Schätzung von Bodenmerkmalen und Modellparametern für die Waldökosystemsimulation auf Basis einer Großrauminventur

To amend the database needed for a large-scale model-based assessment of Austrias forests with regard to climate change impacts and adaptation strategies soil pools and model parameters were estimated for the grid of the Austrian National Forest Inventory (ÖWI). Based on data of the Austrian Forest Soil Survey (WBZI) and categorized site information of the ÖWI a combined statistical and knowledge-based approach was chosen for the estimation of carbon and nitrogen pools as well as pH-value. In addition, site-specific parameters for the soil module of the forest ecosystem model PICUS vl.41 were estimated in order to allow a consistent initialisation of the model at sample plots of the ÖWI. Estimated average soil pools were evaluated against data from the WBZI, showing good agreement over soil types and ecore- gions. Moreover, estimated mean soil C stocks compared well to estimates in the literature. However, the achieved precision in predicting plot-specific soil characteristics was limited yet in the range of previous statistical approaches. To test the full initialisation set (i.e. soil characteristics and model parameters) holistically simulations with PICUS vl.41 were conducted over a wide ecological gradient, showing good agreement between observed and predicted stemwood productivities for a subset of the ÖWI.