Quantitative Risk Assessment of Agricultural Drought in the Jinghe Watershed of Western China
泾河流域农业旱灾风险综合评估研究

Though drought is a recurrent phenomenon in the Jinghe watershed, very little attention has been paid to the mitigation and preparedness of droughts in this area. Current research on risk or vulnerability of agricultural drought are mostly based on administrative units and ignore the interior differences between administrative units. This paper presents a method for the spatial assessment of agricultural risk in the Jinghe watershed of western China at a 1km grid scale. A conceptual framework which emphasizes the combined roles of hazard and vulnerability in defining risk was used, and the Z index method in a GIS environment was used to map the spatial extent of drought severity hazard using monthly precipitation data between 1957-2009 from the China Meteorological Data Sharing Service System. The key social and physical factors that define agricultural drought vulnerability in the context of the Jinghe watershed are climate, soil, irrigation and topography, respectively indicated by seasonal crop water deficiency (SCWD), available water capacity (AWC), irrigation percentage of the whole area of county-level and slope. Agricultural risk was computed by the integration of drought severity and vulnerability. The agricultural severity calculation results showed that 10.8% of the total Jinghe watershed was grouped into the high severity class, 71.8% into the moderate class and 17.4% into the low class. Vulnerability calculation results showed that 35.3% of the total arable land area falls into the high vulnerable class, and 51.7% into the moderate vulnerable class and 13.0% into the low vulnerable class. Agricultural drought risk calculations indicated that 40.5% of the total arable area suffered high risk and 32.2% and 27.3% was grouped into moderate and low risk classes. Generally, the northern Jinghe watershed falls into high risk class, the middle part into moderate risk, and the southern part into low risk; however, spatial variation occurred across the eastern part of the study area as there were scattered spots classified as low risk here.