Spatiotemporal analysis of rainfall erosivity in Oklahoma

The Universal Soil Loss Equation (USLE) and its family of models have been used for soil loss prediction and erosion mitigation. These empirical models relied on precipitation data predating 1957 to calculate the Rainfall Erosivity (R-factor) value; however, the isoerodent map published in AH703 is still widely used in soil loss estimations today. Climatic and precipitation changes have presented questions about the validity and reliability of using these estimation methods. Additionally, instrumentation, precipitation gauging networks, and data availability have improved since the original publication of the AH703 isoerodent map. This study conducted a spatiotemporal analysis in the GIS environment to estimate modern rainfall erosivity across Oklahoma using high-resolution rainfall data. Average annual and monthly rainfall erosivity factors, R-factor and R m -factor , respectively, were estimated using 5-min interval rainfall data collected from 111 Oklahoma Mesonet sites. The sites had an average historical precipitation record of 28 years. Using new rainfall erosivity values, spatial variation was assessed within two geographical segments: a) NOAA-defined state climate divisions and b) EPA-defined Level III ecoregions. Temporal analysis revealed that rainfall erosivity occurring between April and October contributed 86% of the annual R-factor. This study also developed an updated isoerodent map for the state of Oklahoma. The updated R-factor significantly differed from the original AH703 isoerodent map. Specifically, comparing the isoerodent maps revealed that the R-factor changed between −20% and 112%. The reasons contribute to the discrepancies between the two maps are also discussed. • We provide insights from detailed spatiotemporal analysis of rainfall erosivity in Oklahoma using data from 111 Mesonet sites. • A new isoerodent map for Oklahoma, called the Updated R-factor Map , has been developed using modern climate data. • The updated map shows differences of −20% to +112% versus the original AH703 map • The map leverages high-resolution precipitation data to improve soil loss estimates and guide targeted conservation strategies. • The methodology and dataset can be used to train machine-learning models to predict R factor in other regions.