Evaluation of model behaviour with respect to the biogeochemistry at the Solling spruce site

The performance and predictions of 11 biogeochemical models, applied to time series data from a spruce site in Solling, Germany were evaluated. All models are deterministic, process-oriented models, and represent a wide range of modelling approaches with respect to time and space resolution and complexity. Although process formulations vary, the basic processes are shared by most models. The scope of the models is to calculate soil chemistry characteristics relevant to assess stress factors for forests (e.g. unfavourable high inorganic Al concentrations and AlCa rations in soil solution) induced by acid deposition. The evaluation showed that the general trends and levels in chemical variables, such as soil solution pH, and concentrations of base cations, SO2−4 and Al, as well as base saturation can be reproduced by the models. Most models are incapable of modelling correctly pH and Al concentrations simultaneously. Modelling NO−3 concentrations was generally not very successful, despite several complex modelling approaches. The modelling efforts highlighted that the concentrations of base cation, SO2−4 and Ali are coupled in a complex and nonlinear way, which has strong implications for the response to reductions in acid deposition. The rate of chemical weathering obtained by budget studies was supported by the models. The comparison of the models showed that the links between soil chemistry and to forest growth are incomplete. It is recommended to continue to work on certain processes such as Al-chemistry and N transformations, and on the interaction between soil chemical status and biota.