The importance of climate and anthropogenic influence in precipitation partitioning in the contiguous United States

Understanding the process of precipitation partitioning into evapotranspiration and streamflow is fundamental for water resource planning. The Budyko framework has been widely used to evaluate the factors influencing this process. Still, its application has primarily focused on studying watersheds with minimal human influence and on a relatively small number of factors. Furthermore, there are discrepancies in the literature regarding the effects of climatic factors and land use changes on this process. To address these gaps, this study aims to quantify the influence of climate and anthropogenic activities on streamflow generation in the contiguous United States. To accomplish this, we calibrated an analytical form of the Budyko curve from 1990 to 2020 for 383 watersheds. We developed regional models of w, a free parameter introduced to account for controls of precipitation partitioning not captured in the original Budyko equation, within different climate zones. We computed 49 climatic and landscape factors that were related to w using correlation analysis and stepwise multiple linear regression. The findings of this study show that human activities explained a low variance of the spatial heterogeneity of w compared with the watershed slope and the synchronization between precipitation and potential evapotranspiration, nevertheless, urban development emerged as a factor in temperate climates, whereas irrigated agriculture emerged in cold climates. In arid climates, mean annual precipitation explains less than 20% of the spatial variability in mean annual streamflow; furthermore, this climate is the most responsive to changes in w. These results provide valuable insights into how land use and climate interact to impact streamflow generation differently in the contiguous United States contingent on the regional climate, explaining discrepancies in the literature.