气候变化领域集成服务门户

Agricultural activity plays a significant role in the atmospheric carbon balance as a source and sink of greenhouse gases (GHGs) and has high mitigation potential. The agricultural emissions display evident geographical differences in the regional, national, and even local levels, not only due to spatially differentiated activity, but also due to very geographically different emission coefficients. Thus, spatially resolved inventories are important for obtaining better estimates of emission content and design of GHG mitigation processes to adapt to global carbon rise in the atmosphere. This study develops a geoinformation approach to a high-resolution spatial inventory of GHG emissions from the agricultural sector, following the categories of the United Nations Intergovernmental Panel on Climate Change guidelines. Using the Corine Land Cover data, a digital map of emission sources is built, with elementary areal objects that are split up by administrative boundaries. Various procedures are developed for disaggregation of available emission activity data down to a level of elementary emission objects, conditional on covariate information, such as land use, observable in the elementary object scale. Among them, a statistical scaling method suitable for spatially correlated areal emission sources is applied. As an example of implementation of this approach, the spatial distribution of methane (CH4) and Nitrogen Oxide (N2O) emissions was obtained for areal emission sources in the agriculture sector in Poland with a spatial resolution of 100 m. We calculated the specific total emissions for different types of animal and manure systems as well as the total emissions in CO2-equivalent. We demonstrated that the emission sources are located highly nonuniformly and the emissions from them vary substantially, so that average data may provide insufficient approximation. In our case, over 11% smaller emission was estimated using spatial approach as compared with the national inventory report where average data were used. In addition, we quantified uncertainties associated with the developed spatial inventory and analysed the dominant components in total emission uncertainties in the agriculture sector. We used the activity data from the lowest possible (municipal) level. The depth of disaggregation of these data to the level of arable lands is minimal, and hence, the relative uncertainty of spatial inventory is smaller when comparing with traditional gridded emissions. The proposed technique allows us to discuss factors driving the geographical distribution of GHG emissions for different categories of the agricultural sector. This may be particularly useful in high-resolution modelling of GHG dispersion in the atmosphere.