Soil Moisture Retrieve Using GNSS-R Observation Data

Soil moisture has always been the surface physical quantity concerned by people.It is an indicator of the global surface water cycle or a key point in the fields of agriculture,meteorology and hydrology research.It is an important parameter for describing surface conditions.So accurate observation of soil moisture over a large area is of great significance to agriculture,hydrology and meteorology.Microwave remote sensing can realize large-area,non-contact,real-time continuous soil moisture detection in the existing methods.Furthermore,GNSS-R,as an important branch of microwave remote sensing technology,is widely used in remote sensing monitoring in various fields.Due to the complex land reflection conditions on the land surface and the attenuation effect various factors on the signal,GNSS-R soil moisture detection technology is still in the exploration stage,and mainly focused on the ground and airborne observation modes,and it is difficult to carry out large-scale surface soil moisture detection.Therefore,the realization of Space-borne GNSS-R soil moisture detection is also a problem to be solved at present.The current situation is studied in this paper,which can play a certain role in promoting the development of GNSS-R remote sensing applications.In this paper,on the basis of the theoretical study of GNSS reflection signal characteristics,related work is performed toward on soil moisture retrieval methods under two different observation platforms of airborne/Space-borne GNSS-R.Aiming at the traditional Two-antenna GNSS-R retrieval model,considering the influencing factors in the retrieval process of airborne platform and given the correction method,the retrieval model suitable for airborne GNSS-R soil moisture is obtained.At the same time,the feasibility of using the Space-borne GNSS-R observation data to retrieve soil moisture is analyzed.The correlation model between DDM SNR and soil moisture is proposed.Finally,the above model was verified by using the airborne SMEX02 data and the Space-borne TDS-1 data.The experimental results show that under the airborne observation mode,the retrieval soil moisture results in the absence of without rainfall are in good agreement with the measured trend of soil moisture change.There is a strong correlation between them of the error is generally controlled within 0.05m~3/m~3,the correlation coefficient is R=0.76 and the RMSE is 0.0185 m~3/m~3.There is a large deviation between the soil moisture results and the measured results,but the correlation coefficient between the corrected soil moisture retrieval result and the measured result has been greatly improved:before correction,R=0.2,RMSE=0.023m~3/m~3;after correction,R=0.6,RMSE=0.019m~3/m~3,which indicates that the correction method has effective.However,there is still a large retrieval error in the state of rainfall,so the rainfall state data should be abandoned in the subsequent retrieval process.And under the Space-borne observation mode,aiming at the problem that effective information of Space-borne DDM data is difficult to distinguish and extract in the retrieval process,a DDM waveform classification method is proposed,verified the method validity,and given the DDM data suitable for Space-borne soil moisture retrieval after classification,is regular waveform data.Two research areas are obtained by processing.Between the DDM SNR and the SMOS satellite soil moisture data of the two selected study areas have higher correlation coefficients are:R=0.69 and R=0.75.Which indicate that it is feasible to use Space-borne GNSS-R DDM data for soil moisture retrieval.