Input uncertainty analysis of the dynamic soil model SMART2 using Monte Carlo sampling

A dynamic soil model SMART2 has been developed to estimate long-term chemical changes in soil and soil water in response to changes in atmospheric deposition. The aim of this study was to rank the input parameters on the basis of their contribution to model uncertainty in order to determine which additional data would best improve the reliability of predictions. The Monte Carlo technique was used in combination with regression analysis. The uncertainty study was conducted for stream water pH and NO3 concentration and for soil base saturation. Parameters defining nutrient mineralization in soil and plant uptake of nitrogen and base cations were also studied. Nitrogen uptake was mainly influenced by N concentration in stem and leaves, but also by parameters defining N mineralization in soil and N denitrification. Weathering rates and nutrient concentrations in vegetation appear to be the most important inputs explaining base saturation of mineral soil, NO3 concentration and pH of stream water.