Higher tier estimation of ammonia emissions from synthetic nitrogen fertilizers in emission inventories: data collation, methodology and analysis

Synthetic fertilizers are a key source of ammonia (NH 3 ) emissions globally and national emission inventories are an essential element of international agreements aimed at reducing these emissions. Here an analysis of published measurements of NH 3 emission from synthetic fertilizer found that emissions were primarily determined by the experimental approach, the chemical composition of the fertilizer, environmental variables and the method of application. Measurements made using wind tunnels, semi-open chambers and the micrometeorological method produced similar results whereas those made with closed or ventilated chambers, Dräger tubes or 15 N were lower. Measurements made outdoors were lower than those made in the laboratory. The most important factors determining the NH 3 emissions were the method of application, the chemical composition of the fertilizer and, for fertilizers not based on solid forms of urea, the soil pH. National inventory compilers normally take account of these factors but are encouraged to use more detailed methodologies. Here we present a range of statistical models that would allow inventory compilers to also take into account application rate, air temperature, rainfall and soil clay content. However, to take advantage of these models would require the availability of additional, good quality input data. The conditional R 2 of the models developed accounted for 44–55 % of the variance in NH 3 emission measurements but the marginal R 2 only 19–27 %. We recommend further investigation into the source of the low marginal R 2 and, in particular, into the differences caused by the use of the different measurement methods.