Zhao et al. (2026) Generating a consistent long-term L-band soil moisture record through Bayesian merging of SMOS and SMAP data

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-08-22
• Authors: Aoxing Zhao, Zushuai Wei, Linjie Guan, Beibei Yang, Jianxin Zhang, Bingsong Zhang, Tianjie Zhao, Lingkui Meng
• DOI: 10.1016/j.jhydrol.2026.136299

Research Groups

• School of Surveying and Land Information Engineering, Henan Polytechnic University
• School of Artificial Intelligence, Jianghan University
• Changjiang Survey, Planning, Design and Research Co., Ltd.
• Changjiang Space Information Technology Engineering Co., Ltd.
• Information Center of Ministry of Water Resources of China
• Meteorological Information and Technology Support Center of Hubei Province
• State Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, Chinese Academy of Sciences
• School of Remote Sensing and Information Engineering, Wuhan University

Short Summary

The study develops a Bayesian framework to harmonize SMOS soil moisture data with SMAP as a reference, creating a consistent long-term L-band soil moisture record from 2010 to 2024. The resulting merged product significantly increases the annual frequency of effective observations.

Objective

• To resolve systematic biases between SMOS and SMAP satellite missions and create a consistent, high-frequency, long-term global L-band soil moisture dataset.

Study Configuration

• Spatial Scale: Global
• Temporal Scale: 2010–2024

Methodology and Data

• Models used: Bayesian merging framework.
• Data sources: SMOS L-band brightness temperature, SMAP soil moisture (reference), MERRA-2 soil temperature, and ISMN ground measurements (validation).

Main Results

• SMOSReg Accuracy: The corrected SMOS record (SMOSReg) showed strong agreement with SMAP (2015–2022) with an ubRMSE of $0.033\text{ m}^3/\text{m}^3$, a correlation coefficient (CC) of 0.769, and a bias of $-0.002\text{ m}^3/\text{m}^3$.
• Ground Validation: Against ISMN data, SMOS_Reg achieved an ubRMSE of $0.053\text{ m}^3/\text{m}^3$ and CC of 0.662, performing comparably to SMAP (ubRMSE = $0.055\text{ m}^3/\text{m}^3$, CC = 0.695) but with a lower bias ($0.013\text{ m}^3/\text{m}^3$ vs $0.028\text{ m}^3/\text{m}^3$).
• Observational Frequency: The merged SMAP-SMOS_Reg product increased the mean annual number of effective observation days by 61.

Contributions

• Provides a harmonized, long-term (2010–2024) L-band soil moisture dataset that overcomes sensor-specific biases.
• Significantly enhances the spatiotemporal density of soil moisture observations by leveraging the orbital synergies of two different satellite missions.

Funding

• Not specified in the provided text.

Citation

@article{Zhao2026Generating,
  author = {Zhao, Aoxing and Wei, Zushuai and Guan, Linjie and Yang, Beibei and Zhang, Jianxin and Zhang, Bingsong and Zhao, Tianjie and Meng, Lingkui},
  title = {Generating a consistent long-term L-band soil moisture record through Bayesian merging of SMOS and SMAP data},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.136299},
  url = {https://doi.org/10.1016/j.jhydrol.2026.136299}
}