EVALUATING SIMPLIFICATIONS USED IN REGIONAL APPLICATIONS OF THE SAFE AND MAKEDEP MODELS

The importance of different additional simplifications used when movin g from single site applications to regional applications of the SAFE a nd MAKEDEP models was studied. The SAFE model is a dynamic soil chemis try model, which apart from physical and chemical soil parameters also needs the historic evolution of atmospheric deposition and nutrient u ptake to simulate the effects of acidic deposition on the chemical sta tus of forest soils. The historic evolution of atmospheric deposition and nutrient uptake are reconstructed by the MAKEDEP model, using info rmation on current deposition and nutrients status together with gener al deposition trends. Additional simplifications are enforced by limit ations in available regional data sets and the costs involved in compl eting regional data sets. The simplifications studied here are the sam e as those used in an actual regional application to Scania, southern Sweden: (1) Use of default values for certain soil parameters such as dissolved organic carbon (DOC) and soil bulk density; (2) Use of gener al depth trends to extrapolate soil parameters, such as mineralogy and surface area, measured for one soil layer to the other three soil lay ers considered in the SAFE model application; (3) Use of a simplified version of MAKEDEP. where some of the key feedbacks, such as the effec ts of a growing canopy on dry deposition rates have been removed in or der to limit the input data needed to run the MAKEDEP model. The model s where applied to two well documented sites, Selling, Germany and Rot hamsted, UK. Model results using the above simplifications and subsets thereof were compared with soil chemistry measurements and model resu lts based on previously published single-site applications for the two sites. The study shows that the trends and the present day values of pH and base cation concentration are reproduced Fairly well even when all of the simplifications are used. In order to increase the reliabil ity of the regional applications, however, the non-simplified version of MAKEDEP should be used rather than the simplified version. Regional data sets of standing biomass and nutrient contents are therefore nee ded. (C) 1998 Elsevier Science B.V. All rights reserved.