Soil Moisture Retrieval Research Based On L-band Passive Remote Sensing

Soil moisture describes the proportion of water in soil, which has great significance in drought warning, flood forecasting, vegetation monitoring and climate research. Currently, soil moisture probing methods include optical remote sensing, infrared remote sensing, and passive microwave remote sensing. However, due to the poor penetration ability, optical and infrared remote sensing are unable to get inner information of soil, so their applications are limited. Passive microwave remote sensing, which makes use of the strong penetration of microwave to obtain radiation information of soil, is proved to be an effective method. Among all frequency bands, L-band, which is most sensitive to changes in soil moisture and slightly affected by clouds and atmosphere, is believed to be one of the most suitable bands for soil moisture probing.In this paper, the main content is based on the L-band passive remote sensing of the soil moisture retrieval algorithm. Firstly, mathematical theory of soil terrain(including bare soil, vegetation soil and forest soil) brightness temperature models and the iterative retrieval algorithm are introduced. We conduct the simulation analyses to research the different parameters' effects on retrieval accuracy. It reveals that calibration accuracy has much more influence than the radiometer noise on retrieval accuracy. And for the bare soil terrain, the impact of surface roughness error on the retrieval accuracy is larger than soil constituents error; for vegetation terrain, the impact of leaf surface index error on retrieval accuracy is largest, followed by the surface roughness error and soil constitutes error.On the basis of theoretical simulation, we conduct processing and analysis of satellite data and filed data. According to the SMOS satellite measuring data, we conduct soil moisture retrieval procession globally. We analyze and compare our results with the SMOS official data and the global sensors' field data, finding that 81.43% of our results have less than 5% error when compared with SMOS official data and 71.74% when compared with field data. Apart from that, we compare and analyze the official retreival results of Aqurius and SMOS satellites, finding that SMOS's official results are more accurate than Aqurius' s official results on the reference of field data within a certain region of North America. We analyze the cuases and find out SMOS satellite can obtain more information on brightness temperature of land due to SMOS satellite's higher resolution and wider coverage of incident angles, which can effectively improve the accuracy of retrieval results.