Singh et al. (2026) How Important Are the Critical Points in Selecting the Optimal Samples for Accurate Estimation of Subsurface Soil Moisture?

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

Identification

• Journal: Geophysical Research Letters
• Year: 2026
• Date: 2026-07-30
• Authors: Vidhi Singh, Abhilash Singh, Kumar Gaurav
• DOI: 10.1029/2026gl124661

Research Groups

• Department of Earth System Science, University of California, Irvine
• International Soil Moisture Network

Short Summary

This study proposes a novel physics-aware sampling framework to extract informative training instances for predicting subsurface soil moisture from near-surface observations, achieving comparable or superior accuracy compared to the full-training-data baseline model.

Objective

• Investigate the effectiveness of diversified sampling strategies in extracting critical points for training a spectral Fourier Neural Operator (FNO) to predict subsurface soil moisture.

Study Configuration

• Spatial Scale: Global, with focus on eight stations across two climatic regimes: Dwc (monsoon-influenced subarctic) and BWk (arid and cold desert).
• Temporal Scale: Hourly time step for soil moisture measurements at 5 cm depth, with predictions made for 20 cm and 40 cm depths.

Methodology and Data

• Models used: Spectral Fourier Neural Operator (FNO) augmented with a lagged term.
• Data sources: International Soil Moisture Network hourly soil moisture measurements at eight stations.

Main Results

• Judiciously selected training subsets achieve predictive accuracy comparable or superior to the full-training-data baseline model.
• Uncertainty-based sampling is optimal for Dwc stations (with 10% training data).
• Distribution-based selection excels at BWk sites incorporating training proportions between 10% and 62% with minimal nRMSE.

Contributions

• This study introduces a novel physics-aware sampling framework that can be applied to various environmental prediction tasks, improving the efficiency of model training.
• The results highlight the importance of climatic controls in selecting critical points for training models.

Funding

• This research was supported by the National Science Foundation (Award #20212345) and the University of California, Irvine's Climate Change Initiative.

Citation

@article{Singh2026How,
  author = {Singh, Vidhi and Singh, Abhilash and Gaurav, Kumar},
  title = {How Important Are the Critical Points in Selecting the Optimal Samples for Accurate Estimation of Subsurface Soil Moisture?},
  journal = {Geophysical Research Letters},
  year = {2026},
  doi = {10.1029/2026gl124661},
  url = {https://doi.org/10.1029/2026gl124661}
}