Rivers in many parts of the world are carrying less water even as they receive more rain and snow, according to a new study that challenges a basic assumption about how climate change affects freshwater supplies. Researchers at Pennsylvania State University analyzed precipitation and river flow records from more than 10,000 rivers between 1951 and 2020 and found that in nearly one-third of them, flow trends did not match the changes observed in rainfall and snow.
The finding complicates a straightforward equation that has long guided water managers: more precipitation should mean more water in rivers. Global heating is intensifying the hydrological cycle, making wet regions wetter and dry regions drier, but the study shows that the effect on river flow is not uniform and cannot be predicted from rainfall records alone.
In some rivers, the mismatch reflects the growing share of precipitation falling as rain rather than snow. Snowpack acts as a natural reservoir, releasing water gradually through the spring and summer. When snow turns to rain, that water runs off quickly instead of being stored, leaving less available during dry months. In other cases, warmer temperatures increase evaporation from soils and vegetation, so more of the rain that falls never reaches the river channel.
Land use and water withdrawals may also play a role. Dams, irrigation, urban development, and groundwater pumping can alter flow independently of precipitation trends, and the study's broad geographic scope suggests that such factors are widespread enough to register in long-term records. The researchers did not attribute every mismatch to a single cause, but the pattern indicates that climate-driven changes to the water cycle are more complex than a simple shift in rainfall totals.
The implications for water management are significant. Utilities, farmers, and industries that rely on river flow often plan using historical precipitation trends as a proxy for future supply. If that proxy fails in a substantial fraction of rivers, those plans may overestimate available water. The study suggests that monitoring networks should track river flow directly rather than inferring it from rainfall, and that water managers should account for the possibility that a wetter future does not guarantee fuller rivers.
The research adds to a growing body of evidence that the hydrological cycle is being reshaped by warming in ways that vary by region and season. Previous work has documented earlier snowmelt, more intense but shorter rain events, and declining groundwater levels in many aquifers. The new analysis extends that picture to river flow on a global scale, using records that span seven decades and thousands of catchments.
For scientists, the results point to the need for better integration of precipitation, temperature, evaporation, and land-use data when projecting future water availability. For policymakers, they underscore the value of investing in direct streamflow monitoring and in infrastructure that can buffer against both floods and droughts. The study does not predict which rivers will decline next, but it shows that the relationship between rain and river water is not as simple as it once seemed.
13
