Methodology for Spatially Distributed Rainfall Erosivity Calculations at the Conterminous United States to Support Soil Erosion Studies

Research and action agencies in the US work collaboratively to develop and use soil erosion technology to support the development of field-specific conservation plans. These tools and accompanying databases are applied in all counties throughout the country covering a wide range of natural and anthropogenic physical conditions. Climate, particularly precipitation, constitutes one of the key drivers directly related to soil detachment and transport. Observations spanning over 30 years have demonstrated that estimated long-term average annual soil loss in agricultural fields is the result of the cumulative effect of many small and moderate-sized storms along with the impact of occasional severe ones. In the Revised Universal Soil Loss Equation version 2 (RUSLE2) technology, the effect of rainfall is represented by the rainfall runoff erosivity index R. This index is designed to serve as an estimation of the potential storm energy specific to each location. In this study, we propose and evaluate a methodology to generate continuous surfaces of monthly R for the continental US from discrete 15-min precipitation data. Over 2000 stations covering more than 50 years of 15-min precipitation data were used. Storm identification algorithms were implemented and evaluated through comparison with existing tools. Outlier events were identified and removed using a 50-year recurrent interval calculated for each station. Using 30-years of recorded data, a custom universal kriging algorithm was employed to generate a smooth continuous surface as a raster grid. This step included a boxcox transformation of the station data, directional variogram fitting, and the removing of external trends using elevation, long-term annual precipitation totals, and distance to the coast. Predicted surfaces were compared with existing RUSLE2 surfaces for the same time period with great level of agreement. The proposed methodology is intended to be comprehensive and reproductible such that it can serve as a template for future updates of erosivity maps for the entire continent at a county-scale. This methodology provides the means for future systematic updates to the RUSLE2 climate database to account for climatic changes and to support continued national efforts in reducing soil erosion and conserving natural resources.