| Dynamic acidification models are used to predict future change in soil and surface water chemistry under various prescribed deposition scenarios. However, uncertainties in the application of dynamic acidification models, including the choice of the model (structure), input data, and model calibration, will influence prediction.;All models could be calibrated against observed data (stream and soil chemistry), but predictions and calibrated parameter space differed among models, suggesting that a range of equally reliable predictions coexisted. Nonetheless, the overall agreement of prediction patterns illustrated that all models are reliable and robust management tools with respect to acid emission control.;An ensemble of calibrated models predicted that future recovery at HBEF will be limited under currently legislated acid emission, but with further reductions in emission, a faster and greater recovery was predicted.;Keywords: dynamic acidification models, evaluation, uncertainty, calibration, Hubbard Brook, Monte Carlo simulation, scenario analysis.;In this thesis, uncertainties in predictions of four widely used dynamic acidification models (MAGIC, PnET-BGC, SAFE and VSD) were objectively evaluated using the long-term data record at the Hubbard Brook Experimental Forest (HBEF), New Hampshire, USA. The evaluation incorporated input data standardisation, Monte Carlo simulation, systematic calibration, and an ensemble model analysis, to ensure equal treatment of the models. |
