Zhang et al. (2026) Multi-sensor (since 1997) global soil moisture mapping with enhanced Spatio-temporal coverage through machine learning framework fusion

Identification

• Journal: Remote Sensing of Environment
• Year: 2026
• Date: 2026-01-07
• Authors: Haojie Zhang, T. C. Zhao, Zhiqing Peng, Jingyao Zheng, Yu Bai, Nemesio Rodriguez-Fernadez, Donghai Zheng, Huazhu Xue, zhanliang yuan, Qian Cui, Peng Guo, Zushuai Wei, Peilin Song, Lixin Dong, Panpan Yao, Qiangqiang Yuan, L Y Meng, Jiancheng Shi
• DOI: 10.1016/j.rse.2025.115221

Research Groups

• State Key Laboratory of Remote Sensing and Digital Earth, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, China
• National Tibetan Plateau Data Center, State Key Laboratory of Tibetan Plateau Earth System, Environment and Resources, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing, China
• University of Chinese Academy of Science, Beijing, China
• Department of Earth System Science, Tsinghua University, Beijing, China
• Centre d'´Etudes Spatiales de la Biosph`ere (CESBIO), Universit´e de Toulouse, Centre National d'´Etudes Spatiales (CNES), Centre National de la Recherche Scientifique (CNRS), Institut de Recherche pour le D´eveloppement (IRD), Universit´e Paul Sabatier, Institut National de Recherche pour l'Agriculture, l'Alimentation et l'Environnement (INRAE), Toulouse, France
• School of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo, China
• Information Center (Hydrology Monitor and Forecast Center), Ministry of Water Resources, Beijing, China
• College of Information Science and Engineering, Shandong Agricultural University, Tai’an, China
• School of Artificial Intelligence, Jianghan University, Wuhan, China
• Key Laboratory of Physical Electronics and Devices, Ministry of Education, Faculty of Electronic and Information Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi, China
• Key Laboratory of Radiometric Calibration and Validation for Environmental Satellites, National Satellite Meteorological Center (National Center for Space Weather), China Meteorological Administration, Beijing, China
• School of Geodesy and Geomatics, Wuhan University, Wuhan, Hubei, China
• School of Remote Sensing and Information Engineering, Wuhan University, Wuhan, China
• National Space Science Center, Chinese Academy of Sciences, Beijing, China

Short Summary

This study developed a two-stage machine learning framework to fuse multi-sensor passive microwave observations, generating a global daily soil moisture product with enhanced spatio-temporal coverage and consistency from 1997 to 2023. The resulting product demonstrates high accuracy and improved land coverage, inheriting the performance of SMAP L-band observations.

Objective

• To overcome the spatio-temporal inconsistency and poor coverage of global soil moisture products derived from multiple passive microwave sensors.
• To generate a daily global soil moisture product with enhanced spatio-temporal coverage and high accuracy by fusing multi-sensor observations through a machine learning framework.

Study Configuration

• Spatial Scale: Global, 25 km resolution.
• Temporal Scale: Daily, 1997–2023.

Methodology and Data

• Models used:
  • Multi-channel collaborative algorithm (MCCA) for SMAP L-band observations.
  • Long Short-Term Memory (LSTM) network for global gridded soil moisture dynamics.
  • Machine learning framework fusion.
• Data sources:
  • Satellite passive microwave sensors: SMAP (L-band), TMI, GMI, AMSR-E, AMSR2 (brightness temperature observations).
  • Ground measurements from 24 dense global observation networks for validation.
  • Other datasets for cross-validation.

Main Results

• The developed MCCA-ML soil moisture product exhibits spatial distribution and seasonal variation patterns closely matching those of MCCA SMAP, reflecting global climatic and geographic features.
• Validation against 24 dense global observation networks showed a global correlation coefficient (r) of 0.76, a Root Mean Square Error (RMSE) of 0.068 m³/m³, and an unbiased Root Mean Square Error (ubRMSE) of 0.059 m³/m³.
• The daily global land coverage of MCCA-ML soil moisture typically exceeds 80% during periods when two or more satellites were operational, demonstrating good soil moisture detection capability.

Contributions

• Introduction of a novel two-stage machine learning framework to address spatio-temporal inconsistencies and coverage gaps in multi-sensor global soil moisture mapping.
• Generation of a long-term (1997–2023) daily global soil moisture product with a 25 km resolution, significantly enhancing spatio-temporal coverage and consistency.
• Successful inheritance of the high-quality retrieval accuracy of SMAP L-band observations through the fusion of multiple passive microwave sensors.

Funding

• Not specified in the provided text.

Citation

@article{Zhang2026Multisensor,
  author = {Zhang, Haojie and Zhao, T. C. and Peng, Zhiqing and Zheng, Jingyao and Bai, Yu and Rodriguez-Fernadez, Nemesio and Zheng, Donghai and Xue, Huazhu and yuan, zhanliang and Cui, Qian and Guo, Peng and Wei, Zushuai and Song, Peilin and Dong, Lixin and Yao, Panpan and Yuan, Qiangqiang and Meng, L Y and Shi, Jiancheng},
  title = {Multi-sensor (since 1997) global soil moisture mapping with enhanced Spatio-temporal coverage through machine learning framework fusion},
  journal = {Remote Sensing of Environment},
  year = {2026},
  doi = {10.1016/j.rse.2025.115221},
  url = {https://doi.org/10.1016/j.rse.2025.115221}
}