A Research On The Arid Index Of Agricultural Disaster Adaptability And Its Extreme Value Distribution During The Winter Half Year In Southwest China

Drought is a kind of disaster phenomenon that nature affects human society. In recent decades, the global aridification trend has been aggravating. So has our country’s agricultural drought disaster. Generally, drought can be divided into meteorological drought, agricultural drought, hydrologic drought and society drought. The correlation of these four types of drought is very complicated. Among them, meteorological drought is an important factor which can cause other types of drought. And the connection between meteorological drought and other types of drought has been given considerable attention by academic field and operating departments. This thesis mainly choosing southwest China as the research area, adopts, the two methods, lead Correlation Similarity and Skill Score, for the selecting of the most suitable drought index related to the agricultural disasters, and study the agricultural drought disaster risk caused by meteorological drought, based on a variety of meteorological drought indexes and agricultural drought data. Samples of drought index’s selecting, models of drought index extreme value’s establishing and extremum model parameters influenced by the ENSO’s diagnosis are acquired by the relation between the most suitable drought index and the risk of agricultural drought occurrence. The main conclusions are as follows:Z index that has a good respond to the relation between meteorological drought and agricultural drought in southwest China is the most suitable drought index to evaluate the risk of agricultural drought occurrence. The risk of the agricultural drought disaster rate more than eight percent and the agricultural drought crop failure rate more than five percent can respectively aggregate0.53and0.37or above, and the highest risk area is located in the northwest of Guizhou province, when Z Index<-0.84in the winter half year.By using Z index≤-0.84, to determine the spatial distribution of running days, when the rainfall of various stations in southwest China comes to the thresholds. The distribution shows that the maximum appeared in Songpan, Sichuan for133days, and the minimum appeared in Yuxi, Yunnan for31days. The parameters, including the position parameters k and the scale parameters β, acquired by GEV simulation, when the each year’s longest running days of meteorological drought in various stations are selected, have significant spatial differences. The maximum of K appeared in Ruo Ergai, Sichuan for157.22, and the minimum appeared in Yuxi, Yunnan for62.04. And the maximum area of β appeared in the area of Zhanyi, Yunnai for32.11. The running days of continuing drought will be longer in Sichuan, Chongqin and the northwest area of Guizhou, and the risk of drought during winter half year will be bigger, when the reappearance period of meteorological drought (Z index≤-0.84) in southwest China is certain.The differences of GEV distribution parameters in high value year and low value year of SOI, mainly manifest in the southeast of Guizhou, the southeast of Yunnan and the Hengduan Mountain Range areas. When the SOI is low, both the value of the scale parameter (3and the location parameter k are high. And the risk of agricultural drought is bigger. The agricultural disaster data show that the agricultural drought occurs in four years, accounting for80percent of the five years, from1984to2007, during which the value of the SOI is low (El Nino). From the data, we can conclude that the SOI with the low value can indicate the occurrence of agricultural disaster in southwest China.