Retrieving Soil Moisture From FY-3Satellite Data

Soil moisture is the quantity of water contained in soil,which is an important indicator for monitoring soil degradation and drought.it is also one of essential parameters in the study of hydrology, meteorology, soil science, ecology and agricultural sciences. Currently there are a lot of soil moisture inversion method based on remote sensing technology at home and broad, using mainly concentrated in the optical wavelength band (visible-near infrared, thermal infrared) and microwave bands. The four main method are visible-near infrared soil moisture monitoring method, thermal infrared soil moisture monitoring method, visible and near-infrared and thermal infrared soil moisture monitoring, microwave remote sensing of soil moisture.The FY-3series is the second generation of chinese sun-synchronous meteorological satellite,The first of this series was developed and launched from the Taiyuan Satellite Launch Centre in China on27May,2008. It is equipped with both sounding and imaging payload,providing three-dimensional,quantitative,multi-spectrum global remote sensing data under all weather conditions. Its satellite data can be used to study the inversion of surface soil moisture.In order to take full advantage of domestic satellite data, this paper applies FY-3B/MWRI data develop a microwave soil moisture inversion algorithm. Simultaneous analysis FY-3A/MERSI visible-near-infrared data quality and drought monitoring capabilities, and inversion the relative humidity of the soil in Hebei use TVDI index. Main conclusions are:(1) In this paper, the singlefrequency dual polarization algorithm was used to study the bare soil moisture retrieval based on FY-3B/MWRI data. The estimated soil moisture were evaluated by measured data. The results show that the determination coefficients between the estimated and measured soil moisture is0.547,the RMSE is0.0685cm3/cm3, the retrieved soil moisture is in good agreement with the measured data. However, the measured value is less than the inversion results in high values of soil moisture, and thus through the measured soil moisture data revisions to the original model.(2)Using another group of FY-3B/MWRI data to test the accuracy and stability of the revised model. The RMSE between measured data and original model inversion results is0.0638cm3/cm3, The RMSE between measured data and correction model inversion results is0.0361cm3/cm3, The correction model to obtain a higher retrieval accuracy, can be used for soil moisture retrieval of bare soil areas. The results also show that the use of measured data to revision inversion model is necessary.(3)The NDVI and LST are calculated in the study area based on MERSI data and comparative with MODIS related products. The results show that exist significant correlation between MERSI NDVI、LST and MODIS products.(4)There is a very similar pattern of spatial distribution between MERSI-TVDI and MODIS-TVDI.These index also have a good correlation with measured data and accumulated precipitation. This indicates that the drought monitoring results from MERSI is credible, the MERSI-TVDI can be used as indicators of soil drought monitoring.(5) The RMSE between measured data and MERSI inversion results is0.145. Compared with the MODIS data inversion results, the MERSI data is more accurate.Its indicate that MERSI data quality is reliable, can be used for soil moisture retrieval in Hebei Province.