Soil Moisture Retrieval Based On AMSR-E And Modis Data Fusion

As an important state parameters of the global energy balance, soil moisture is an important part of global water circulation. It has become a frontier topic in academic circles by using soil moisture monitoring drought, floods and other extreme disaster. Traditional soil moisture monitoring is the field based on data points. These methods have higher precision, but time consuming and can’t satisfy the need of large scale, long time period, the real-time dynamic monitoring. With the development of 3S technology, using remote sensing(RS) technology for soil moisture monitoring has become a hot research at home and abroad. At present the use of a single remote sensing data can be retrieved soil moisture, but the effect and accuracy still needs to be improved, resulting in special cases of missing data, or affected by the weather, the inversion precision lower. Multi-source remote sensing data fusion is to provide different remote sensing image data source information to be integrated, in order to obtain high quality information. In this paper, taking Huaihe River Basin as the study area,the moderate resolution imaging spectroradiometer MODIS data collected in 2006-2010 with all day,all-weather, and reflect the surface condition,some penetrating properties of microwave remote sensing data of AMSR-E surface temperature characteristics on the surface, as a method of wavelet fusion using environment in the MATLAB industry, and fuse the data. From the time and space correlation analysis fusion inversion data and source data, measured soil moisture and precipitation data, the following preliminary conclusions:Five different frequency data of AMSR-E (Tb6.9, Tb10.7, Tb18.7. Tb36.5, Tb89.0) and the two kinds of products of MODIS (MOD11A2,MOD13A2) in the inversion of soil moisture in Huaihe basin,Tb6,9 (i.e.,frequency of 6.9GHz) is the highest correlation with the measured data, and it shows the best effect. It is mainly due to the Tb6.9 frequency penetration strong, anti-radio interference.Soil moisture inversion effect of AMSR-E and MODIS data based on wavelet transform fusion is better than that of soil moisture retrieval from the single remote sensing data.It is more sensitive based on the MODIS and AMSR-E fusion data at each turn of the season on soil moisture changes, especially in spring, summer and autumn.Based on MODIS and AMSR-E fusion data at each turn of the season on soil moisture changes more sensitive, especially in spring, summer and autumn. Tb6.9 SM data varies with the cumulative change in rainfall data size, this is mainly due to AMSR-E data in versing of surface 1cm-2cm of soil moisture. However, soil moisture between 1cm and 2cm is very sensitive to atmospheric precipitation. In the aspect of the precipitation time series, Tb6.9 SM data have certain advantages, but it also can’t be the only basis to determine the condition of soil water of the study area. In the inversion of soil moisture changes, Tb6.9 SM trend is too gentle, while based on AMSR-E and MODIS fusion data can be closer to the measured data, the fusion date performs better effect, because MODIS data is better than the AMSR-E data in the inversion of deep soil moisture.The inversion of soil moisture precision is different in the Basin of Huaihe. From the time series, high inversion precision is the spring of 2009 and 2010 of spring and winter: from the spatial series, high inversion precision are Zhumadian and Bozhou. the worst is Huaiyin and Shangqiu. The study found the inversion precision difference is mainly due to:AMSR-E or MODIS missing data, resulting in the inversion accuracy reduced: topography, vegetation coverage and precipitation and other caused by factors such as the remote sensing data of actual soil moisture inversion uncertainty.In this study, taking Huaihe River Basin as a study area, soil moisture retrieval from AMSR-E and MODIS data fusion based on single source data inversion results than traditional effect is good, and it has use value and good application prospect of drought monitoring of multi-source remote sensing data fusion, but we found that the quality of data fusion effect by the source data resolution, the study area terrain, weather effects. If we use terrain factor, the higher resolution satellite data fusion, inversion effect would be better.