Wang et al. (2026) Filling data gaps in soil moisture monitoring networks via integrating spatio-temporal contextual information

Identification

• Journal: Hydrology and earth system sciences
• Year: 2026
• Date: 2026-07-06
• Authors: Weixuan Wang, Yizhuo Meng, Zushuai Wei, Linguang Miao, Hui Wang, Wen Zhang
• DOI: 10.5194/hess-30-4191-2026

Research Groups

• School of Remote Sensing and Information Engineering, Wuhan University, China
• Department of Computer Science, University of Southern California, USA
• School of Artificial Intelligence, Jianghan University, China

Short Summary

This study proposes a novel spatiotemporal reconstruction framework called ST-GapFill to fill missing values in soil moisture monitoring networks. The framework integrates multi-source contextual information through two key innovations: spatial correlation-guided neighbor selection and long short-term memory (LSTM) network.

Objective

• To develop an accurate method for filling missing values in soil moisture data using spatiotemporal contextual information

Study Configuration

• Spatial Scale: Regional scale, covering the Shandian River Basin in China
• Temporal Scale: Long-term, with data collected from 2019 to 2020

Methodology and Data

• Models used: ST-GapFill framework, which combines spatial correlation-guided neighbor selection and LSTM network
• Data sources: Soil moisture data from the SMN-SDR network in Shandian River Basin, China; precipitation data from nearby stations

Main Results

• The proposed ST-GapFill framework achieved high accuracy in filling missing values, with a correlation coefficient of 0.91 with ground truth under low missing-rate conditions (less than 50%)
• Comparative analysis revealed that ST-GapFill outperformed traditional methods such as IDW and ARIMA

Contributions

• This study contributes to the development of accurate and robust methods for filling missing values in soil moisture data, which is essential for climate change impact assessments and precision agriculture

Funding

• Not specified

Citation

@article{Wang2026Filling,
  author = {Wang, Weixuan and Meng, Yizhuo and Wei, Zushuai and Miao, Linguang and Wang, Hui and Zhang, Wen},
  title = {Filling data gaps in soil moisture monitoring networks via integrating spatio-temporal contextual information},
  journal = {Hydrology and earth system sciences},
  year = {2026},
  doi = {10.5194/hess-30-4191-2026},
  url = {https://doi.org/10.5194/hess-30-4191-2026}
}