Predictive capabilities of erosion models for different storm sizes

• Published in: Transactions of the ASAE Vol. 35; no. 2
• Main Authors: Bingner, R.L. (USDA, ARS, National Sedimentation Lab, Oxford, MS), Mutchler, C.K, Murphree, C.E
• Format: Journal Article
• Language: English
• Published: 01.03.1992
• Subjects: BASSIN VERSANT; CUENCAS HIDROGRAFICAS; EAU DE RUISSELLEMENT; EROSION; ESCORRENTIA; LLUVIA; MISSISSIPPI; MODELE DE SIMULATION; MODELOS DE SIMULACION; PLUIE; RENDEMENT; RENDIMIENTO; SEDIMENT; SEDIMENTO; TEMPETE; TERRE EN PENTE; TERRENO EN DECLIVE; TORMENTAS
• ISSN: 0001-2351; 2151-0059
• DOI: 10.13031/2013.28628

When conservation practices are evaluated for erosion control effectiveness, the erosion or sediment yield produced by all the various storms that occur throughout the year should he considered. The Universal Soil Loss Equation neglects storms less than 13 mm unless the storm includes excessive rainfall intensities. Other erosion models have been developed that can easily consider all sizes of storms. This study evaluates the capability of several erosion models (CREAMS, SWRRB, EPIC, ANSWERS, and AGNPS) to predict runoff and erosion from rainfall events of 0 to 13, 13 to 25, 25 to 50, 50 to 75, and greater than 75 mm. Runoff and erosion were simulated on three watersheds in Mississippi: a sloping upland watershed, a flatland watershed, and a terraced watershed with underground outlets. Results show that the models predicted a linear relation of runoff and erosion with rainfall. However, the slope of the model regressions varied widely from data values. Only CREAMS and SWRRB were successful in predicting annual average runoff and sediment yield within 20% of measured amounts from the upland watershed and only SWRRB and AGNPS were successful in prediction for the flatland watershed. None of the models predicted within 20% of measured amounts from the terraced watershed. In all cases, errors were greatest for the larger storms