Application of BIOME-BGC model to managed forests: 1. Sensitivity analysis

A process-based model BIOME-BGC designed for simulation of biogeochemical element cycling in terrestrial ecosystems was prepared for application to managed forest ecosystems in temperate Europe. New routines were implemented that permit specification of thinning, felling and species change when planting new forest. Other changes were implemented to water cycling routines, specifically to precipitation and evaporation, simulation of industrial nitrogen deposition and fine roots mortality. The major aim of the paper was to conduct a sensitivity analysis of the adapted model. We specifically analysed the effects of site and eco-physiological parameters on the modeled state variables (carbon pools in biomass, litter and soil and net primary production (NPP)). The analysis revealed a high sensitivity of all tested variables to the following site parameters: total precipitation, rooting depth, sand fraction (for sandy soils only), ambient CO2 and parameters of nitrogen input. Similarly, the tested variables were shown to be highly sensitive to the following eco-physiological parameters: leaf and fine root C:N ratio, new stem C to new leaf C ratio, new fine root C to new leaf C ratio, specific leaf area, maximum stomatal conductance, fire mortality and fraction of N in Rubisco (specifically for deciduous species). Additionally, the whole plant mortality had a high effect on carbon pools, but a small effect on NPP.