Abstract: Traditionally,drought monitoring models based on evapotranspiration do not consider the weather influence in the calculating of solar radiation. In fact,clouds have a significant impact on the solar radiation,so that the calculated evapotranspiration results have great errors in cloudy weather,which reduce the drought monitoring accuracy. Because of the lag of vegetation to precipitation,the traditional remote sensing drought model based on temperature and vegetation index has the hysteresis,and the remote sensing model without considering the differences between crop types and only relying on the vegetation index derived by sensor,and the response mechanism of different crops to drought varied widely,all of these are not reflected by the remote sensing data, so that the monitoring results of drought are always unsatisfactory. In this paper,SEBS model is applied to localization, and the time series expansion method and the pixel accuracy were improved,the land surface parameters with high spatial accuracy and real-time meteorological elements of high temporal precision are fully utilized,at the same time,the FY-2E data is used to calculate the objective ground solar radiation. This paper monitored drought based on improved SEBS model and Crop Water Stress Index(CWSI) in Shanxi Province with remote sensing date from FY-3A/3B and FY-2E satellite,and meteorological date from automatic weather stations,agricultural meteorological stations and soil moisture stations. The results showed as follows:(1) CWSI is influenced by topography and vegetation,the drought is underestimated in lush vegetation area,and overestimated in city area;(2) The accuracy of CWSI for drought monitoring reached more than 80%,can meet the precision requirements of the daily work;(3) CWSI and soil moisture and precipitation showed a consistent distribution,and CWSI was proved to be more timely for drought monitoring. The perfect combination of data from meteorological observation stations and from remote sensing data effectively improved the accuracy of the monitoring results in space and time,it overcome the shortage of the general model.

Key words: drought monitoring, SEBS model, crop water stress index