Comparision Of Three Drought Indexes Based On Gridded Dataset And Judgment Of Drought Trend In The Loess Plateau Of Gansu Province

The frequency of extreme climate events havean increasing trendunder the context of global warming, and the amount ofnatural disaster is also increasingaccordingly.In addition, the severity of the drought disasters, losses and influence in the natural disasters is the most significant. At present, the awful drought has an impact on production and living of people, and becominga heated theme around the world. The precipitation amount of 0.5°×0.5° gridded dataset(V2.0) is established by the national meteorological information center of China, and three drought indexesthat the percentage of precipitation anomaly, Z index and standardized precipitation index(SPI) are employed to evaluate their credibility in the loess plateau ofGansu province, moreover, some analysis methods are employed, such as the climate trend rate, Inverse distance weighted interpolation and commensurability information to discuss the drought trend in future. The key consequences are as follows:(1) The analysis results from the inter-annual and seasonal drought features by three drought indexes show that the drought trend is enhancing, and the same trend can be found in spring and autumn, and the opposite trend of wetting trend exists in winter, but any variation trend be found in summer in recent 53 years.(2) Result of three drought indexes showed the similar spatial features, especially the Z index and SPI. The wetting trend is significant in winter, but in other seasons, the drought trend gradually intensifies.(3) Some differences can be obtained in the drought frequency of three indexes, but the drought frequency of heavier drought is underestimated by Pa index, next is SPI and Z index.(4) In stark contrast to the actual drought years, three drought indexes all underestimated the drought level, by contrast, Z index is more accurately in three ndexes. By comparing the typical drought record with three drought index, we find that Z index is more accurate than other two indexs in the spatial drought distribution of 1962 and the spring drought in 2000.(5) The next possible year that appearing moderate and above drought is 2017 recently, the stochastic probability occupy 25%, and the next spring drought may be 2015, 2017 or 2024, the stochastic probability is 60%, 50% and 70%, respectively, in addition, the next autumn drought may be 2016 or 2020, the stochastic probability respectively occupy 43%and 71%.