For decades, global soil moisture monitoring from space has depended on reference datasets. Satellite observations, while indispensable, are rarely used alone; instead, their retrieval algorithms are typically calibrated or constrained using external soil moisture products derived from other satellites, models, or reanalysis systems. This practice has helped stabilize retrievals, but it has also introduced fundamental limitations—reducing transparency, constraining transferability across regions, and complicating long-term consistency as reference products evolve. A growing question in Earth observation is whether this dependence is truly unavoidable.
In a study published (DOI: 10.34133/remotesensing.0939) on January 7, 2026, in the Journal of Remote Sensing, researchers from the Chinese Academy of Sciences, Peking University, and the China Meteorological Administration present PHYsics-based Soil rEflectivity Retrieval (PHYSER)—a physics-based framework for spaceborne GNSS-R soil moisture retrieval. The study demonstrates that global soil moisture can be retrieved independently, without relying on any external soil moisture reference products.
A long-standing constraint in satellite soil moisture retrieval
Soil moisture governs the exchange of water, energy, and carbon between the land surface and the atmosphere, influencing droughts, floods, ecosystem functioning, and agricultural productivity. Satellite remote sensing has become essential for monitoring soil moisture at regional to global scales, yet existing approaches face persistent challenges.
Conventional microwave sensors provide physically meaningful measurements but often struggle to balance spatial resolution, temporal coverage, and mission cost. More recently, Global Navigation Satellite System Reflectometry (GNSS-R) has emerged as a promising alternative. By passively receiving L-band signals continuously transmitted by navigation satellites such as GPS and BeiDou, GNSS-R offers low power consumption, all-weather capability, and dense spatiotemporal sampling.
Despite these advantages, most GNSS-R soil moisture retrieval methods still rely on empirical or semi-empirical relationships calibrated against external soil moisture products. This reliance weakens the physical interpretability of the results and limits their robustness when applied across regions, time periods, or future satellite missions. As GNSS-R constellations rapidly expand, the absence of an independent, physics-based retrieval framework has become a critical bottleneck.
Retrieving soil moisture from physical principles
PHYSER addresses this bottleneck by rethinking GNSS-R soil moisture retrieval from first principles. Rather than fitting GNSS-R observations to existing soil moisture datasets, the framework derives soil moisture directly from the physical interaction between navigation signals and the land surface.
At the core of PHYSER is the accurate reconstruction of soil surface reflectivity from GNSS-R measurements. This is achieved through a stepwise physical correction strategy. First, system-related biases inherent to the GNSS-R "multi-transmitter, single-receiver" observation geometry are corrected using inland water bodies as stable natural calibration targets. This step ensures consistency across different navigation signals and viewing geometries.
Second, land surface effects—particularly vegetation attenuation and surface roughness—are explicitly corrected using a physically based radiative transfer model. These land surface factors are shown to introduce larger uncertainties than satellite system errors, underscoring the importance of addressing them through physics-based correction rather than statistical adjustment.
With these corrections applied, soil reflectivity is transformed into soil permittivity using Fresnel equations. Soil moisture is then retrieved using established dielectric mixing models informed by global soil texture data.
Independent validation across space and ground observations
The PHYSER framework was evaluated using one year of observations from the BuFeng-1 A/B twin satellites, China's first spaceborne GNSS-R mission designed for technology demonstration. The retrieved soil moisture fields were compared with SMAP satellite products, ERA5-Land reanalysis data, and hundreds of in situ measurement sites worldwide.
Across diverse climatic and land surface conditions, the PHYSER-based retrievals show strong spatial and temporal consistency with these independent datasets. While retrieval errors are comparable to—or only slightly higher than—those of empirical GNSS-R approaches, PHYSER achieves this performance while remaining fully independent of reference soil moisture products.
"This work shows that GNSS-R soil moisture retrieval does not have to be a statistical imitation of other products," said a member of the research team. "By grounding the retrieval in physics, we gain transparency, robustness, and the ability to extend the method to future missions without retraining against external datasets."
Implications for future Earth observation missions
As GNSS-R missions multiply and satellite constellations become denser, the need for scalable and physically interpretable retrieval methods is becoming increasingly urgent. PHYSER provides a pathway toward soil moisture monitoring that is not tied to any specific reference product or satellite mission.
The framework has the potential to strengthen climate reanalysis, improve hydrological forecasting, and support agricultural decision-making, particularly in data-sparse regions. With further refinement—especially in densely vegetated environments—PHYSER could help enable operational GNSS-R soil moisture products that complement, and potentially stand alongside, traditional microwave remote sensing systems.
###
References
DOI
10.34133/remotesensing.0939
Original Source URL
https://spj.science.org/doi/10.34133/remotesensing.0939
Funding information
This study is supported by the Chinese Academy of Sciences, the Shandong Provincial Natural Science Foundation (Grant No. ZR2024QD048), the National Natural Science Foundation of China (NSFC) project (Grant No. 42471511), the BUFENG-1 Application Extension Program of the China Spacesat Co., Ltd., the ESA-MOST China Dragon5 Programme (ID.58070), the Fengyun Application Pioneering Project (FY-APP-2021.0301), the Beijing Nova Program (Grant Nos. 20230484327 and 20240484540), and the Hunan Provincial Natural Science Foundation project (Grant No. 2024JJ9186).
About Journal of Remote Sensing
The Journal of Remote Sensing, an online-only Open Access journal published in association with AIR-CAS, promotes the theory, science, and technology of remote sensing, as well as interdisciplinary research within earth and information science.