刘全儒 et al. (2026) Depth-dependent predictive information in monitored soil moisture profiles: Contrasting roles of persistence, forcing, and HYDRUS-derived states

Identification

• Journal: Agricultural Water Management
• Year: 2026
• Date: 2026-09-19
• Authors: 刘全儒, Zongzhi Wang, Ying Bai, Kun Wang, Wenhua Wan, Liang Cheng, Huihua Du, Wenqi Wang, Yao Zhenzhu, Qi Ding, Yongbing Zhang
• DOI: 10.1016/j.agwat.2026.110798

Research Groups

• The National Key Laboratory of Water Disaster Prevention, Nanjing Hydraulic Research Institute, China
• Shandong Water Conservancy Vocational College, Rizhao, China
• Rizhao Yushan Tea Industry Co., Ltd., Rizhao, China

Short Summary

This study evaluates the incremental predictive value of HYDRUS-derived states in monitored soil moisture profiles. The results show that antecedent observed-state information (Persistence) is a strong benchmark for same-day SWC estimation, and HYDRUS-derived states do not provide a robust incremental accuracy gain in the tea-plantation profile.

Objective

• To investigate the role of persistence, forcing, and HYDRUS-derived states in monitored soil moisture profiles
• To evaluate the incremental predictive value of HYDRUS-derived states using paired models with identical non-HYDRUS inputs

Study Configuration

• Spatial Scale: Local scale (10-100 cm depth)
• Temporal Scale: Daily time step, 90-d protocol for HYDRUS calibration and validation

Methodology and Data

• Models used: LightGBM, HYDRUS-1D
• Data sources: Soil moisture sensors, meteorological observations, satellite-derived NDVI and FVC

Main Results

• Persistence provided a strong benchmark for same-day SWC estimation in both the tea-plantation and grassland profiles
• In the tea-plantation profile, HYDRUS-derived states did not provide a robust incremental accuracy gain under the primary 90-d protocol
• In the grassland profile, bootstrap-supported improvements were concentrated at 50-80 cm

Contributions

• This study provides new insights into the role of persistence and HYDRUS-derived states in monitored soil moisture profiles
• The results highlight the importance of antecedent observed-state information as a benchmark for same-day SWC estimation

Funding

• National Key Laboratory of Water Disaster Prevention, China (project code: XXXX)
• Shandong Water Conservancy Vocational College, China (project code: YYYY)

Citation

@article{刘全儒2026Depthdependent,
  author = {刘全儒 and Wang, Zongzhi and Bai, Ying and Wang, Kun and Wan, Wenhua and Cheng, Liang and Du, Huihua and Wang, Wenqi and Zhenzhu, Yao and Ding, Qi and Zhang, Yongbing},
  title = {Depth-dependent predictive information in monitored soil moisture profiles: Contrasting roles of persistence, forcing, and HYDRUS-derived states},
  journal = {Agricultural Water Management},
  year = {2026},
  doi = {10.1016/j.agwat.2026.110798},
  url = {https://doi.org/10.1016/j.agwat.2026.110798}
}