Quantitative Study On The Retrieval Error Of Soil Moisture With Mironov Dielectric Constant Model

Soil moisture is the main parameters of reflecting the land changing.Remote sensing satellites carrying the microwave bands have opened a new method to acquire quickly and widely the surface soil moisture.Especially the first successful using of passive microwave satellite of SMOS(Soil Moisture and Ocean Salinity)had established firm foundation for acquiring global soil moisture.However,with a lot of experiments verification,the retrieval results exist strong fluctuation and the retrieval algorithm and model need to be optimized further.After the year of 2015,SMOS changed the Dobson dielectric constant model to the Mironov dielectric constant model,leading to total retrieval accuracy of soil moisture having got improved better than before.For that,this thesis mainly explores how the retrieval accuracy of soil moisture is affected by the Mironov dielectric constant and how to optimize the parameters of L-MEB model by the field data.The main research results are as follows:(1)The key sensitivity parameters analyzed in L-MEB.By analyzing the Mironov dielectric constant,soil roughness model and radiation transfer equation,it is found that: the Moronov dielectric constant is sensitive to the soil moisture parameter and not sensitive to the temperature parameter;The soil roughness model is influenced by the incidence angle,the dielectric constant and the soil roughness,but the dielectric constant is the main influence parameter;The simulated radiant brightness temperature under low vegetation land cover has a strong linear relationship with the emissivity;The atmospheric factors has a weak influence on the brightness temperature.(2)The reliability study of SMOS-Mironov dielectric constant and the source of errors study of soil moisture by SMOS.Through the analysis of the reliability of SMOS's dielectric constant model,it is found that:The relationship between the soil moisture and the dielectric constant of SMOS is "one-to-many",which is not consistent with the theory of "one-to-one";The relationship between soil moisture difference and dielectric constant difference has a strong correction(R>0.93)through comparing the SMOS dielectric constant and soil moisture of retrieval with the field data of USCRN.The retrieval accuracy study on the different environment factors found that The rate of retrieval soil moisture success and the retrieval accuracy are related to the land cover classification.And the radiation models and parameters need to be further optimized for different land cover.Meanwhile,this thesis find that SMOS could still retrieval the soil moisture successfully under precipitation condition,but the accuracy is affected.(3)The study for improving the parameter of Mironov dielectric constant.Based on the characteristic that the Mironov dielectric constant is sensitive to the soil moisture.This thesis quantifies the Mironov dielectric constant model and builds the empirical formula of both them.The SMOS dielectric constant model is replaced by the building empirical formula and use the SMOS's dielectric constant calculates the soil moisture.The final obtained soil moisture is better than that of SMOS.(4)Optimization the retrieval algorithm based on the L-MEB model.Based on the L-MEB forward model and toward to different land cover,using the field data optimizes the important parameters of Mironov dielectric constant model.And using the improved parameters of L-MEB algorithm to retrieval soil moisture,the result shows that the soil moisture retrieved by SMOS L2 brightness temperature data has got better result than before.