Parameter selection and testing the soil nitrogen dynamics model SOILN

. The SOILN model with a crop growth submodel for grass and cereal crops and the associated soil water and heat model SOIL were selected out of a number of similar models to simulate nitrogen cycling in a soil/crop system. The main parameter values required by the model were selected on the basis of a combination of field experiments and literature sources. Experimental data measured on grassland at Dumfries in the West of Scotland and on arable land at Bush Estate near Edinburgh were used to test the model. Simulated biomass yields and nitrogen contents of harvested biomass were in reasonable agreement with measured values for both grass and cereal crops. There were similar trends in accumulated leached nitrate between the simulations and experiments at the sites. Any discrepancy between simulated and measured nitrate leached appeared to correspond to similar discrepancies between simulated and measured water flow. The comparison between simulated and experimental results suggests that the model with the selected parameter values can simulate nitrogen and carbon cycling both in grassland and in arable land, and make convincing predictions about the effects of varying soil, crop, fertilizer and manure management practices. A basic sensitivity analysis carried out on the parameters determining the biological and biochemical processes showed the model predictions of annual N-leaching are relatively insensitive to all but two of the plant parameters. However, the model predictions of annual N-harvested and dry mass production are sensitive to numerous plant parameters.