Bai et al. (2026) A decade-long seamless-continuity daily L-band soil moisture product derived from SMOS observations since 2010

Identification

• Journal: Scientific Data
• Year: 2026
• Date: 2026-02-11
• Authors: Yu Bai, Li Jia, Tianjie Zhao, Zhiqing Peng, Jingyao Zheng, Chaolei Zheng, Ping Tang, Jiancheng Shi
• DOI: 10.1038/s41597-026-06756-9

Research Groups

• State Key Laboratory of Remote Sensing and Digital Earth, Aerospace Information Research Institute, Chinese Academy of Sciences, China
• Department of Earth System Science, Institute for Global Change Studies, Tsinghua University, China
• Aerospace Information Research Institute, Chinese Academy of Sciences, China
• National Space Science Center, Chinese Academy of Sciences, China

Short Summary

This study developed a fully automated gap-filling method, Discrete Cosine Transformation with Partial Least Squares (DCT-PLS), to generate the first decade-long (2010-2020) seamless-continuity global daily L-band soil moisture product (MTMA-SC_SM) from SMOS observations. The resulting product achieved high fidelity, comparable accuracy to original retrievals, and 100% spatiotemporal coverage, providing a robust dataset for climate and land surface studies.

Objective

• To develop and apply a fully automated gap-filling method (DCT-PLS) to reconstruct missing data in SMOS L-band soil moisture products, thereby generating a decade-long, global, spatiotemporally seamless daily soil moisture dataset to enhance its utility for eco-hydrological studies and climate trend assessment.

Study Configuration

• Spatial Scale: Global, 25 km spatial resolution (EASE-GRID 2.0 projection).
• Temporal Scale: Decade-long (June 2010 to December 2020), daily resolution.

Methodology and Data

• Models used: Discrete Cosine Transformation with Partial Least Squares (DCT-PLS) for gap-filling.
• Data sources:
  • Satellite: SMOS (Soil Moisture and Ocean Salinity) Multi-Temporal and Multi-Angular (MTMA) L-band H-pol multi-angular brightness temperature data (retrieved at 6:00 A.M.).
  • In-situ validation: Soil moisture data from 22 ground observation networks (398 sites) across five continents (Asia, Europe, Africa, North America, Oceania) from the International Soil Moisture Network (ISMN) and Long-Term Agroecosystem Research (LTAR) network, covering 2010-2019, primarily at 0-5 cm depth.
  • Simulated gaps: Artificially generated time-series and regional gaps in SMOS MTMA data for method validation.

Main Results

• The DCT-PLS method demonstrated high fidelity in reconstructing synthetic gaps, achieving a correlation coefficient (R) > 0.9, root mean squared error (RMSE) < 0.04 m³/m³, and mean absolute error (MAE) < 0.04 m³/m³.
• When evaluated against 22 in-situ soil moisture networks, the MTMA-SC_SM product achieved an overall R > 0.7 (specifically, 0.746) and an unbiased RMSE (ubRMSE) of 0.057 m³/m³, performing comparably to the original SMOS MTMA retrievals (accuracy difference < 0.01 for all metrics).
• The MTMA-SC_SM product achieved 100% spatiotemporal coverage, effectively filling gaps caused by orbital coverage, radio frequency interference, and retrieval failures.
• Spatially, the seamless product preserved mesoscale patterns and seasonal amplitudes across all climate zones, with no discernible boundary artifacts around reconstructed regions.
• The reconstructed soil moisture may not respond effectively to sudden rainfall or irrigation events, indicating an area for future improvement.

Contributions

• Generation of the first decade-long (2010-2020), gap-free, global daily L-band soil moisture record (MTMA-SC_SM) at 25 km resolution, addressing a critical limitation of existing satellite products.
• Introduction of a novel, fully automated gap-filling method (DCT-PLS) that leverages the intrinsic spatiotemporal coherence of SMOS data without relying on external ancillary data, thereby minimizing uncertainties.
• Provides a robust and continuous dataset essential for long-term climate trend assessment, land surface modeling, and eco-hydrological studies at the global scale.

Funding

• National Natural Science Foundation of China (Grant No. 42090014)
• National Key Research and Development Program of China (No. 2021YFB3900104)
• National Natural Science Foundation of China (Grant No. 42501450)

Citation

@article{Bai2026decadelong,
  author = {Bai, Yu and Jia, Li and Zhao, Tianjie and Peng, Zhiqing and Zheng, Jingyao and Zheng, Chaolei and Tang, Ping and Shi, Jiancheng},
  title = {A decade-long seamless-continuity daily L-band soil moisture product derived from SMOS observations since 2010},
  journal = {Scientific Data},
  year = {2026},
  doi = {10.1038/s41597-026-06756-9},
  url = {https://doi.org/10.1038/s41597-026-06756-9}
}