Drought risk assessment in complex landform area

Abstract: This article intended to achieve drought risk assessment in complex topography area, such as Sichuan Province, in the support of natural disaster risk theory and multi-source data, especially satellite data which was widely used to reflect the properties of study area from different angles including meteorology, hydrology, topography. In detail, TRMM satellite precipitation data was used to make up for the shortage of data from meteorological stations. And the latitude, longitude and digital elevation model were introduced to improve the accuracy of temperature interpolation; furthermore, drought frequency instead of drought degree was calculated using Normalized Vegetation Index data to more accurately quantify drought risks. Meanwhile, an index of water acquisition cost was constructed to quantify the affections of terrain on water distribution and acquisition. In addition, raster data were used as far as possible, such as raster populations and economies. Thus, an index system which revealed variable properties of study area in different angles was constructed. And then, weights of the index system were calculated using entropy method. Finally, the drought risk in Sichuan Province was assessed. The accuracy of this assessment was tested by agricultural drought risk loss. It showed a high correlation between the agricultural loss and the drought risk achieved from this paper (R2=0.8347, P<0.001). Thus, the reliability of this research in aspect of drought risk assessment in complex topography areas was proved. Through the analysis of drought risk assessment result, it showed that: 1) Most of Sichuan Basin, part of Panzhihua and Liangshan showed high drought risks; 2) Northern Guangyuan, Bazhong, Dazhou, most of Liangshan and Panzhihua, southern Yibin and Luzhou reached average risk degrees, as well as some valleys in the Western Sichuan Plateau; 3) However, there was quite low risk in most of the Western Sichuan Plateau. Through this study, it proved that the index system and method in drought risk assessment had some reference significance especially in complex topography areas. At the same time, there were some shortages which needed to be improved. 1) Using Entropy method, statistical characteristics of data was fully taken into consideration, and weights of the indexes were determined according to the variation of each index. However, it failed to distinguish the importance of each index. To improve the accuracy of weight determination in further study，it was best to combine the objective method and subjective method. 2) The spatial resolution was greatly improved by introducing a large number of spatial data into study, which was a great breakthrough in drought risk assessment, for most studies just took administrative region as the basic unit. However，uncertainty was actually inevitably increased because of the accuracy problem of spatial data itself. 3) According to the requirements of natural disaster risk theory, 4 factors and 11 indexes were selected to construct the assessment system, which basically met the need of drought risk assessment in the study area. But the assessment system may need a further refinement and improvement in order to be able to reflect the complex background information of the complex topography area in all aspects more accurately and comprehensively.