Benchmarking reliable erosion indices from quarter-hour station data for climate studies in the southeastern United States

Abstract

Precipitation is one of many aspects of our changing climate that is under investigation by the scientific community. Although climate scientists generally predict that the amount of precipitation has and will continue to increase for the United States, the characteristics of that changing precipitation are not well understood. Some have reported significant observed and projected changes in precipitation characteristics, but existing studies have not adequately addressed this issue. The precipitation characteristics particularly vulnerable to change are energy, intensity, duration, and frequency, which are responsible for causing changes in the erosivity of rainfall. Furthermore, there are published discrepancies in observed erosivity. Hence, the goal of this study was to update the erosion indices (EI), specifically those in the southeastern United States, for two reasons. First, it is a region highly likely to be influenced by climate change and climate variability, and second, it is the setting of published discrepancies mentioned above. Observed quarter-hour precipitation data from 616 National Oceanic and Atmospheric Administration (NOAA) National Centers for Environmental Information (NCEI) land-based stations (DSI-3260) for 11 states in the Southeast were used to calculate EI values for each station. The data were screened (to reduce gaps in the data) before calculation, resulting in 172 gauge locations to represent the 11 states. A water balance was performed to validate the observed data set and to select the best of seven station screening methods. Annual and single-storm EI statistics were calculated for stations passing the preferred screening method using 15-minute precipitation data. Since fixed-interval data slightly underestimate the maximum 30-minute intensity of a storm, an adjustment to annual EI calculations was suggested to account for this underestimation. Annual EI estimations for the Southeast were found to be significantly larger than Wischmeier and Smith (1978) and Renard et al. (1997), confirming the findings of McGregor et al. (1995) while significantly increasing the spatial resolution of EI observations in the Southeast. The regional increase on average was 18.6% over previous EI values from Renard et al. (1997), with a range of −19.5% to +57.5%. There was no consistent significant change in the single-storm EI for the study area, with an average decrease of 2.3% for the 10-year storm EI, although regional results varied. A comparison of the 10-year storm EI showed significant changes in some inland areas mostly arising from an unorthodox contouring method used in Renard et al. (1997), which was also pointed out by Hollinger et al. (2002). Also, the Atlantic Coast may have experienced lower single-storm EI for the period observed due to a negative phase of the Atlantic Multidecadal Oscillation (AMO) compared to observations during a strong-positive AMO phase for previous EI estimations. This study presents a novel, simplified methodology for estimating EI from more reliable precipitation data as well as a benchmark for climate impact studies related to erosivity for conservation efforts in the Southeast.