Research On Collaborative Forest Surface Soil Moisture Inversion Method Of Landsat 8 And GF3 Image Data

Soil moisture is an important part of soil parameters research.With the advancement of science and technology and the development of the times,the enriched remote sensing modeling and the increasingly mature remote sensing inversion theory make remote sensing technology an important means of estimating soil indicators.Combined with remote sensing satellite data,real-time dynamic large-area monitoring and estimation of surface soil moisture information in the area can be used.At present,the research focus of scholars at home and abroad has shifted to the study of soil moisture in multi-source remote sensing data.Daxing'anling area is an important forestry base in China.The forest farms in the study area have dense virgin forests.The forest soil moisture serves as the carrier of the ecosystem cycle,and has a complex relationship with the local topography,soil and vegetation cover types.In recent years,researchers from home and abroad have conducted a large number of experiments on remote sensing inversion of soil moisture,and there have been many breakthroughs.However,the research scope is mainly concentrated in areas with low vegetation coverage such as bare land,or related research on predicting the ecological status of arid areas and sparse vegetation coverage areas supporting agricultural production,and the soil moisture research on forest cover.Less,so it is necessary to study the inversion of surface soil moisture under high vegetation cover.In order to eliminate the influence of vegetation and surface roughness on surface soil moisture inversion results in the area with high vegetation coverage in the study area,Landsat 8 OLI/TIRS image is selected as the optical data source,which satisfies the improved temperature vegetation.The calculation of the drought index model is in line with the requirement of soil moisture inversion accuracy under the cover of forest vegetation,and assists in the correction of DEM data.Firstly,the Landsat 8 remote sensing image data is used to invert the surface temperature of the study area,calculate NDVI,NDMI,vegetation coverage and other parameters,establish the feature space(TVDI)of surface temperature and vegetation index,and obtain the dry and wet feature space.The edge equation is used to compare the parameters with better correlation into the regression model for inversion of soil moisture,and then the soil moisture in Mohe City of Daxing'anling area is inverted,and the soil moisture is verified by the ground measured data obtained from field sampling.The accuracy of the inversion results.In order to improve the accuracy of the model,the microwave remote sensing data is used to collaboratively retrieve the soil moisture.The radar data source is the high-resolution 3 remote sensing image.The classical vegetation scattering model is used to improve the water cloud model.Part of the measured data is used for modeling,and the remaining observed soil.The moisture data values were used to verify the accuracy of the results of the collaborative inversion model.Finally,the two improved inversion models were evaluated.The main results show that:(1)For the soil moisture inversion of vegetation cover surface,the NDMI extracted by Landsat 8 data has better effect on the removal of vegetation;(2)Inversion of the geothermal temperature of the study area was carried out by comparing the method with higher precision radiation transfer equation method.Based on the land surface temperature(Ts)and NDMI obtained from this,the TVDI improved by terrain correction was constructed.The characteristic model has a good negative correlation with the measured soil moisture and TVDI inversion values;(3)Using two polarization methods to eliminate the influence of the vegetation layer,it can be seen that after removing the influence of the vegetation layer,the backscattering coefficient is numerically reduced,compared to the HV polarized high-scoring satellite No.3.Data,HH polarization data can better reflect soil moisture conditions,and can obtain higher soil moisture inversion accuracy;(4)Collaborative high-scoring satellite data and Landsat 8 image data can obtain higher precision soil moisture inversion results.Using this model to invert the accuracy of surface soil moisture under high forest vegetation coverage is more accurate than single optical remote sensing inversion.The correlation coefficient between the soil moisture value obtained by the model and the soil moisture value obtained by the field measurement is 0.668,and the root mean square error is 0.028.The results show that the model has better inversion effect and can be applied to future vegetation cover.Inversion of soil water content in the area.