Prange et al. (2025) Elucidating the loose tie between precipitation and streamflow sensitivities to warming across the contiguous United States

Identification

• Journal: npj Climate and Atmospheric Science
• Year: 2025
• Date: 2025-11-20
• Authors: Marc Prange, Ming Zhao, Elena Shevliakova, Sergey Malyshev
• DOI: 10.1038/s41612-025-01257-9

Research Groups

• Atmospheric and Oceanic Sciences, Princeton University, Princeton, NJ, USA
• Geophysical Fluid Dynamics Laboratory (GFDL), NOAA, Princeton, NJ, USA

Short Summary

This study uses a moderately high-resolution global climate model in a counter-factual warming scenario to elucidate the hydrological processes driving regional streamflow sensitivities to warming across the contiguous United States. It finds that while the West Coast and eastern US experience increased high-flows driven by atmospheric rivers, the mountainous western US sees dwindling streamflows due to snow loss fueling evapotranspiration at double the rate of precipitation changes.

Objective

• To elucidate the hydrological processes driving regionally varying mean and extreme streamflow sensitivities to warming across the contiguous United States (CONUS) in a counter-factual warming scenario.
• To isolate the thermodynamic component of the precipitation response to warming by nudging atmospheric winds, and attribute changes in river floods and their drivers to atmospheric river (AR) precipitation, non-AR precipitation, and snowmelt.

Study Configuration

• Spatial Scale: Contiguous United States (CONUS), analyzed at a model resolution of approximately 50 kilometers, with results aggregated over HUC4 hydrological sub-regions.
• Temporal Scale: Simulations cover the historical period from 1951 to 2020. Analysis focuses on daily, monthly, seasonal, and annual mean changes, with high-flow events defined by the 99.7th percentile of daily streamflow (annual recurrence rate).

Methodology and Data

• Models used:
  • GFDL’s global coupled atmosphere-land model AM4.0/LM4.0 (approximately 50 km resolution).
  • LM4.0 includes a grid-scale river network.
  • Atmospheric river (AR) detection using the Guan and Waliser (2015) algorithm, validated with the Mundhenk et al. (2016) algorithm.
  • Potential evapotranspiration (PET) calculated using the Penman–Monteith Equation over an open water surface.
  • Thornthwaite-style water balance model for observational comparison.
• Data sources:
  • Observed daily sea surface temperatures (SSTs) for the control simulation (CMIP6 HighResMIP protocols).
  • NCEP reanalysis data (6-hourly) for nudging atmospheric winds in both control and warming simulations.
  • Observationally deduced 1 km resolution water balance dataset (1980–present) for CONUS, forced by Daymet Daily data version 3 (daily precipitation and temperature).

Main Results

• Precipitation Response: Mean precipitation increases by 3.6% K⁻¹ across CONUS (consistent with observed scaling), while extreme precipitation (99.7th percentile) increases by 5.4% K⁻¹, often exceeding the Clausius-Clapeyron scaling (7% K⁻¹). Mean AR precipitation increases by 4.2% K⁻¹.
• Streamflow Response:
  • Mountainous Western US: Mean and extreme streamflows decline (up to -10% K⁻¹ for extremes) despite significant precipitation increases, particularly from ARs. This is attributed to snow loss fueling evapotranspiration (ET) at double the rate of precipitation changes.
  • Eastern US: Streamflows increase significantly (up to 10% K⁻¹ for extremes), exceeding precipitation changes.
  • West Coast: Streamflows increase positively (up to 4% K⁻¹ for extremes) but at a lower magnitude than CONUS-averaged precipitation changes.
• Surface Water Balance and ET Constraints:
  • Mountainous Western US: Experiences a substantial increase in ET relative to precipitation, balanced by a decrease in runoff. The winter-time mountainous west shifts from a water-limited to an energy-limited ET regime due to enhanced liquid water supply (liquid precipitation and melt) outweighing increased energy supply from reduced snow cover.
  • Eastern US: ET/P decreases while runoff/precipitation increases.
  • West Coast: Exhibits only slightly positive changes in ET/P.
• High-Flow Drivers and Seasonality:
  • West Coast and Eastern US: Liquid precipitation, predominantly from ARs, is the main driver of high-flows. High-flow frequency increases with warming in these regions, with the West Coast showing a more pronounced winter seasonality and the Eastern US a weakened seasonality with increases throughout the year.
  • Mountainous Western US: Melt is the exclusive driver of high-flows in the present climate. A strong reduction in melt-induced high-flows (exceeding -30% K⁻¹ regionally) occurs with warming, leading to an annual mean runoff loss of -4.1% K⁻¹.
• Antecedent Soil Moisture: Over the mountainous western US, the loss of upstream water inputs (especially snowmelt) outweighs increased antecedent soil saturation, leading to fewer high-flows. In contrast, increased high-flows over the eastern US are driven by a shift towards higher antecedent soil saturation and more extreme upstream water inputs.

Contributions

• Provides a process-based understanding of the regionally varying mean and extreme streamflow sensitivities to warming across CONUS using a moderately high-resolution global coupled atmosphere-land model.
• Isolates the thermodynamic component of the warming response by tightly nudging atmospheric winds, allowing for a focused analysis on the most robust and well-understood part of the precipitation response.
• Attributes changes in high-flow frequency to specific upstream water inputs (AR precipitation, non-AR precipitation, and snowmelt), seasonal energy and water constraints on evapotranspiration, and antecedent soil moisture conditions.
• Highlights the critical role of atmospheric rivers as the strongest driver of high-flows across CONUS, not just the West Coast, and their sensitive response to warming.
• Demonstrates how complex, regionally confined hydrological processes (e.g., snow dominance, ET limitations) lead to diverse streamflow responses even under a comparatively homogeneous precipitation response to warming.

Funding

• National Oceanic and Atmospheric Administration (NOAA), U.S. Department of Commerce, under awards NA22OAR4050663D and NA23OAR4050431I.

Citation

@article{Prange2025Elucidating,
  author = {Prange, Marc and Zhao, Ming and Shevliakova, Elena and Malyshev, Sergey},
  title = {Elucidating the loose tie between precipitation and streamflow sensitivities to warming across the contiguous United States},
  journal = {npj Climate and Atmospheric Science},
  year = {2025},
  doi = {10.1038/s41612-025-01257-9},
  url = {https://doi.org/10.1038/s41612-025-01257-9}
}