with EDGAR2

On this page the key differences between EDGARV2 and EDGARV3 are descriped.

Aim of EDGAR3

The overall aim for Version 3.0 was to update the inventories from 1990 to 1995, and for direct greenhouse gases also to 1970, to include new greenhouse gases. After consultation of the users, the objectives have been somewhat changed and extended. Thus, specific aims were:

  • update/extension from 1990 to 1995;
  • extend time series for direct greenhouse gases to 1970-1995;
  • include new 'Kyoto' greenhouse gases HFCs, PFCs, SF6;
  • greenhouse gas emissions also on per country basis using IPCC source categories;
  • include NH3;
  • improve/include uncertainty estimates and time profiles.

For updating and extended time series different priorities were given for the following groups of gases:

  • direct greenhouse gases CO2, CH4, N2O and new gases HFCs, PFCs, SF6: 1970-1995;
  • ozone precursors CO, NOx, NMVOC as well as SO2 and NH3: update 90 and 95;
  • extend CFCs, halons, HCFCs to 1900-1995

Special attention was given to the compilation of a reference dataset for new gases as none was available. For the update of the current Version 2.0, we followed the following principles:

  • Activity data: update by including relevant statistics for the period 1970-1995, after checking for possible changes of source categories; this implies the inclusion of the 'new' countries, e.g. for the former USSR.
  • Emission factors: only to be changed for 1990 if validation showed major discrepancies; only to be changed for 1995 compared to 1990 if there are concrete indications that there major changes have occurred that cannot be neglected; the same holds for factors for 1970, in particular for direct greenhouse gases.
  • Grid maps: only to be updated if maps available of better quality or better applicability.
  • Additional sources: coal fires, oil fires, vegetation fires in temperate regions, domestic waste combustion and wastewater handling were added, based on the significance in some countries for specific emissions.
Updating emission factors for 1990

As a result of the validation of EDGAR 2.0 with other global and regional emission inventories it was decided that several items should be modified for the reference year 1990. Compared to Version 2.0 the following amendments have been made for 1990:

  • The emission factors for 1990 for direct greenhouse gases CO2, CH4 and N2O have been brought more in line with the defaults recommended in the Revised IPCC Guidelines for Greenhouse Gas Inventories and for reference purposes any departures from them will be clearly identified. For CO2 from fossil fuel use, emission factors per detailed fuel type will be used (in V2.0: one aggregated factor for coal, oil and gas). This also means that the agricultural emissions will be affected considerably by the inclusion of some 'indirect' emissions. Other examples of areas where emission factors will be updated are CH4 from rice and landfills.
  • Global default emission factors for NOx, CO and NMVOC for the following non-road transport activities are updated: Rail transport, Inland water, Other land - non-road and Non-specified transport. Emission factors are entered for coal, diesel oil and gasoline when applicable. 
  • Global default emission factors for NOx and SO2 for sea ships have been updated; in particular the emission factor for NOx has increased significantly.
Activity data 1970-1995
Activity data were collected for the period 1970-1995. A major part could be drawn from IEA (energy), UN (supplementary energy, industrial production) and FAO (agriculture) databases. But for some source like biofuels and specific industrial production of commodities like adipic acid, nitric acid and fluorinated carbons country statistics are not readily available. For each of these latter compounds additional data sources were found and used. 
In the process of updating 1990 activity data with more recent statistical datasets, these levels are often changed to a lesser or larger degree. This is caused by the phenomenon that statistics of activity data of the most recent years tend to change during a couple of years after the first compilation. This happens in particular in non-OECD countries, however, also in industrialised countries this phenomenon can be observed, although in these countries the changes are often only minor. In addition, data for the former USSR have become rather weak due to inconsistencies between the sum of the new countries and the 1990 data for the former USSR.
For biofuels we use the previous V2.0 dataset for less developed countries and FAO fuelwood plus IEA data for OECD countries. In addition, for the IPCC sources 'Land-use change and forestry' (LUCF) and 'Waste' there is no readily available data in time-series per country. Here, in line with the approach taken for the compilation of the GEIA NH3 inventory, biomass burning data (vegetation fires) for LUCF were based on FAO reports providing 10 year averaged estimates. For agricultural waste burning too the activity data were essentially based on the methodology used for NH3, however using updated fractions for the amount of agricultural waste per unit of net crop production, and using much lower fractions burned in OECD countries. For waste, the activity data per country are based on a fit with of international waste generation figures per capita with per capita income per country.
Emission factors for 1995
For the update of the 1990 EDGAR 3.0 emission factors to 1995 the following three sectors have been selected: large combustion plants, mobile sources and solvent use. Updated emission factors, or another mix of sub-activities, for 1995 were required for sources such as coal mining, gasoline cars, shifting type of rice cultivation, landfills with gas recovery. In addition, also for power plants and some industries in countries where additional control technology e.g. for SO2 and NOx has been installed updated emission factors have been considered. For extension of emission factors for CH4 and N2O towards 1970 similar considerations have been made. Given the limited resources available for the 1995 update it was decided to focus only on the most important developments and changes which might influence emission factors. There are many sectors in which important emission reduction measures have been implemented.
Major differences compared with version 2

The largest difference with EDGAR 2.0 emissions is in the following sources:

  • wastewater treatment has been added, which is a substantial source of CH4;
  • indirect emissions of N2O from agriculture have been added (complying with the Revised IPCC Guidelines of 1997);
  • agricultural waste burning emissions have been decreased substantially, in particular for CO, due to lower fractions assumed to be burned and a substantial decrease of the CO emission factor;
  • temperate forest fires show considerable emissions, though highly variable between years;
  • NMVOC from domestic waste burning, in Version 2.0 called 'uncontrolled waste burning', have decreased substantially after a re-analysis of the source data;
  • fossil fuel fires have been added, increasing fuel-related emissions in China considerably;
  • NOx from international shipping has increased substantially due to an update of the emission factors;
  • the spatial distribution of sources allocated with the population maps has changed substantially, due to the introduction of another base map from Li (1996) and applying urban and rural maps where appropriate;
  • the use of other vegetation maps for allocating deforestation and savanna burning on the grid.

The 1990 emissions have not only changed due to updates of emission factors, but also since international statistics of activity data of the most recent years tend to change during a couple of years after the first compilation. In addition, data for the former USSR have become rather weak due to inconsistencies between the sum of the new countries and the 1990 data for the former USSR (Olivier et al., 2001).

Update of grid maps

Although no specific effort has been made to improve the current grid maps used for allocating country emissions of specific sources to the 1x1 grid, the following new maps have been included:

  • human population distribution, based on a new GEIA population map by Li, also split into an urban and rural population map;
  • steel production plants by process type, which covers a large fraction of coal/coke use in the industry sector (also used for coke production locations);
  • cement production plants;
  • nitric acid production plants;
  • aluminium smelters;
  • rice production in Asia, based on NOP-MLK results by Denier van der Gon;
  • coal fire map for China.
New sources

In EDGAR 3.0 the following new sources have been added:

  • Wildfires/vegetation fires in non-tropical regions: Recognising the importance of emissions related to biomass burning, temperate vegetation fires have been added as an emission source based on the UN/ECE forest fire statistics for 1990 and 1995.
  • Waste handling: Recognising the possible importance of this source category, wastewater treatment and domestic waste combustion were added as sources.
  • Coal fires: Unintentional coal fires at shallow coal deposits have been added in EDGAR as an emission source category for China. This source appeared to be considerable and was so far lacking in EDGAR 2.0. Only for China this emission source has been taken into account, although these fires are known also to occur in other countries (e.g. USA, India, Indonesia). 
  • Oil fires: The Kuwait oil fires in 1992 due to the Gulf war have been included as separate source.
Changes in regional subdivisions

Renaming

The regional subdivision has been adjusted to the RIVM's new regional subdivision (Kreileman et al., 1998), which is also used by other global projects such as IMAGE 2.1. EDGAR 3 uses these regional definitions, with aggregations for Latin America and Africa, which in the basic subdivision are divided into 2 and 4 smaller regions, respectively. In practice, the largest changes have occurred in the EDGAR 2.0 regions "India region", "China regions" and "East Asia", which have been renamed as "South Asia", "East Asia" and "Southeast Asia", respectively. In Figure 1 these regions and the corresponding EDGAR 3 regions are shown on a world map. 

Change in definitions

In order to comply with the standard RIVM regional subdivision, the definition of the three Asian regions has been changed as follows:

  • Afghanistan (AFG):   moved from 10 (H) to 11 (South Asia)
  • Korea (South) (KOR):  moved from 13 (K) to 12 (East Asia)
  • Papua New Guinea (PNG): moved from 13 (K) to 04 (Oceania)
  • Cambodia (KHM):   moved from 12 (J) to 13 (Southeast Asia)
  • Laos (LAO):    moved from 12 (J) to 13 (Southeast Asia)
  • Vietnam (VNM):   moved from 12 (J) to 13 (Southeast Asia)

Furthermore, since the former DDR and the BRD have been merged into Germany (United), which is part of the "OECD Europe" region, compared with EDGAR 2.0 the following change was made:

  • Former DDR (DDR):  moved from 6 (F) to 03 (OECD Europe)

Besides these changes between regions, the following major changes occurred within regions:

  • Former USSR [07 (G)]:   split into 15 new countries
  • Former Yugoslavia [06 (F)]:   split into 5 new countries
  • Former Czechoslovakia [06 (F)]:  split into 2 new countries
  • Former Ethopia [09 (D)]:   split into 2 new countries

Finally, a number of small countries/islands have been added to the old regional definitions.

Reordering
In addition, the order of regions in EDGAR 3 has been changed; now the OECD, Economies-In-Transition (EIT) and Less Developed Countries (LDC) regions have been grouped together, respectively. These three regional clusters are somewhat homogeneous with respect to the development of the national statistical systems and the emission factors for a number of source categories.
EDGAR 3.0 data

The largest differences with EDGAR 2.0 emissions are in the following sources:

  • wastewater treatment has been added, which is a substantial source of CH4;
  • indirect emissions of N2O from agriculture have been added;
  • agricultural waste burning emissions have been decreased substantially, in particular for CO;
  • temperate forest fires show considerable emissions, though highly variable between years;
  • NMVOC from Domestic waste burning, in version 2.0 called Uncontrolled waste burning, have decreased substantially; the same holds for Miscellaneous industrial processes (i.e. non-combustion);
  • fossil fuel fires have been added, increasing fuel-related emissions in China considerably;
  • NOx from international shipping has increased substantially;
  • the spatial distribution of sources allocated with the population maps has changed substantially, due to the introduction of another base map and applying urban and rural maps where appropriate;
  • the use of other vegetation maps for allocating deforestation and savanna burning on the grid.

The 1990 emissions have not only changed due to updates of emission factors, but also since international statistics of activity data of the most recent years tend to change during a couple of years after the first compilation. This happens in particular in non-OECD countries, however, also in industrialised countries this phenomenon can be observed, although in these countries the changes are often only minor. In addition, data for the former USSR have become rather weak due to inconsistencies between the sum of the new countries and the 1990 data for the former USSR.

 
Last update: 1 October 2004