Application of the Projection Pursuit Technique in Assessment of Agricultural Drought Vulnerability in Dalian

Drought can directly affect agricultural output and food shortages. Its continuous accumulation can result in significant land degradation, water resources depletion, and environmental damage, which can restrict the sustainable development of agriculture. The status of agricultural production is prominent in Dalian. It often suffers from drought disaster, which has restrained agricultural development in recent years. Unraveling characteristics of agricultural drought of Dalian is therefore critical. The authors synthetically considered mechanisms of various interactions of agricultural drought, introduced reduced-order thoughts of projection, and built an agricultural drought model by the Projection Pursuit Technique to evaluate agricultural drought vulnerability of Dalian. A vulnerability evaluation index system was built, including population density, agricultural output value proportion, wet area proportion, precipitation, sowing area ratio, the agricultural population proportion, food production per capita, irrigation index, net income of farmer average per capita, and agricultural fertilizer quantity per unit area. We then selected the best projection direction, took characteristic quantities of direction as the mete yard, and evaluated the agricultural drought vulnerability of seven agricultural areas in Dalian based on the panel dataset during the period 2000-2007. Characteristics of variations in spatial-temporal differences of vulnerability assessment were comprehensively analyzed. Then, the zonal map of agricultural drought vulnerability of the study area was drawn. The following conclusions were drawn. The average values of agricultural drought vulnerability assessment in descending order are Zhuanghe City, Pulandian City, Wafangdian City, Jinzhou District, Lvshunkou District, Changhai County and Ganjingzi District. The agricultural drought vulnerability of the north is higher than the south and central Dalian. Results show that the Projection Pursuit Technique can give appropriate assessment of agricultural drought vulnerability which involves multiple factors, and provide a new way for decision-making. In the meanwhile, the developed model seems to be reliable and feasible, and it would be an effective technique to evaluate agricultural drought vulnerability over other regions.