INVESTIGATING THE VALUE OF HYDROLOGIC MODEL COMPLEXITY AND SPATIAL INFORMATION USING MULTICRITERIA METHODS

摘要

The focus of this study is to assess the performance improvements, within a multicriteria framework, of lumped and semi-distributed applications of a hydrologic model to a high-altitude, snow-dominated, watershed in the Sierra Nevada near Lake Tahoe, Nevada, USA. Specifically, performance comparisons are made to evaluate the benefit of "distributing" different estimates of model input precipitation and model parameters in terms of different measures of streamflow simulation error (e.g., absolute percent bias and Nash Sutcliff measure of efficiency) during different periods of watershed behavior (e.g., spring runoff, summer baseflow, and fall-winter low flows). The initial results demonstrate the value of the approach in identifying specific hydrologic model improvements associated with both the conceptual representation of the watershed and different distributions of precipitation as model input.