Research Of Summer Soil Moisture Retrieval Model In Henan Province

Drought is a complex natural disaster that occurs frequently on a global scale, which has a great impact on the global natural ecological environment and human socio-economic activities, and the drought disaster has been one of the hot issues. Henan Provinces, as the main crops producing province of China, located in the North China Plain, the annual distribution of precipitation is uneven by the impact of the monsoon, the impact of drought happed more frequent and serious in the summer. Soil moisture is one of an important indicator of drought, There is a strong nonlinear coupling between soil moisture and various influence factors, it is necessary to take full account of the various factors that affect the monitoring and to establish a monitoring model that fits the actual needs. Hence this paper takes Henan Province as the main research area, combined with the 2007-2012 summer MODIS(Moderate-Resolution Imaging Spectroradio meter) surface temperature products, surface reflectance products, Land cover type produce, and in-suit data, the soil moisture retrieval model was conducted by neural network algorithm and TVDI index, respectively. The main achievements of this paper are as follows:(1) This paper using the reflectance bands of MODIS produce which is reflecting changes in soil moisture, combine with several important soil moisture influencing factors: band surface temperature band, land cover type and DEM, and in-suit data a researched of soil moisture retrieval model which based on Neural Network Algorithm and TVDI Index was conducted,Establishment of Summer Soil Moisture Retrieval Model in Henan Province.The two inversion models can objectively reflect the spatial distribution of soil moisture, But the two methods have different value of retrieval outcome,TVDI index method has lower soil moisture inversion value.(2) An analysis of the relationship between the two inversion models and the measured soil moisture was conducted, founded that the soil moisture inversion model based on the neural network algorithm is more accurate than the TVDI index inversion model, It is shown that the algorithm model can better describe the spatial distribution of soil moisture.(3) Based on the soil moisture model retrieved by neural network algorithm, the distribution pattern of summer drought disaster in Henan Province in 2012, summer was established. The result shows that the summer drought in 2012 was less severe, while the disaster is more serious in June,the drought disaster of August were lighter, nearly no drought disaster happed in July. And the frequency of summer drought is in northern, western and the Central of Henan province in the higher frequency of drought.The research work above has established a soil moisture retrieval model which can invert high temporal resolution, and achieves the fine expression of soil moisture.