Wang et al. (2026) A novel approach for Quasi-Global daily continuous surface soil moisture downscaling at 500-m resolution using CYGNSS observations

Identification

• Journal: International Journal of Applied Earth Observation and Geoinformation
• Year: 2026
• Date: 2026-01-31
• Authors: Jundong Wang, Ting Yang, Wanxue Zhu, Shiji LI, Zixuan Tang, Wei Wan, Zhigang Sun
• DOI: 10.1016/j.jag.2026.105137

Research Groups

• Key Laboratory of Ecosystem Network Observation and Modeling, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China
• CAS Engineering Laboratory for Yellow River Delta Modern Agriculture, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China
• University of Chinese Academy of Sciences, Beijing, China
• Shandong Dongying Institute of Geographic Sciences, Dongying, China
• National-Regional Joint Engineering Research Center for Soil Pollution Control and Remediation in South China, Guangdong Key Laboratory of Integrated Agro-environmental Pollution Control and Management, Institute of Eco-environmental and Soil Sciences, Guangdong Academy of Sciences, Guangzhou, China
• College of Land Science and Technology, China Agricultural University, Beijing, China
• Key Laboratory of Water Cycle and Related Land Surface Processes, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China
• Institute of Remote Sensing and Geographic Information System, School of Earth and Space Sciences, Peking University, Beijing, China

Short Summary

This study developed a novel framework to generate daily, continuous, quasi-global surface soil moisture (SSM) at 500-meter resolution by integrating CYGNSS observations with optical imagery, without requiring in-situ SSM inputs. The approach successfully enhances spatial detail while maintaining accuracy, achieving a mean unbiased root-mean-square error (ubRMSE) of 0.061 cm³/cm³ and a positive GDOWN value of 0.017.

Objective

• To develop a practical framework for generating daily, continuous, quasi-global surface soil moisture (SSM) at 500-meter resolution by integrating CYGNSS observations with optical remote-sensing data.

Study Configuration

• Spatial Scale: Quasi-global (pan-tropical areas between 38°N and 38°S), with downscaled SSM at 500-meter resolution and coarse SSM at 9-kilometer resolution.
• Temporal Scale: Daily continuous, applied and validated for August 2019 – December 2022, with POBI algorithm validation for January 2023 – December 2023.

Methodology and Data

• Models used:
  • Previously-Observed Behavior Interpolation (POBI) algorithm
  • Modified Reflectivity-Vegetation-Roughness (R-V-R) model
  • Enhanced OPtical TRApezoid Model (OPTRAM)
  • Bayesian algorithm (for downscaling)
  • Linear algorithm (for comparison)
  • Random Forest (RF) algorithm (for comparison)
• Data sources:
  • Cyclone Global Navigation Satellite System (CYGNSS) Level 1 data (L-band microwave signals)
  • SMAP-E dataset (Soil Moisture Active Passive Enhanced L3 Radiometer Global and Polar Grid Daily 9 km EASE-Grid Soil Moisture V006)
  • MODIS SDC500 data (Global Daily Seamless Surface Reflectance Data Cube at 500-meter resolution, bands 1-7)
  • International Soil Moisture Network (ISMN) in-situ SSM measurements (0-5 cm depth, 194 sites)
  • MODIS Land Cover Product (MCD12Q1) for IGBP land cover types

Main Results

• The POBI algorithm increased the quasi-global daily CYGNSS Surface Reflectivity (SR) coverage from less than 30% to approximately 95%, with an overall mean bias of 0.06 dB and a standard deviation of 2.98 dB.
• At the 9-kilometer scale, CYGNSS SSM (derived using the R-V-R model) showed good agreement with SMAP SSM, exhibiting R values of 0.5–0.95 and ubRMSE of 0.02–0.07 cm³/cm³ across most regions.
• The enhanced OPTRAM successfully and automatically delineated wet and dry edges in the Shortwave Infrared Transformed Reflectance (STR)-Fractional Vegetation Cover (FVC) space across diverse land cover and climatic conditions.
• The 500-meter downscaled SSM demonstrated improved consistency with ISMN in-situ SSM compared to the 9-kilometer CYGNSS SSM, with an overall mean R of 0.542 and a mean ubRMSE of 0.061 cm³/cm³.
• The GDOWN metric, indicating downscaling performance, showed a positive overall mean of 0.017, signifying effective improvement over non-downscaled data, with positive values across most land cover types except evergreen needle forests.
• The downscaled SSM preserved the key accuracy characteristics of the native CYGNSS retrievals while significantly enhancing spatial detail and variability, with absolute differences generally below 0.06 cm³/cm³.
• Comparative analysis showed that the Bayesian algorithm outperformed Linear (mean R=0.364, ubRMSE=0.075 cm³/cm³) and Random Forest (mean R=0.432, ubRMSE=0.056 cm³/cm³) methods, demonstrating stable and consistently strong correlations.

Contributions

• Proposed a novel, physically-grounded framework for daily, continuous, quasi-global 500-meter SSM downscaling by synergistically integrating CYGNSS microwave observations and optical imagery, without requiring in-situ SSM inputs for the downscaling process itself.
• Developed an enhanced OPTRAM with an automated iterative process for wet/dry edge detection, making the optical-based SSM retrieval and downscaling scalable for large-scale applications.
• Demonstrated the effectiveness and robustness of the Bayesian algorithm for disaggregating coarse 9-kilometer CYGNSS SSM into high-resolution 500-meter products, outperforming conventional Linear and Random Forest methods.
• Maximized the information content from satellite Earth Observation data while minimizing dependence on ground-based data, thereby enhancing the robustness and transferability of the SSM product.
• The proposed framework is sensor-agnostic and extensible, allowing for flexible integration of future Earth Observations and continuous algorithm improvements.

Funding

• Shandong Provincial Natural Science Foundation (Grant No. ZR2024QD048)
• National Natural Science Foundation of China (No. 42271349, No. 42101376)
• The GDAS' Project of Science and Technology Development (2022GDASZH-2022020402-3, 2022GDASZH-2022010106)
• National Key Research and Development Program of China (2021YFD1900902)
• Program of Yellow River Delta Scholars (2020–2024)

Citation

@article{Wang2026novel,
  author = {Wang, Jundong and Yang, Ting and Zhu, Wanxue and LI, Shiji and Tang, Zixuan and Wan, Wei and Sun, Zhigang},
  title = {A novel approach for Quasi-Global daily continuous surface soil moisture downscaling at 500-m resolution using CYGNSS observations},
  journal = {International Journal of Applied Earth Observation and Geoinformation},
  year = {2026},
  doi = {10.1016/j.jag.2026.105137},
  url = {https://doi.org/10.1016/j.jag.2026.105137}
}