Abstract: Drought Monitoring using remote sensing technology is one of the most useful methods. Medium Resolution Spectral Imager(MERSI) on board the new Generation Polar-Orbiting Meteorological Satellite of China(FY-3) has 15 spectral channels with 1 km spatial resolution,which enhances the ability to observe fine surface features and provides a new data source for drought monitoring in large area. Satellite images from NOAA/AVHRR,Landsat/TM,TERRA/AQUA-MODIS have been the major data sources used in China's drought applications. Of all these satellite sensors,data from TERRA/AQUA-MODIS has moderate spatial resolution and high spectral and temporal resolution,which is especially favorable in validation of known drought indexes. In this paper,drought conditions of May,July and August in 2014 in Shaanxi Province,China were studied based on FY-3C/MERSI data and MODIS MOD09GA(daily reflectance products) and MODIS MOD11A1(daily land surface temperature products) of the same period,by building the Multiple Band Drought Index(MBDI) using the Normalized Difference Water Index(NDWI) and LST(Land Surface Temperature) indexes,comparing the correlation of MBDI results of MERSI and MODIS,taking correlation analysis of MBDI retrieved by MERSI and MODIS with 10 cm surface soil moisture data,comparing MBDI monitoring spatial distribution map between MERSI and MODIS,and analyzing the change trend of pixels under different drought degrees. The results show as follows:(1) it shows significant positive correlation between MERSI-MBDI and MODIS-MBDI,with R² of all three periods of drought process over 0.7,and the correlation coefficient got through 0.001 level test,indicating there is a good consistency between MERSI-MBDI and MODIS-MBDI;(2) it shows significant positive correlation between MODIS-MBDI and 10 cm surface soil moisture data,which all got through 0.05 level test;MERSI-MBDI has better correlation with 10 cm surface soil moisture data,which all got through 0.01 level test;(3) the retrieval of spatial and temporal characteristics of drought using MERSI-MBDI and MODIS-MBDI has a good consistency, which was also tally with the actual situation of the study area;pixel numbers under different drought degree was also basically identical with the trend. The results indicate that MERSI data can be greatly used in drought monitoring in Shaanxi Province,and FY-3C/MERSI could apply to continuous drought monitoring in large area,as a new generation of polar orbit meteorological satellite data in our country,according to the comparative analysis with MODIS-MBDI results. The conclusion has important significance for promoting satellite remote sensing application ability in China,reducing reliance on foreign satellite remote sensing data,and providing drought monitoring timely when drought occurred.

Key words: FY-3C/MERSI, MODIS, multiple band drought index