Ma et al. (2026) High resolution US water table depth estimates reveal quantity of accessible groundwater

Identification

• Journal: Communications Earth & Environment
• Year: 2026
• Date: 2026-01-14
• Authors: Yueling Ma, Laura E. Condon, Andrew Bennett, Amy Defnet, Danielle Tijerina-Kreuzer, Peter Melchior, R. M. Maxwell
• DOI: 10.1038/s43247-025-03094-3

Research Groups

• High Meadows Environmental Institute, Princeton University, Princeton, NJ, USA
• Department of Hydrology and Atmospheric Sciences, University of Arizona, Tucson, AZ, USA
• Department of Hydrology, Geological Survey of Denmark and Greenland, Copenhagen, Denmark
• Research Computing, Princeton University, Princeton, NJ, USA
• Department of Civil and Environmental Engineering, Princeton University, Princeton, NJ, USA
• Department of Astrophysical Sciences, Princeton University, Princeton, NJ, USA

Short Summary

This study develops a high-resolution (approximately 30 m) machine learning-based estimate of water table depth and accessible groundwater storage across the continental United States, revealing significant local variability and systematic underestimation by coarser resolution products.

Objective

• To develop a high-resolution (approximately 30 m) estimate of water table depth over the continental United States, including uncertainty, and quantify accessible groundwater storage.

Study Configuration

• Spatial Scale: Continental United States (CONUS), covering 7.3 million km², with a resolution of approximately 30 meters (one arc-second).
• Temporal Scale: Long-term mean water table depth, derived from observations spanning 1914 to 2023.

Methodology and Data

• Models used: Machine learning (Random Forest model with 300 decision trees).
• Data sources:
  • Over 1 million well observations (long-term mean water table depth measurements) from three major groundwater databases (1914-2023).
  • Ten spatially gridded climatological and hydrogeological input variables (e.g., precipitation, temperature, potential evapotranspiration, hydraulic conductivity, soil texture, elevation, slope, distances to streams) at resolutions ranging from 1 arc-second to 1 kilometer.
  • Porosity data from the ParFlow CONUS 2.0 platform (10 layers).

Main Results

• A high-resolution (approximately 30 m) water table depth product for the continental United States was developed with a Pearson Correlation Coefficient of 0.79 and a root mean square error of 14.94 m.
• The estimated total groundwater storage over the continental United States is 306,500 km³ (uncertainty range: 291,850–316,720 km³) above a depth of 392 m.
• Approximately 40% of the land area in CONUS has a water table depth shallower than 10 m, and 16% has water table depths shallower than 5 m.
• Coarse resolution products systematically underestimate total groundwater storage (e.g., an 18% underestimation at 100 km resolution compared to 30 m resolution) and significantly underrepresent shallow groundwater areas.
• Uncertainty in water table depth estimates varies regionally, with higher uncertainty generally observed in the more arid western United States.

Contributions

• Provides the highest resolution (approximately 30 m) and most statistically accurate estimate of water table depth for the continental United States to date, including uncertainty quantification.
• Bridges the scale gap between remote sensing and point observations by offering a spatially extensive and locally relevant groundwater product.
• Quantifies total and accessible groundwater storage with depth, specifically highlighting the abundance and importance of shallow groundwater.
• Demonstrates the systematic bias and limitations of coarser resolution groundwater products in capturing local variability and shallow groundwater.

Funding

• U.S. National Science Foundation Convergence Accelerator Program (grant no. CA-2040542).

Citation

@article{Ma2026High,
  author = {Ma, Yueling and Condon, Laura E. and Koch, Julian and Bennett, Andrew and Defnet, Amy and Tijerina-Kreuzer, Danielle and Melchior, Peter and Maxwell, R. M.},
  title = {High resolution US water table depth estimates reveal quantity of accessible groundwater},
  journal = {Communications Earth & Environment},
  year = {2026},
  doi = {10.1038/s43247-025-03094-3},
  url = {https://doi.org/10.1038/s43247-025-03094-3}
}