Huang et al. (2026) Satellite soil moisture as an additional observational constraint for machine learning-based irrigation water use modeling

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Environmental Research Letters
• Year: 2026
• Date: 2026-06-16
• Authors: Xin Huang, Qing He, Naota Hanasaki, Taikan Oki
• DOI: 10.1088/1748-9326/ae7e0c

Research Groups

Not specified in the provided text.

Short Summary

This study demonstrates that a cell-wise machine learning framework combined with satellite soil moisture data significantly improves the estimation of high-resolution (9 km) monthly irrigation water use across the conterminous United States compared to conventional pooled learning methods.

Objective

• To evaluate whether spatially explicit machine learning parameterization and the integration of satellite soil moisture observations improve the accuracy of monthly irrigation water use (IWU) estimations by addressing regional heterogeneity and spatial nonstationarity.

Study Configuration

• Spatial Scale: Conterminous United States (CONUS) at a 9 km resolution.
• Temporal Scale: Monthly.

Methodology and Data

• Models used: Machine Learning (comparing a conventional pooled learning strategy against a cell-wise framework).
• Data sources: Satellite soil moisture products and hydro-meteorological predictors.

Main Results

• Framework Performance: The cell-wise framework outperformed the pooled learning strategy, which failed to account for distinct local irrigation regimes.
• Satellite Integration: Incorporating satellite soil moisture improved agreement with benchmark data in approximately 90% of irrigated grid cells.
• Regional Variance: The greatest improvements occurred in semi-arid regions where precipitation is decoupled from the growing season; in these areas, satellite data effectively reduced model overestimation during non-growing seasons.

Contributions

• Provides evidence that high-resolution IWU modeling must account for spatial nonstationarity.
• Establishes that satellite-derived soil moisture provides critical observational constraints that exceed the information available in standard meteorological forcings and model-derived hydrological states.

Funding

Not specified in the provided text.

Citation

@article{Huang2026Satellite,
  author = {Huang, Xin and He, Qing and Hanasaki, Naota and Oki, Taikan},
  title = {Satellite soil moisture as an additional observational constraint for machine learning-based irrigation water use modeling},
  journal = {Environmental Research Letters},
  year = {2026},
  doi = {10.1088/1748-9326/ae7e0c},
  url = {https://doi.org/10.1088/1748-9326/ae7e0c}
}