Abstract:

Groundwater drought not only severely restricts the socioeconomic development of the region， but also brings great challenges to groundwater resource management. Therefore， in this study， taking Jiuquan Oasis as an example， six different probability distribution functions were selected to construct the standardized groundwater index （SGI） by analysing the monthly groundwater level data of 11 observation wells from 1985 to 2017， and the K-S test and the AIC criterion were used for the preference of the distribution functions. The results show that the Gev function and the Beta function show strong applicability in different time scales. Based on this， this paper adopts K-Means clustering method to classify the SGI of 11 observation wells into three categories： continuously declining type， declining-rising type， and mutation type between 2002 and 2006. The results of the decomposition show that the groundwater drought condition of the "Continuously declining" wells deteriorated from non-drought to moderate-severe drought， the “Declining-rising” wells had insignificant groundwater drought conditions， and the "Mutation between 2002 and 2006" wells maintained the groundwater drought condition at the moderate-severe level for a long period of time after experiencing the mutation. In addition， the seasonal peaks of SGI in groundwater observation wells were concentrated in December and February in winter and in March and May in spring. The results of the study provide a basis for the scientific management of groundwater resources in the Jiuquan oasis.

Key words: groundwater droughts, standardized groundwater index, K-Means clustering method, seasonal-trend decomposition using LOESS method, Jiuquan Oasis