Assessment of risk ranking for autumn drought in Guangxi province based on AHP and GIS

Abstract: Guangxi is one of the annual precipitation-rich regions of the country. But seasonal drought occurs very frequently because of spatio-temporal nonuniform distribution of rainfall. Seasonal drought has a rather large influence on the water resources, industrial production, and human life, especially on the agricultural production of Guangxi. The research there has been mostly aimed at some characteristics of a certain domain (such as agriculture) in the application .And some evaluation processes too depended on subjective experience or a simple formula, and did not comprehensively consider the influence to society's economy and environment of drought disaster, the vulnerability of the disaster bearing body, and the ability to prevent and reduce disaster. In order to strengthen the risk assessment and emergency management capability of the autumn drought disasters of Guangxi, an autumn drought risk assessment indicator system was established, and its indicators were determined according to drought risk, sensitivity of disaster environment, vulnerability of disaster bearing body, ability of disaster prevention and reduction, included rainfall anomaly, little rain days anomaly, topography, hydrographic net, Karst landform, population density, gross domestic product(GDP), arable area, pecuniary loss and real GDP per capita, by use of meteorological data, such as daily rainfall of 88 meteorologic stations, from 1961 to 2010, basic geographic information, consisting of 1:50 000 scale county boundary and hydrographic net, digital elevation model(DEM) (about 100m resolution), and Karst landform(about 1000m resolution), and socio-economic data, containing population density, GDP, real GDP per capita, arable area and pecuniary loss, taking the county as a unit. Then factor weights were obtained by an analytic hierarchy process (AHP), which relied on a judgement matrix and its eigenvalues and eigenvectors, and a consistency test of the matrix, and comprehensive assessment models for agriculture and social economy were established and calculated to get the autumn drought disaster risk index, which ranked distribution by a geographical information system (GIS).The distribution indicated that the higher risk area contains the west of Chongzuo, central and south of Baise, east of Hechi, south of Liuzhou, east and south of Guilin, central and northwest of Laibin, urban of Guigang, and some counties of Nanning. The lower risk area contains the northern mountainous area of Baise, Hechi, Liuzhou and Guilin, most of Fangcheng and Beihai, central and south of Yulin, and most of Pingnan, Guiping, Zhaoping, and Mengshan. Finally the condition of drought disaster was used to validate the distribution. The correlation coefficients were 0.58441 and 0.60393, respectively, of agriculture and social economy, through significance test by 0.05, using the correlate analysis method, by the multivariate analysis tool of ARCGIS. The results showed that the distribution of autumn drought disaster risk is basically consistent with the spatial distribution of drought disaster losses, which in the high risk area are mainly distributed in the middle basin and mountain area in the west, and are low in the mountain areas in the northwest and north, coastal area in south, and parts of the southeast of Guangxi. The results reflected preferably regional differences of drought risk, which were due to the distinction of formation, environment, bearing body and prevention of autumn drought disaster of Guangxi. The drought assessment model and method used combine the advantage of AHP and GIS, and can make drought evaluation procedure and evaluation results more scientific, increasing practicability and maneuverability.