<i>Evaluation of Different Precipitation Inputs to WEPP</i>

Abstract. Almost 20 years of observed 1-minute weather data from the most complete and longest operational NOAA Automated Surface Observation Station (ASOS) network station were used to prepare seven climate inputs (the raw, unadjusted, quality checked, and gap-filled gauge data and six temporal aggregations of the same data) to the Water Erosion Prediction Project (WEPP) model. The nearest stations from both the old (1995) and new (2015) CLIGEN databases were also used to compare climate inputs and corresponding soil loss predictions from WEPP as it would typically be applied in the field for research or soil conservation planning. The results showed that WEPP simulations driven by CLIGEN closely approximated the vigor (erosivity) of breakpoint quality climate inputs despite using only 15-minute precipitation data for CLIGEN input file parameterization. Erosion prediction models and corresponding theory were developed initially based on breakpoint rainfall data, so climate inputs that more closely approximate that kind of data should be considered ideal for soil erosion modeling efforts. These results call into question the use of coarser resolution climate inputs such as 15-minute precipitation data, which are most common, without applying sufficient adjustments and/or corrections to better approximate breakpoint data. Our findings also suggest that intensity corrections of 4% used by Hollinger et al. (2002), USDA-ARS (2008, 2013), and McGehee and Srivastava (2018) for the Revised Universal Soil Loss Equation (RUSLE2) erosivity inputs may not be sufficient for at least the climate near these stations in the Northeastern United States. A national study of precipitation intensity and erosivity based on high resolution climate inputs like those available from the ASOS network may be warranted.