Exploration of Tianmu-1 and CYGNSS to estimate soil moisture with GNSS-R in southwest China

Abstract The Tianmu-1 satellites (TM-1) is China’s first commercial BDS/Global Navigation Satellite System (GNSS) remote sensing and detection constellation that is compatible with the four major navigation systems, namely BDS, GPS, GLONASS, and Galileo. It could conduct three-dimensional and integrated operational detection of the global ocean, atmosphere, and ionosphere around the clock. The Cyclone Global Navigation Satellite System (CYGNSS) is an Earth System Science Pathfinder mission of the National Aeronautics and Space Administration (NASA) in the United States. This mission represents the first frequent space-based measurements of surface wind speeds in the inner cores of tropical cyclones. Comprising eight microsatellites, it is capable of providing near-seamless coverage of the Earth. This paper utilizes the observation reflection data of the TM-1 and CYGNSS in 2023. It respectively studies the capabilities of these two satellite constellations in estimating soil moisture (SM, 0–5 cm). This research opens broader prospects for obtaining high-precision SM products using satellite-based GNSS reflection data. Using Global Navigation Satellite System-reflectometry (GNSS-R) signal data in conjunction with the Normalized Difference Vegetation Index (NDVI), precipitation, Digital Elevation Model (DEM), slope, and geographic coordinates, the SM data from the Soil Moisture Active Passive (SMAP) satellite serves as the benchmark for developing a model based on the Random Forest (RF) method, facilitating SM value estimation in the Sichuan-Tibet transportation corridor. Research results indicate that the generated SM Pearson correlation coefficient (R) is 0.806, with a root mean square error (RMSE) of 0.054 cm 3 cm −3 . Based on this, the extrapolated monthly SM values were obtained, with the maximum RMSE of the monthly retrieval results not exceeding 0.065 cm 3 cm −3 , and the absolute value of the bias not exceeding 0.037 cm 3 cm −3 . Utilizing the CYGNSS, the R soil moisture in the study area is 0.765, with an RMSE of 0.066 cm 3 cm −3 . The extrapolated monthly SM values indicate that the maximum RMSE for the monthly retrieval results does not exceed 0.088 cm 3 cm −3 , while the absolute value of the bias remains within 0.068 cm 3 cm −3 . Overall, although the SM results of the two satellites are underestimated in the study area, the SM retrieval effect achieved by TM is better than that of the CYGNSS satellite. Simultaneously, it is proved that using TM to realize spaceborne GNSS-R SM retrieval is feasible, providing a new data source for studying spaceborne GNSS-R SM retrieval.