Ranking of parameters on the basis of their contribution to model uncertainty

The sensitivity and uncertainty model UNCSAM was implemented to study the influence of model parameters on phosphorus losses (sediment bound P and soluble P in surface runoff and deep percolation as well as P taken up by the crop). The simulations were performed using ICECREAM, a management model for predicting field-scale losses of phosphorus. The UNCSAM study provided deeper insight into the simulation with ICECREAM: certain assumptions made about the input parameters were confirmed, whereas other, unexpected parameters gained importance. Consequently, the parameters determining crop growth, for example, need to be further refined. Further, it could be shown that the same parameters appear to influence the percolated P amounts and the amounts of P taken up by crop, which in turn affect percolation of water and transpiration. Contrary to this, particularly soluble P in surface runoff was mainly influenced by parameters, which affect the P pools in soil instead of those, which affect surface runoff of water.