Wang et al. (2025) Disentangling climate and policy uncertainties for the Colorado River post-2026 operations

Identification

• Journal: Nature Communications
• Year: 2025
• Date: 2025-09-29
• Authors: Bowen Wang, Benjamin Bass, Alex Hall, Stefan Rahimi, Lei Huang
• DOI: 10.1038/s41467-025-63635-4

Research Groups

• Department of Atmospheric and Oceanic Sciences, University of California, Los Angeles, USA
• Department of Civil and Environmental Engineering, Massachusetts Institute of Technology, USA
• Department of Atmospheric Science, University of Wyoming, USA

Short Summary

This study evaluates the future conditions of Lakes Powell and Mead in the Colorado River Basin under climate change and various water management policies using dynamically downscaled CMIP6 climate models and unique uncertainty methods. It finds that existing policies lead to high risks of dead pool conditions for both reservoirs by 2060, while proposed alternative policies reduce but do not eliminate these risks, necessitating larger reductions for long-term sustainability.

Objective

• To evaluate the future conditions and risks to Lakes Powell and Mead, and their associated water supply and hydropower generation, under projected climate change and existing versus alternative water management policies for post-2026 operations.

Study Configuration

• Spatial Scale: Colorado River Basin (CRB), Upper Colorado River Basin (UCRB), Lakes Powell and Mead.
• Temporal Scale: 2027–2060 (near-term planning period), 2060–2100 (longer-term perspective), 21st century. Simulations are conducted at an annual time step.

Methodology and Data

• Models used:
  • Water Budget Model (WBM): A custom, rapid, and flexible model simulating reservoir operations at an annual time step.
  • Noah-Multiparameterization (Noah-MP) land surface model: Used to simulate streamflow.
  • Weather Research and Forecasting (WRF) model: For dynamical downscaling of climate data.
  • 10 CMIP6 Global Climate Models (GCMs): Under the SSP3-7.0 emission scenario.
  • Monte Carlo simulation approach: To probabilistically sample and superimpose internal climate variability.
  • Decision Making under Deep Uncertainty (DMDU) approach: To remove internal variability and evaluate equilibrium states.
  • Autoregressive model of order 1 (AR(1)) and Pearson-III distribution: Used for generating streamflow anomalies in Monte Carlo simulations.
• Data sources:
  • Dynamically downscaled and bias-corrected regional climate simulations (from CMIP6 GCMs).
  • ERA5 (ECMWF Reanalysis v5): Used to force WRF for calibration of Noah-MP.
  • Observed natural streamflow data for the Colorado Basin (USBR).
  • Historical storage, regulated inflow, release, and evaporation data for Lakes Powell and Mead (USBR).
  • Observed streamflow data at the Colorado River above Diamond Creek gauge (USGS site number 09404200).
  • Hydropower generation data (EIA).
  • Upper Colorado River Commission (UCRC) projections for Upper Basin (UB) water demand.
  • Water Accounting Reports for Lower Basin (LB) consumptive use.

Main Results

• All 10 dynamically downscaled GCMs project significant temperature increases in the UCRB (ensemble mean of 0.57 ± 0.10 °C per decade).
• The GCM ensemble-mean streamflow trend since 1984 shows a decrease of –2.6 ± 1.2% per decade, leading to a projected 20.0% (± 9.3%) decrease in streamflow by 2060 and 30.5% (± 14.1%) by 2100.
• Under the existing policy (DCP2019), Lakes Powell and Mead face substantial risks of reaching dead pool conditions at least once before 2060 (85% and 83% likelihood, respectively).
• Alternative policies (DCP2019+, LB2024, UB2024) considerably reduce risks at Lake Mead (e.g., LB2024 yields a 17% cumulative risk of dead pool by 2060), but Lake Powell remains highly vulnerable.
• LB2024, the only policy requiring reductions to UB depletions, yields the lowest risks at Mead and the highest hydropower output, but relegates higher risks to Powell.
• All policies exhibit "tipping points" where reservoir storage can rapidly decline from stable to critical levels (inactive or dead pool) with modest decreases in natural flow.
• The equilibrium reservoir storage under some policies (LB2024, UB2024) does not always change monotonically with natural flow, revealing unintended incoherences where individual reservoir storage can fluctuate despite declining combined storage.
• The existing policy (DCP2019) provides the greatest total water delivery in the 2027–2060 period but at the cost of higher dead pool risks and lower hydropower generation, with its water supply advantage diminishing by the end of the century.

Contributions

• First study to utilize 10 dynamically downscaled CMIP6 GCMs for comprehensive streamflow projections in the Colorado River Basin.
• Introduces unique methodologies that both probabilistically sample internal climate variability (Monte Carlo) and isolate policy effects by removing internal variability (DMDU approach), providing complementary insights.
• Offers a timely evaluation of recent, formally submitted policy proposals (DCP2019+, LB2024, UB2024) for post-2026 Colorado River operations.
• Identifies and elucidates "tipping points" and non-monotonic responses in reservoir equilibrium states under different policies, highlighting potential unintended consequences and incoherences in policy design.
• Provides novel scientific and policy-oriented information for sustainable water management in the CRB, offering insights not previously utilized by the Bureau of Reclamation.

Funding

• Department of Commerce NOAA project “Towards Predicting Drought and Subsequent Water Resource Challenges at Landscape-resolving Scales Across the Western U.S.” (award no. NA23OAR4310633)
• Department of Energy HyperFACETS project (award no. DE-SC0016605)
• State of California support under California Energy Commission project “Development of Climate Projections for California and Identification of Priority Projections” (award no. EPC-20-006)
• NCAR Computational and Information Systems Laboratory (CISL) for computational support

Citation

@article{Wang2025Disentangling,
  author = {Wang, Bowen and Bass, Benjamin and Hall, Alex and Rahimi, Stefan and Huang, Lei},
  title = {Disentangling climate and policy uncertainties for the Colorado River post-2026 operations},
  journal = {Nature Communications},
  year = {2025},
  doi = {10.1038/s41467-025-63635-4},
  url = {https://doi.org/10.1038/s41467-025-63635-4}
}