Accuracy assessments of gridded precipitation data: A case study from LTAR Texas Gulf

Abstract

Abstract Extreme meteorological events contribute to 80% of agroecosystem loss indemnities. Daily, seasonal, and annual precipitation dynamics also affect plant production, erosion, sediment and nutrient loading, and soil health across agroecosystems, largely depending on precipitation timing, amount, frequency, and intensity. Gridded meteorological data have been publicly available since the mid‐2000s and used to assess spatiotemporal precipitation dynamics. The accuracy of gridded compared to site‐level precipitation is, however, rarely evaluated because most long‐term meteorological data are already calculated into gridded databases. At the Long‐Term Agroecosystem Network Texas Gulf site in central Texas, however, a nearly 90‐year precipitation record exists from a suite of 15 on‐site rain gauges that have not been used in gridded dataset development. The objective of this study was to evaluate the accuracy of historical precipitation acquired from 15 on‐site monitoring stations with the common gridded databases of the Parameter‐Elevation Regressions on Independent Slopes Model (PRISM), the Daily Surface Weather and Climatological Summaries (DayMET), and the Gridded Surface Meteorological Dataset (GridMET). Our findings suggest precipitation is highly variable spatiotemporally. Annual and seasonal gridded data from all sources were significantly correlated with on‐site weather station precipitation, but GridMET produced the strongest correlations with on‐site data. Across time, the accuracy of gridded precipitation data improved, especially between 1980 and 2000 decades. Extreme daily precipitation events acquired from gridded data sources, however, were poorly correlated with actual precipitation at rain gauge sites. These results suggest that gridded data can be helpful for long‐term management planning but also showcase a limited utility of gridded data for monitoring assessments, especially as they relate extreme precipitation to erosion, crop loss, and insurance indemnities.