Impacts of model structure and data aggregation on European wide predictions of nitrogen and green house gas fluxes in response to changes in livestock, land cover, and land management.

Various model approaches have been developed for assessing emissions of different forms of reactive nitrogen in various parts of Europe at various geographic resolutions and for various time periods. The modeling approaches include emission factor approaches, empirical models, simple process-based models, and detailed ecosystem models. In this study, we compared three relatively simple process-based models, developed for the national scale (Integrated NITrogen Impact AssessmenT model On a Regional Scale (INITIATOR2)), European scale (MITERRA) and global scale (integrated model to assess the global environment (IMAGE)), with respect to their response to structural and technological changes in the agricultural systems based on the IPCC B2 baseline scenario for the period 2000-2030. Changes are predicted by the IMAGE model and relate to crop yield, crop area, animal numbers, and N fertilizer inputs. The predicted relative changes by IMAGE are used in INITIATOR2 and MITERRA while relating the animal categories and crop categories in IMAGE to those in the latter models. A comparison was made of NH3, N2O and NOx emissions and N leaching to ground water. We compared predictions for the years 2000 and 2030 for: (i) the Netherlands between INITIATOR2 and MITERRA and (ii) Europe (EU-27 countries) between MITERRA and IMAGE. The results of the comparison are presented and evaluated in view of differences in model structure and the effect of aggregating input data at larger spatial scales. [ABSTRACT FROM AUTHOR]