Applicability of Machine Learning tool for the 2020-2099 forecasting of reservoir capacity

Reservoir storage is crucial for managing water seasonal fluctuations and climate change impacts on U.S. water supplies. Due to their strategic location and the continuous loss of their storage, current reservoirs are considered a non-renewable resource for the country. Previous efforts developed a supervised machine learning (ML) tool that estimated reservoir capacity losses from reservoir design information and watershed characteristics. The objective of this study is to evaluate the capability of this ML tool to forecast reservoir capacity under future climate scenarios. The evaluation was run for three trial reservoirs of different initial capacities and basin characteristics. Results showed ML tool results aligned with extrapolating results, a common practice when projecting reservoirs storage, with the addition of confidence intervals that would incorporate the uncertainty of future climate conditions. Results also exhibited the possibility of implementing the ML tool for forecasting storage of reservoirs with limited surveys. Time step for capacity loss estimation was found to be a critical indirect parameter of the ML tool, with larger basins requiring larger time step for the computations. Projected reservoir capacity losses showed a higher temporal correlation with maximum precipitation values for small basins. Projected percentages of reservoirs’ capacity loss per year with respect to initial capacity were larger for the trial basin with largest maximum rainfall events, and steepest channel slope regardless of the average precipitation or the basin area.

Authors

Institutions

Publication Details

Journal
EarthArXiv (California Digital Library)
Published
2026-11-21
DOI
https://doi.org/10.31223/x5zn6x
Primary Topic
Water resources management and optimization
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Applicability of Machine Learning tool for the 2020-2099 forecasting of reservoir capacity

Travis A. Dahl, İbrahim Demir, Marian Muste, Alejandra Botero‐Acosta et al.
EarthArXiv (California Digital Library)
Water resources management and optimization
preprint

Applicability of Machine Learning tool for the 2020-2099 forecasting of reservoir capacity

Travis A. Dahl, İbrahim Demir, Marian Muste, Alejandra Botero‐Acosta, Paul Boyd, Amanda L. Cox
preprint en

Abstract

Reservoir storage is crucial for managing water seasonal fluctuations and climate change impacts on U.S. water supplies. Due to their strategic location and the continuous loss of their storage, current reservoirs are considered a non-renewable resource for the country. Previous efforts developed a supervised machine learning (ML) tool that estimated reservoir capacity losses from reservoir design information and watershed characteristics. The objective of this study is to evaluate the capability of this ML tool to forecast reservoir capacity under future climate scenarios. The evaluation was run for three trial reservoirs of different initial capacities and basin characteristics. Results showed ML tool results aligned with extrapolating results, a common practice when projecting reservoirs storage, with the addition of confidence intervals that would incorporate the uncertainty of future climate conditions. Results also exhibited the possibility of implementing the ML tool for forecasting storage of reservoirs with limited surveys. Time step for capacity loss estimation was found to be a critical indirect parameter of the ML tool, with larger basins requiring larger time step for the computations. Projected reservoir capacity losses showed a higher temporal correlation with maximum precipitation values for small basins. Projected percentages of reservoirs’ capacity loss per year with respect to initial capacity were larger for the trial basin with largest maximum rainfall events, and steepest channel slope regardless of the average precipitation or the basin area.

EarthArXiv (California Digital Library)
University of Iowa (US), United States Army Corps of Engineers (US), Tula University (RU), Coastal and Hydraulics Laboratory (US), Saint Louis University (US), U.S. Army Engineer Research and Development Center (US)
Water resources management and optimization
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon