Evaluation of Runoff and Erosion Models

Abstract

Runoff and sediment yield for 30 runoff events on three experimental watersheds are calculated using the agricultural non‐point‐source pollution (AGNPS), areal non‐point‐source watershed environmental response simulation (ANSWERS), and chemicals runoff and erosion from agricultural management systems (CREAMS) runoff‐erosion models. The results are compared with measured runoff and sediment yield. The computed and measured runoffs show reasonable to poor agreement. The average ratios of computed to measured sediment yields for the various storms and watersheds show a large scatter. ANSWERS provides the most consistent results for estimates of runoff and sediment yield. All three models tend to underestimate sediment yield for large storms. For high intensity and low intensity storms on two small watersheds, the detachment models in ANSWERS and CREAMS have biases (ratio of calculated to measured sediment yields) that range between 0.9–1.0 and 0.4–1.6, respectively.