Changsheng Li
Institute for the Study of Earth, Oceans and Space, University of New Hampshire, Durham, NH, USA

1. The strengths and weaknesses in the current GEIA emission databases for NH3

The current GEIA NH₃ file contains data sets for eleven NH₃ sources for the year 1990, as published by Bouwman et al. (1997). The sources are (1) Domesticated animals, (2) Synthetic fertilizers, (3) Crops, (4) Humans, (5) Biofuel combustion, (6) Savannah burning, (7) Deforestation, (8) Agricultural residue burning, (9) Industrial processes,  (10) Fossil fuel combustion, and (11) Oceans. The GELA files include global maps for each of the eleven sources at 1° x 1° grid resolution, as described by Bouwman et al. (1997) and references therein. With an emission factor approach, the authors estimated NH₃ emissions at country scale based on the statistics and selected factors for the countries. The GEIA NH₃ data sets indicate that 70% of global NH₃ emission is related to food production with a dominant source of excreta from domestic animals.

Uncertainty for the data sets is not presented in the GEIA files. However, Bouwman et al. (1997) discussed uncertainty in their paper although only focusing on the statistical data. The emission factors can be an important source of uncertainty due to their high variations induced by the substrate quality, climatic conditions, soil properties, vegetation, and anthropogenic activities. In June 2002 the U.S. National Academy of Sciences (NAS) released a report, The Scientific Basis for Estimate Emissions from Animal Feeding Operations (NAS, 2002), based on a rigorous scientific review of air emission factors as related to current animal feeding and production systems in the U.S.  The NAS report stresses that emissions of NH₃ and other relevant gases from animal feeding operations can be highly variable both in time and space at any one facility and between different facilities of the same classification, and thus supplying a single emission factor without an estimate of uncertainty is deemed inadequate. 

2. Alternative emission distributions

Van Aardenne et al. (2001) have extended the global, anthropogenic emissions of NH₃ to the period 1890-1990 with 10 year intervals based on the historical activity statistics and selected emission factors adopted in the Emission Database for Global Atmospheric Research (EDGAR 2.0) (Olivier et al., 1999). In the new results, uncertainty analysis basically remained at the same level as in the works done by Bouwman et al. (1997). Several latest developments on NH₃ emission factors at regional or country scale (e.g., Battye et al., 1994; James et al., 1997; MRI, 1998; Misselbrook et al., 2000; Hutchings et al., 2001) can be utilized for updating the GEIA databases.

Mass balance approach and process-based models need to be considered as alternatives to emission factors. The process-oriented models of soil NH₃ emission (e.g., Sherlock and Goh, 1984; Singh and Nye, 1986; Izaurralde et al., 1990; Jayaweera and Mikkelsen, 1990; Li, 2000; Plochl, 2001; Potter et al., 2001; Sogaard et al., 2002) have incorporated impacts of environmental factors on NH₃ production/deposition, and hence could serve as a basis for improving NH₃ inventory studies.

3. Emission trends

Van Aardenne et al. (2001) estimated global NH₃ emissions increased from 9 Tg NH₃–N in 1890 to 43 Tg in 1990, dominated by agricultural sources. Continuous increases in NH₃ emissions at global scale would be expected due to the changes in food structure as well as livestock population in many developing countries, especially in Asia. 

4. The seasonality of the emissions

No observations for seasonality of NH₃ emissions at regional or global scale have been reported although numerous researchers observed seasonal variations in NH₃ fluxes from agricultural lands at site scale (e.g., Parton et al., 1988; van Hove et al., 2002; Robarge et al., 2002). Climatic conditions (e.g., air temperature, precipitation etc.) seemed dominating the seasonality through their effects on decomposition, ammonification, nitrification, denitrification and other NH₃ production–related processes occurring in the soils. Regional or global seasonality of NH₃ emissions could be obtained by means of model simulations in conjunction with observations at large scales (e.g., utilizing the sensors on aircraft or spacecraft).    

5. Natural emissions as a function of soil characteristics

Observations for NH₃ emissions from natural soils are sparse. A recent modeling study indicated that cultivated soils in China emitted 13 Tg NH₃-N in 1990 although there were overlaps with animal manure- and fertilizer-induced NH₃ (Li et al., 2002). Driven by biological degradation of proteins in soil organic matter as well as ammonium/ammonia equilibria, NH₃ production occurs in a wide range of soils even without any animal manure or fertilizer applied (Dawson, 1977; Lemon and Van Houtte, 1980; Vlek et al., 1981). The soil “background” emissions, with relatively low rates but from a vast area, could have a non-negligible contribution to the global NH₃ budget.

6. New GEIA inventory projects

Limitation of the emission factor approach will become obvious when it is applied for NH₃ inventories driven by management alternatives or climate change scenarios. Future GEIA activities could turn to model-centered approaches, which involve intensive measurements, GIS databases holding characterization of animal wastes and other major NH₃ sources, and development and application of process-based models.