Donkor et al. (2026) Modeling Spatiotemporal Streamflow Patterns in the Missouri River Basin Under Future Climate Scenarios

Identification

• Journal: Water
• Year: 2026
• Date: 2026-04-02
• Authors: Benjamin Donkor, Zhulu Lin, Siew Hoon Lim
• DOI: 10.3390/w18070858

Research Groups

• Department of Agricultural and Biosystems Engineering, North Dakota State University, Fargo, ND, USA
• Department of Agribusiness and Applied Economics, North Dakota State University, Fargo, ND, USA

Short Summary

This study utilized the Soil and Water Assessment Tool (SWAT) driven by five downscaled and bias-corrected CMIP6 global climate models to assess historical (2008–2024) and future (2025–2049) streamflow patterns in the Missouri River Basin, revealing spatially variable hydrological responses with reduced extreme events in the upper basin and increased extreme events in the lower basin.

Objective

• To evaluate spatiotemporal streamflow patterns in the Missouri River Basin under future climate scenarios to provide quantitative insights for regional water resource management.

Study Configuration

• Spatial Scale: Missouri River Basin (MoRB), approximately 1.37 million square kilometers, encompassing parts of ten U.S. states and one Canadian province.
• Temporal Scale:
  • Warm-up period: 2000–2007
  • Calibration period: 2008–2017
  • Validation period: 2018–2024
  • Historical analysis period: 2008–2024
  • Future projection period: 2025–2049

Methodology and Data

• Models used:
  • Hydrological model: Soil and Water Assessment Tool (SWAT)
  • Climate models: Five downscaled and bias-corrected CMIP6 Global Climate Models (GCMs) under the Shared Socioeconomic Pathway scenario (SSP2-4.5), including ACCESS-CM2, BCC-CSM2-MR, MPI-ESM1-2-HR, MRI-ESM2-0, and NorESM2-MM.
  • Software: QSWAT (QGIS plugin), RSWAT, MATLAB® (Version 2024b), R (Version 4.5.1).
  • Calibration algorithm: Sequential Uncertainty Fitting version 2 (SUFI-2).
  • Extreme value analysis: Generalized Extreme Value (GEV) method.
• Data sources:
  • Topography: 30-meter Digital Elevation Model (DEM) from the U.S. Geological Survey.
  • Land cover: 30-meter USGS National Land Cover Database (NLCD) 2016.
  • Soils: U.S. General Soil Map (STATSGO2) database.
  • Weather: 25-year NOAA daily precipitation and daily minimum and maximum air temperatures. Additional meteorological variables (relative humidity, solar radiation, wind speed) generated by SWAT's WGEN.
  • Reservoir data: Daily reservoir outflow records (2000–2024) from the U.S. Army Corps of Engineers (USACE) Omaha District for six major dams.
  • Observed streamflow: Daily streamflow (2000–2024) from the USGS National Water Information System (NWIS) at six gaging stations.
  • Climate projection data: High-resolution downscaled and bias-corrected CMIP6 hydro-climate projections for the Conterminous U.S. (CONUS) under the SSP2-4.5 scenario.

Main Results

• The SWAT model demonstrated satisfactory performance during calibration and validation, with Nash–Sutcliffe Efficiency (NSE) > 0.5, Kling–Gupta Efficiency (KGE) ≥ 0.5, Coefficient of Determination (R²) > 0.5, and Percentage Bias (PBIAS) within ±25% at most USGS gauge stations, particularly along the mainstem of the Missouri River.
• Future streamflow projections indicate spatially and temporally variable hydrological responses:
  • The upper basin (represented by Bismarck, North Dakota) is projected to experience lower flows across most percentiles and reduced extreme events. The 99th percentile discharge is projected to decrease by 25.5% (NorESM2-MM) to 45.1% (ensemble mean), and the 1st percentile discharge by 76.9% (NorESM2-MM) to 83.8% (ensemble mean).
  • The lower basin (represented by Hermann, Missouri) shows decreased median flows but higher extreme events. The 99th percentile discharge is projected to increase by 15.6% (NorESM2-MM) to 9.7% (ensemble mean), while the 1st percentile discharge is projected to decrease by 20.8% (NorESM2-MM) to 6.4% (ensemble mean).
• Recurrence interval analysis suggests that 100-year flows may decline by 11% at Bismarck and increase by 37.4% at Hermann.
• These spatial differences are primarily driven by variations in precipitation and temperature changes across the basin, with stronger warming in the upper basin increasing evaporative demand and reducing runoff efficiency, while the lower basin exhibits higher runoff efficiency and precipitation elasticity.

Contributions

• Provided a comprehensive spatial perspective by analyzing the entire Missouri River Basin, contrasting with previous studies that focused on localized areas within the large basin.
• Incorporated the latest CMIP6 climate projections into the hydrological analysis, offering an updated assessment of climate change impacts and, in some instances, presenting conclusions that differ from earlier studies.
• Combined hydrological modeling forced by climate projections with percentile-based streamflow analysis to account for various uncertainties inherent in climate change impact assessments.

Funding

• U.S. Department of Agriculture’s National Institute of Food and Agriculture (AFRI grant, award #: 2024-67023-42549)
• Hatch project (project #: ND01498)

Citation

@article{Donkor2026Modeling,
  author = {Donkor, Benjamin and Lin, Zhulu and Lim, Siew Hoon},
  title = {Modeling Spatiotemporal Streamflow Patterns in the Missouri River Basin Under Future Climate Scenarios},
  journal = {Water},
  year = {2026},
  doi = {10.3390/w18070858},
  url = {https://doi.org/10.3390/w18070858}
}