Evaluation of an acidification model with data from manipulated catchments in Norway

Abstract

INTERNATIONAL efforts to control the emissions of the acidifying compounds SO₂ and NO_x to the atmosphere, and to establish critical loads, are in part aimed at predicting the long-term response of soils and waters to changes in acid deposition. Estimation of future long-term trends in acidification requires the use of models, the strict verification of which requires years or decades to determine whether predictions match observed responses. Experimental manipulations with whole ecosystems provide data with which predictive models can be evaluated. The MAGIC model^1,2 is a widely used, intermediate-complexity, process-oriented model for soil and water acidification in catchments. Here we follow up a preliminary application³of the model to the first two years of data from the manipulated catchments of the RAIN (Reversible Acidification In Norway) project⁴. We use a more extensive four-year data set as inputs to a new version of the model. The four-year record provides a much more rigorous test of the predictive ability of the model because these data show major changes in runoff chemistry due to the experimental change in deposition. Sensitivity analysis shows that MAGIC is successful at predicting future acidification.