Evaluation of Four Parameterization Strategies for the APEX Model

Abstract

Abstract.State-of-the-art model parameterization consists of calibrating and validating the model using monitored data. When data are not available, many studies have relied on alternate strategies, including parameter sets obtained for eco-hydrologically similar watersheds and best professional judgment (BPJ). The objectives of this study were to (1) test the accuracy of four APEX parameterization strategies to predict runoff quantity and quality in a small agricultural watershed and (2) compare the effects of the different parameter sets on relative and absolute water quality outcomes for different conservation practice scenarios. A BPJ and three parameter sets obtained on nearby and more distant sites, including two fully calibrated parameter sets, i.e., for runoff, sediment, and total phosphorus (TP), and one partially calibrated parameter set, i.e., for runoff and TP only, were evaluated based on edge-of-field runoff, sediment loss, and TP loss, as well as for long-term annual predictions and relative changes for six conservation practice scenarios. Only the parameter sets that were fully calibrated met the model performance criteria on the test watershed and produced similar 30-year average annual predictions for the conservation practices. Relative changes in conservation outcomes were similar for the fully and partially calibrated parameter sets. They were different for the BPJ parameter set for at least one conservation practice. In the absence of site-specific data, the best parameterization strategy was to use parameter sets from a model calibrated for runoff, sediment, and nutrient losses from an eco-hydrologically similar site. Partial calibration of the model was sufficient to obtain consistent relative effects of the conservation practices. Keywords: Conservation practices, Hydrologic modeling, Model parameterization, Phosphorus, Sediment.