A process-oriented model of N2O and NO emissions from forest soils 2. Sensitivity analysis and validation

The process-oriented model PnET-N-DNDC describing biogeochemical cycling of Cand N and N-trace gas fluxes (N2O and NO) in forest ecosystems was tested for its sensitivity to changes in environmental factors (e.g., temperature, precipitation, solar radiation, atmospheric N-deposition, soil characteristics). Sensitivity analyses revealed that predicted N-cycling and N-trace gas emissions varied within measured ranges. For model validation, data sets of N-trace gas emissions from seven different temperate forest ecosystems in the United States, Denmark, Austria, and Germany were used. Simulations of N2O emissions revealed that field observations and model predictions agreed well for both flux magnitude and its seasonal pattern. Differences between predicted and measured mean N2O fluxes were ,27%. An exception to this was the N-limited pine stand at Harvard Forest, where predictions of fluxes deviated by 380% from field measurements. This difference is most likely due to a missing mechanism in PnET-N-DNDC describing uptake of atmospheric N2O by soils. PnET-N-DNDC was also validated for its capability to predict NO emission from soils. Predicted and measured mean NO fluxes at three different field sites agreed within a range of 613%. The correlation between modeled and predicted NO emissions from the spruce and beech stand at the Höglwald Forest was r 5 0.24 (spruce) and r 5 0.35 (beech), respectively. The results obtained from both sensitivity analyses and validations with field data sets from temperate forest soils indicate that PnET-N-DNDC can be successfully used to predict N2O and NO emissions from a broad range of temperate forest sites.