The Construction Of Simplified Evapotranspiration Stress Index And Its Application For Agricultural Drought Monitoring

On the background of Global Climate Change, the frequency and intensity of agricultural drought disasters presents an increasing trend, which challenges China’s agricultural production. Meteorological drought indices are generally applied in traditional drought monitoring services. Regional evaportranspiration models based on remote sensing method, by contrast, provide technology support for agricultural drought monitoring at large scale.Based on the Priestley-Taylor equation and land surface temperature-vegetation index (LST-Ⅵ) feature space, simplified Evapotranspiration Stress Index (SESI) is derived by semi-empirical evapotranspiration model in this study. MODIS land products and soil moisture observations were used for verification, in comparison with Temperature Vegetation Dryness Index (TVDI), and applicability of SESI was further evaluated. This thesis provides a new agricultural drought index combining mechanism and practicality, which has practical significance for drought monitoring in China’s agrometeorological service. The results are as follows:(1) Three different schemes, LST-NDVI, LST-EVI, LST-Fr, were applied in feature space modeling. Scatter diagrams and the results of dry-edge fitting showed that LST-NDVI feature space is relatively more stable during the whole study period, which is appropriate for long-term continuous application, while LST-EVI and LST-Fr feature space are only appreciate for summer modeling at the plain areas of the Beijing, Tianjin and Hebei, because of the abundance of foliage.(2) According to verification results based on relative soil moisture from two depth (10cm and20cm, both SESI and TVDI have favorable indicative function of surface soil layer moisture. The results of correlation analysis showed SESI had better performance than TVDI in general, although significant test results of both were a<0.01. In addition, better correlation was found under winter wheat growing and cotton growing condition than maize growing condition.(3) The temporal and spatial distribution of SESI, compared with meteorological record, demonstrated this drought index was feasible scheme for agricultural drought monitoring, and its comparability during different periods was better than TVDI. However, insufficient monitoring capability of SESI was revealed in summer.(4) Despite the comparability during multi-period, it is necessary to differentiate periods, regions and crops in further research and application, considering the temporal change of climate condition, the spatial variation of spectral signature of underlying surface, and the diverse requirement of water supply between different crops.