Wilbanks et al. (2026) Evaluation of streamflow trends and the drivers of long-term change for 33 river basins in the southeastern U.S.

Identification

• Journal: Limnologica
• Year: 2026
• Date: 2026-01-28
• Authors: Kelsey A. Wilbanks, Darold P. Batzer
• DOI: 10.1016/j.limno.2026.126311

Research Groups

• Department of Entomology, College of Agriculture and Environmental Science, University of Georgia, Athens, GA, United States

Short Summary

This study assessed long-term (1957–2022) streamflow trends and their drivers across 33 river basins in the southeastern U.S., revealing widespread declines in mean annual streamflows, with spatial variability primarily linked to precipitation patterns and drainage area, while regional temperature and population increases were not directly correlated.

Objective

• To assess long-term mean, minimum, and maximum annual streamflow trends (1957–2022) for 33 river basins in the South Atlantic-Gulf Drainage, southeastern United States.
• To understand the spatial variability of streamflow changes by examining relationships with static spatial indicators (drainage area, elevation, ecoregions).
• To identify the drivers of streamflow changes by assessing long-term trends in total precipitation, maximum and minimum temperatures, population, and groundwater level.
• Hypotheses: (H1) Streamflow patterns would show regional drying, exacerbated from northern to southern basins, with minimum streamflows showing higher drying than maximum and mean. (H2) Drivers would be consistent with both climate change and anthropogenic factors, with anthropogenic and climate factors outweighing or exacerbating precipitation changes. (H3) Geomorphic features (larger drainage area, greater slope) would predict streamflow drying.

Study Configuration

• Spatial Scale: South Atlantic-Gulf Drainage (03 Hydrologic Unit Code, HUC), southeastern U.S., covering 367,740 km². The study focused on 33 6-digit HUC basins across Florida, South Carolina, and parts of Alabama, Georgia, Louisiana, Mississippi, North Carolina, Tennessee, and Virginia.
• Temporal Scale: 65-year period, from 1957 to 2022.

Methodology and Data

• Models used:
  • Mann-Kendall trend test: To statistically detect monotonic trends in streamflow, precipitation, temperature, population, and groundwater level.
  • Thiel-Sen estimator (Sen’s slope): To quantify the magnitude and direction of linear rate of change over time for all variables.
  • Regression analysis: To determine relationships between mean annual streamflow changes and potential correlative metrics.
  • Akaike’s Information Criterion (AIC) model selection: To identify the best-fit models for predicting mean annual streamflow changes.
• Data sources:
  • United States Geological Survey (USGS): Daily mean streamflow (m³/s) from 374 gages (minimum 50 years of data), daily groundwater level (depth in metres below surface) from 376 wells (minimum 30 years of data), drainage area (km²), watershed HUC, GPS coordinates, Watershed Boundary Dataset, and 1-meter digital elevation model.
  • National Oceanic and Atmospheric Administration (NOAA) National Centers for Environmental Information (GHCNd): Daily total precipitation (mm), total snowfall (mm), maximum and minimum temperatures (°C) from 463 precipitation stations, 357 maximum temperature stations, and 355 minimum temperature stations (minimum 50 years of data).
  • U.S. Census Bureau: Decennial (1950–2020) population estimates for counties, aggregated to estimate mean population density (# individuals/km²).
  • Environmental Protection Agency (EPA): Level II ecoregions (Mississippi Alluvial and Southeast USA Coastal Plains, Ozark, Ouachita-Appalachian Forests, Southeastern USA Plains, Everglades).

Main Results

• Overall Streamflow Trends (1957–2022):
  • Mean annual streamflows declined at 56.1% of gages, with 15.5% showing significant decreases. The average Sen’s slope across all basins was –0.06 ± 0.61 (m³/s)/year.
  • Minimum annual streamflows showed significant declines at 34.0% of gages, with an average Sen’s slope of –0.03 ± 0.19 (m³/s)/year.
  • Maximum annual streamflows declined at 56.4% of gages, with a larger magnitude of decline averaging –0.30 ± 1.7 (m³/s)/year.
  • Spatially, 69.7% of basins exhibited decreasing mean annual streamflows, primarily in the central region. Increases were observed in the western and southern regions (24.2% of basins) and the Roanoke Drainage Basin (northern).
• Factors Linked to Streamflow Changes:
  • Total annual precipitation was positively correlated with streamflow changes (ß = 41.6, 95% CI: 10.4 – 72.9), explaining 17% of streamflow variation. Spatial patterns of precipitation changes mirrored those of streamflows.
  • Drainage area was associated with mean annual streamflow changes (ß = 8.5, 95% CI: –31.6 – 48.5), explaining 14% of streamflow variation, with larger watersheds tending to show greater declines.
  • AIC model selection indicated that a combined model of total precipitation, groundwater level, and drainage area was the strongest predictor of mean annual streamflow change (AIC = –44.11), accounting for 100% of the total explanatory predictive power.
• Factors Not Significantly Linked (at regional scale):
  • Maximum and minimum annual temperatures increased across the region (average 0.06 ± 0.96 °C/year and 0.04 ± 1.14 °C/year, respectively) but were not directly related to regional mean annual streamflow patterns.
  • Population density increased across all basins (average 15.1 ± 19.9 (# individuals/km²)/decade) but was not directly related to regional mean annual streamflow patterns.
  • Elevation was not related to changing mean annual streamflows.
  • Groundwater levels showed declines (average –0.03 ± < 0.01 m/year) but did not show consistent patterns or significant relationships with mean annual streamflow variation in regression analysis, possibly due to limited well coverage.

Contributions

• Provides a comprehensive, long-term (65-year) evaluation of streamflow trends and their drivers across 33 river basins in the South Atlantic-Gulf Drainage, southeastern U.S., highlighting significant spatial variability.
• Quantifies the relative contributions of climate (precipitation) and static spatial indicators (drainage area) as primary correlative factors for regional streamflow changes.
• Demonstrates that while temperature and population increases are widespread, their direct regional correlation with streamflow changes is not significant, suggesting complex interactions or localized impacts.
• Identifies a robust predictive model for future streamflow changes based on a combination of precipitation, groundwater level, and drainage area, offering valuable insights for water resource management and ecological conservation.

Funding

• Georgia Environmental Protection Division, Regional Water Plan SEED Grant [Contract Number: 761–200119]
• U.S. Department of Agriculture Hatch Program

Citation

@article{Wilbanks2026Evaluation,
  author = {Wilbanks, Kelsey A. and Batzer, Darold P.},
  title = {Evaluation of streamflow trends and the drivers of long-term change for 33 river basins in the southeastern U.S.},
  journal = {Limnologica},
  year = {2026},
  doi = {10.1016/j.limno.2026.126311},
  url = {https://doi.org/10.1016/j.limno.2026.126311}
}