Zuo et al. (2026) Decoding surface and root-zone soil moisture dynamics for agricultural drought assessment using multi-source climate records (1990–2019)

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-02-05
• Authors: Hao-Nan Zuo, Yingwei Sun, Pei Leng
• DOI: 10.1016/j.jhydrol.2026.135095

Research Groups

• State Key Laboratory of Efficient Utilization of Arable Land in China/Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, Beijing, China
• China Institute of Water Resources and Hydropower Research, Beijing, China

Short Summary

This study investigates the dynamics of surface and root-zone soil moisture using 30 years of ESA-CCI data to assess agricultural drought characteristics in three US states and develops a novel knowledge-guided machine learning model for improved drought prediction. It reveals distinct soil moisture responses to precipitation during prolonged versus short-duration droughts and demonstrates an 8% improvement in root-zone soil moisture prediction accuracy with the new model.

Objective

• To investigate the dominant factors affecting long-term and short-term soil moisture variations at different depths (surface and root-zone).
• To characterize agricultural drought events using 30 years (1990–2019) of satellite-based surface soil moisture (SSM) and root-zone soil moisture (RZSM) climatological records from ESA-CCI.
• To develop and evaluate a novel knowledge-guided machine learning model for agricultural drought prediction.

Study Configuration

• Spatial Scale: Three US states
• Temporal Scale: 30 years (1990–2019)

Methodology and Data

• Models used: Novel knowledge-guided machine learning model, standard machine learning model.
• Data sources: European Space Agency-Climate Change Initiative (ESA-CCI) for surface soil moisture (SSM) and root-zone soil moisture (RZSM) climatological records.

Main Results

• During prolonged drought events, root-zone soil moisture (RZSM) remains relatively stable, primarily influenced by precipitation and infiltration from surface soil moisture (SSM), while SSM exhibits pronounced fluctuations.
• For short-duration drought events, precipitation is consistently the dominant factor controlling both SSM and RZSM, with RZSM showing a lagged response to precipitation.
• The developed novel knowledge-guided machine learning model improves RZSM prediction performance by approximately 8% compared to a standard machine learning model and more accurately reflects drought intensity across the study region.

Contributions

• Provides new insights into the distinct dynamics of surface and root-zone soil moisture under different drought durations.
• Utilizes long-term (30-year) satellite-based soil moisture climatological records for agricultural drought assessment, addressing a gap in existing literature.
• Introduces a novel knowledge-guided machine learning framework that significantly enhances agricultural drought monitoring and forecasting capabilities, particularly for root-zone soil moisture prediction.

Funding

Not specified in the provided text.

Citation

@article{Zuo2026Decoding,
  author = {Zuo, Hao-Nan and Sun, Yingwei and Leng, Pei},
  title = {Decoding surface and root-zone soil moisture dynamics for agricultural drought assessment using multi-source climate records (1990–2019)},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.135095},
  url = {https://doi.org/10.1016/j.jhydrol.2026.135095}
}