Land-Use Emissions Model (LUEM)

The Land-Use Emissions Model (LUEM) computes the emissions of gaseous pollutants (greenhouse gases, gas species involved in ozone chemistry, aerosol formation and acidifying compounds), stemming from natural and land-use related sources. The calculations in IMAGE 2.2 are based on Kreileman and Bouwman (1994). The terrestrial flux of CO2 is computed by the Terrestrial Carbon Model (TCM). Therefore, the land-use emissions model focuses on emissions of other important gases, including:

• greenhouse gases (CH4, N2O)
• ozone precursos (NOx, CO, NMVOC)
• acidifying compounds (SO2)

For an explanation of compounds and groups of compounds, see definitions - chemical compounds.

The major input and output variables of the Land-Use Emissions model include:

Model input	Quantity of C burnt during deforestation
	Crop production (to derive quantity of crop residues)
	Human population
	Harvested rice area
	Feed intake for non-dairy and dairy cattle
	Animal populations for all animal categories distinguished
	Use of synthetic N fertilizers and animal manure N
	Areas and production of leguminous crops
	NPP, soil-fertility index, mean monthly temperature and soil moisture content
	Land cover areas
Model assumptions	All emission factors*
	C flux per unit area for savannas
	Ratio residue : harvested product for each crop and DM and N content for each crop
	Fraction of crop residues burnt*
	Fractions of irrigated + rainfed wetland rice harvested area of total rice harvested area
	Fraction animal manure that is available for application (i.e., all manure minus excretion during grazing and burning of manure)*
Model output	Emission by source and compound
* Scenario assumptions on emission factors and activity levels are made for landfills, savanna and agricultural waste-burning and fertilizer use. All other emissions are model-driven on the basis of the activity levels used

The sources, gas species emitted, geographic detail of the calculations, activity levels and key assumptions and origin of the input data are presented in the summary table below. In general, gas emissions are calculated as:

with:
E = emission
A = activity level (e.g. fertilizer use, feed intake, amount of biomass burnt)
Ef = emission factor (i.e. the emission per unit of the activity level).

Exceptions to this emission-factor approach are:

• several sources for which the emission is a global estimate (see table below, geographic detail)
• N2O emissions from soils under natural vegetation. The calculations are based on a regression model for the natural N2O soil emissions, and fluxes vary with climate, net primary production and soil conditions. This regression model is modified from Kreileman and Bouwman (1994) on the basis of a more extensive set of measurement data covering a wider range of ecosystems.

In contrast to Kreileman and Bouwman (1994), calculations for a number of (anthropogenic) agricultural emissions follow the default IPCC Methodology for National Emission Inventories (IPCC, 1997). To make our calculations consistent with this methodology, some additional sources were included, such as N2O emissions from biological N fixation by leguminous crops (pulses and soybeans), from soil incorporation of crop residues, and indirect sources of N2O (caused by leached N inducing N2O emissions from groundwater and surface waters). In addition, we included NOx emission calculations for agricultural fields and natural ecosystems. The IPCC (1997) default methodology was not used for some sources:

• for CH4 emissions from cattle we use the approach of EPA (1994). In this approach CH4 emissions are directly related to the feed intake of cattle. The feed intake is a function of the energy requirement for maintenance (based on body weight), for obtaining feed (grazing), growth, animal traction and calving. Hence, the CH4 generation is calculated on the basis of the scenario for animal productivity considered, and no additional assumptions are needed. A different approach is taken for the other animals (sheep and goats, pigs, poultry) for which the IPCC (1997) default emission factors were used.
• for CH4 emissions from wetland rice fields (irrigated, rainfed and deep water rice), we follow the approach of Neue (1997). Here, the IPCC methodology could not be applied because (reliable) data on organic amendments in rice cultivation (required for IPCC methodology) are not available for most countries and regions
• for the enhanced N2O emissions following deforestation (land-clearing effects) IPCC (1997) provides no guidelines. These emissions were calculated on the basis of N2O pulses during the first 10 years after clearing, as discussed in detail by Kreileman and Bouwman (1994). In the first year after the disturbance, the fluxes amount to 5 times the flux of the original ecosystem. They decrease linearly to the level of the new ecosystem in the 1oth year, and this is ususally lower than the flux of the original forest.

Except for those sources where a global estimate is used (see table below), emissions from all sources vary according to the scenario considered. For example, the CH4 emissions from cattle per unit of product decrease with increasing productivity. For future years we assumed that emission factors for livestock in less-industrialized countries slowly evolve to the emission factors in industrialized countries. N2O emissions from natural terrestrial ecosystems change according to land-cover and climate changes. Fertilizer-induced N2O emissions change according to both the scenario on fertilizer use and substitution of synthetic fertilizers by animal manure (see scenario assumptions - land use. This occurs when livestock production intensifies and more animal manure becomes available.

In addition to the changing emissions as a result of productivity increases, additional scenario assumptions are made for landfills, savanna and agricultural waste-burning and fertilizer use (see scenario assumptions land-use emissions).

Summary of the computation of land-use and natural emission
Source	Geographic detail	Activity level/key assumptions	Origin of activity level
Biomass burning (deforestation) (CH4, CO, N2O, NOx, NMVOC SO2)	Grid	Quantity of C burnt during deforestation	terrestrial carbon cycle model
Savanna burning (CH4, CO, N2O, NOx, NMVOC)	Grid	Quantity of C burnt in savannas (fixed carbon flux per unit area)	land-cover model
Agricultural residue burning (CH4, CO, N2O, NOx, NMVOC, SO2)	Grid	Crop production, fraction above-ground residues and burning fraction based on Bouwman et al. (1997), Smil (1999); EPA (1994) and emission factors from IPCC (1997) (see also crop residues in agricultural economy model (AEM))	land-cover model
Landfills (CH4)	Regional	Urban population	Scenario
Domestic sewage treatment (CH4, N2O)	Regional	Total human population	Scenario
Wetland rice fields (CH4)	Grid	Harvested areas of irrigated, rainfed and deepwater rice from FAO (1999). Regional emission factor for 1970-1995 based on Neue (1997).	land-cover model/scenario
Animals (CH4)	Grid	Cattle: feed intake (cattle) and total number of animals according to Alcamo et al. (1998). Other animals: animal populations; emission factors from IPCC (1997).	Scenario/land-cover model
Animal waste (CH4, N2O)	Grid	Total number of animals. Emission factors from IPCC (1997).	land-cover model
Arable land (N2O,NOx)	Grid	N2O: N-fertilizer use (synthetic fertilizer + animal manure), N- fixing crops (pulses, soybeans), and crop residue incorporation; incorporation of crop residues is calculated as total above-ground residues minus biofuel use, agricultural waste-burning and use of residues for feed (see feed for animals), plus below-ground residues. Emission factors and general procedure from IPCC (1997).	Scenario/land-cover model
	Grid	NOx: N fertilizer use (synthetic fertilizer + animal manure); emission factor from Veldkamp and Keller (1997). Crop residue incorporation as for N2O	land-cover model
Indirect sources (N2O) from surface and gound water caused by N leaching from soils	Region	Use of synthetic N fertilizers and animal manure N and population according to IPCC (1997).	Scenario
Land-clearing effects (N2O, NOx)	Grid	Forest clearing rates (deforestation) N2O: post-clearing emission rates according to Kreileman and Bouwman (1994).	land-cover model
Aquatic sources (CH4, CO, N2O, SO2)	World	Constant, based on IPCC (CH4, CO); (N2O); oceanic DMS (IPCC, 1995)	-
Natural wetlands (CH4)	Grid	Constant; global emission based on IPCC (2001).	-
Soils under natural vegetation (N2O, NOx)	Grid	N2O: NPP, soil fertility index, monthly mean temperature and soil moisture, soil type, (Modified from Bouwman et al. ,1993; Kreileman and Bouwman, 1994).	climate model, land-cover model
	Grid	NOx: areas of biomes ; emission factors for biomes from Davidson and Kingerlee (1997). These emission factors include canopy reduction factors to describe absorption of NOx by plant leaves.	land-cover model
Natural vegetation (NMVOC)	Grid	Area of natural ecosystems (see land cover, land use). Emission from Guenther et al. (1995).	land-cover Model
Other natural sources (CH4, CO, NOx, SO2)	World	Constant, based on IPCC (1995); includes CH4 from termites and methane hydrates, CO from plants and wildfires, NOx from lightning, SO2 from natural sources (volcanoes)	-