Predicting salinity and alkalinity fluxes of U.S. freshwater in a changing climate: Integrating anthropogenic and natural influences using data-driven models

Abstract

Climate change is an ongoing and intensifying threat. Previous studies indicate that U.S. rivers are undergoing salinization and alkalinization driven by both natural (e.g., temperature and precipitation) and anthropogenic (e.g., population density) factors. In this study, random forest models were developed to predict how the salinity (i.e., sodium) and alkalinity fluxes from 226 U.S. rivers will vary with changing population density and climatic forcings (i.e., temperature and precipitation) from 2040 to 2100 for three socioeconomic development pathways. The models predicted a lower future sodium flux in the northern U.S., likely due to reduced winter road salting under projected warmer winter. In the southern and western U.S., where road salting is uncommon, the models predicted little or no change in future sodium flux, however, a projected warmer and drier climate might exacerbate soil salinization in these regions. The models also indicated that carbonate weathering rates are inhibited when temperatures exceed 10 °C, leading to a lower future alkalinity flux in carbonate-rich watersheds at high temperatures. In siliciclastic-dominated or organic carbon-rich watersheds, rising temperatures are associated with increased riverine alkalinity flux, likely through the acceleration of silicate weathering and decomposition of soil organic carbon. Higher precipitation and enhanced transport capacity were generally deemed to contribute to higher solute fluxes before reaching a plateau. This study underscores the urgency for policymakers and scientists to adapt strategies for managing rivers, focusing on mitigating the impacts of river salinization and shifts in riverine alkalinity driven by global warming.