Synergetic Inversion Of Surface Soil Moisture Based On C/L Band Active And Passive Microwave Remote Sensing

Active and passive microwave remote sensing has great potential for inversion of soil moisture,but active and passive microwave remote sensing has its own characteristics in soil moisture monitoring.Active microwave radar remote sensing has the characteristics of high spatial resolution,but long revisit time and low accuracy of soil moisture retrieval.Passive microwave remote sensing has the advantages of short revisit time and high accuracy of soil moisture retrieval,but low spatial resolution.In order to obtain high spatial resolution and high precision soil moisture products,the collaborative inversion of soil moisture based on active microwave SAR and passive microwave radiometer data is currently the best choice.This dissertation takes full account of the characteristics of active and passive microwave remote sensing sensors and data.Then it,and carries out research on the collaborative inversion of surface soil moisture model for C-band active microwave radar data and L-band passive microwave radiometer brightness temperature.Among them,it mainly includes the following aspects:Based on the analysis of the soil dielectric constant,the simulation accuracy analysis and applicability evaluation of several mixed dielectric constant models were carried out.Research on the simulation accuracy and applicability of several mixed dielectric constant models based on the influence factors of soil dielectric properties.Firstly,the effects of five factors such as soil texture,mineral composition,soil moisture content,temperature and frequency on the soil dielectric properties were studied in depth.Then,collect a variety of experimental data at home and abroad comprehensively.The simulation accuracy analysis and applicability evaluation were carried out on the four mixed dielectric constant models ofincluding physical Mironov model,semi-empirical Dobson model,semi-empirical Wang and Schumgge models,and empirical Hallikainen model from three aspects:soil texture,temperature,and frequency.That the optimal dielectric constant model should be used for different soil types in active and passive microwave inversion of soil moisture is proposed.For different soil texture conditions,the simulation of real and imaginary parts of soil dielectric constants has the highest accuracy,followed by the Haillikainen model,and the Dobson model,Wang and Schmugge models have poor accuracy.As far as the simulations of real component of soil dielectric constant for the above four models,the mean of statistic RMSE were 1.58,5.18,5.02,2.94,and the mean of slope were 1.17,1.46,1.52,1.43 and the mean of R were 0.998,0.994,0.998,0.985 respectiv.ely.For silty sandy loam soil,silty clay loam soil and silty clay soil,the simulation accuracy of the Mironov model decreases with the increase of silt content,and the Dobson model has the best simulation results.For different temperature conditions,the relative changes of the dielectric constants simulated by Mironov,Dobson,Wang and Schmugge in the non-frozen period have a good agreement with the measurements.However,all the four models are not applicable in the frozen period.Furthermore,the Haillikainen model is only suitable for simulation of soil dielectric constant at normal temperature.In different frequency conditions,the four models have good applicability to the simulation of soil dielectric constant,and the relative changes caused by the dielectric constants simulated by the four dielectric constant models with the frequency changes are basically consistent with the measured conditions,indicating that the four models are suitable for the simulation of the dielectric constant at different frequencies.the relative change in dielectric constant of the simulation is basically consistent with the actual measurement.In general,the Mironov model is the first choice for the L-band microwave inversion of soil moisture,and the Dobson model was choosedthe first choice for soils with unusually high levels of silt in silt loam soils,silt clay loam soils and silty clay soils.The study aboutResearch on soil moisture collaborative inversion model of active and passive microwave data in C-band and L-band based on brightness temperature downscaling.Firstly,using the AIEM model and Tor Vergata's discrete backscattering and radiation model,the relationship between V-polarized emissivity and microwave VV polarization backscattering coefficient in the bare soil area and vegetation area of C-and L-band was simulated respectively.The simulation results show a highly linear correlation between the L-band V-polarized emissivity and the C-band VV polarized backscatter coefficient,indicating that the linear relationship between the emissivity and the backscatter coefficient in the same band of C-or L-band is also applicable between the L-band emissivity and the C-band backscatter coefficient.Then,according to the passive microwave model and active microwave backscatter theory,the linear relationship between the backscattering coefficient and the reflectivity is constructed.At the same time,the active microwave polarization decomposition theory is introduced to decompose the active microwave backscatter coefficient,and to analyze the physical and mathematical relationships between surface scattering,two-plane scattering,and volume scattering in the active microwave backscattering mechanism.Based on this,the physical expressions of the linear relation coefficients are solved,and a novel single-temporary-instantaneous-state brightness temperature and space downscaling algorithm based on backscatter coefficients is developed.Besides,combined with the single-channel algorithm SCA-V for passive microwave soil moisture inversion,a soil moisture collaborative inversion model of active and passive microwave data in C-and L-band based on brightness temperature downscaling.Finally,the L-band radiance temperature and the AirSAR C-band backscatter coefficient data of PALS in the SMEX02 experimental area were used to perform downscaling of the brightness temperature and soil moisture retrieval.The inversion results were verified by the measured soil moisture data.The statistical correlation R reached 0.69 and the RMSE was 0.051 cm3·cm-3.The results showed that the model was reliable.The study about soil moisture collaborative inversion model of active and passive microwave data in C-and L-band based on soil moisture product downscaling.Firstly,the relation between microwave VV polarization backscatter coefficient and soil water content in bare soil and vegetation area of C-Band was simulated by using AIEM model and Tor Vergata discrete backscattering and radiation model respectively.The simulation results show that there is a strong linear correlation between the linear form of the backscatter coefficient and the soil moisture content.Then,according tobased on the linear relationship and taking into account the spatial heterogeneity within the coarse scale,a new type of downscaling algorithm for soil moisture products was developed.Combined with the single-channel algorithm SCA-V for passive microwave soil moisture inversion,the soil moisture collaborative inversion model of active and passive microwave data in C-and L-band based on soil moisture product downscaling was established.The C-band AirSAR radar data in the SMEX02 experimental area was used to downscale the soil moisture in the 4000m-low spatial resolution for he PALS L-band passive microwave inversion,and then the soil moisture inversion by the active and passive microwave remote sensing coordination was completed.Finally,the inversion results were verified by the measured soil moisture data.The statistical correlation Reached 0.71 and the RMSE was 0.053 cm3·cm-3.The results showed that the model was reliable.Inversion an of soil moisture at 3km spatial resolution based on spaceborne Sentinel-1 and L-band SMAP radiometer.Firstly,for the pre-processed Sentinel-1 SAR backscattering coefficient data(3km spatial resolution,C-band,VV-and VH-polarized),SMAP brightness temperature(9 km spatial resolution)and the corresponding ancillary data such as soil temperature,soil texture,and vegetation moisture,soil moisture inversion was performed based on the two kinds of active and passive microwave collaborative inversion model.Finally,By the 10cm surface soil moisture monitoring data ofn March 21st,April 1st,April 11th,April 21st,May 1st April 16th and May 11thApril 21st,20178,preliminary evaluation on soil moisture products of two regions from March 18th to April 6th,April 11th to April 19th and May 5th to May 13th in Henan Provincepreliminary assessments of soil moisture products based on two regions on April 16th and April 21st of 2018 in Henan Province were conducted.The results showed that the average RMSE of soil moisture measured data and inversion data based on soil moisture inversion model of brightness temperature downscaling was 0.041 cm3·cm-3,bias was-0.008 cm3·cm-3,ubRMSE was 0.040 cm3·cm-3,and R was 0.61.The RMSE of soil moisture measured data and inversion data based on soil moisture inversion model of soil moisture product downscaling was 0.043 cm3·cm-3,bias was-0.021 cm3·cm-3,ubRMSE was 0.037cm3·cm-3,and the R was 0.673.It shows that the two soil moisture inversion models proposed in this study are effective for soil moisture production combined with temperature data of C-band Sentinel-1 SAR and L-band SMAP,and have good accuracy.