A Forward Model and Inversion Algorithm for Near-Surface Soil Moisture Estimation With GNSS Refraction Pattern Technique

The global navigation satellite system (GNSS) refraction pattern technique makes use of two pairs of GNSS receivers and antennas to collect the refracted signal in the medium and the direct signal in the air, respectively. Due to the sensitive response of the refracted signal to variation of the dielectric constant, the technique is suitable for measuring medium dielectric constant-related parameters such as snowpack density, vegetation water content, and soil moisture. In this article, a forward model related to the power ratio of the refracted signal to the direct one is developed to elucidate the mechanism of soil moisture-induced refracted signal strength attenuation. By making use of the simulating results derived from the model, a second-order polynomial function is established to describe the relationship between the power ratio, soil temperature, and soil moisture. Based on the function, an inversion algorithm, which takes the GNSS carrier-to-noise (C/N-0) observations under high elevation angles (50 degrees-60 degrees) and soil temperature as the inputs, is proposed for the near-surface soil moisture estimation. The proposed algorithm is validated through a dataset collected in an experimental campaign over two years. The results demonstrate that there exists a good agreement between the proposed method-derived soil moisture estimations and ground-truth ones; and the root-mean-square error (RMSE) of the proposed algorithm-derived soil moisture estimation is 0.009 cm(3) cm(-3) when the ground-truth soil moisture is in the range from 0.150 to 0.550 cm(3) cm(-3). Because the observations were collected by using consumer-grade GNSS chips and antennas, this study also provides a basis for the design and development of the low-cost GNSS soil moisture sensor in the future.