Bivariate copula function-based spatial–temporal characteristics analysis of drought in Anhui Province, China

Abstract

With the changing global climate, the occurrence of drought has become more frequent and serious in Anhui Province, an important agricultural production area. It is necessary to investigate and detect characteristics of the frequent droughts, which can provide reasonable suggestions to reduce their adverse influences. In this study, temporal variability characteristics of the Standardized Precipitation Index (SPI) with different time scales were analyzed based on the daily precipitation data of 16 stations during 1961–2010. Drought characteristics, including duration and severity, were then defined using run theory based on SPI with a 3-month time scale. The most suitable distribution function was selected to define the univariate marginal distributions of drought duration and severity. Finally, the bivariate joint probability distributions for different drought partitions were assessed based on the copula function, and spatial distributions of their return periods developed. Results indicate the following: (1) the duration and severity of droughts in Anhui Province have been decreasing and the spatial distribution of return periods for drought duration and severity are similar. Average drought duration was between 3.29 and 3.69 months, increasing gradually from north to south, and average drought severity was between 2.11 and 3.54, increasing from northeast to southwest. (2) When mild droughts happen, the drought bivariate return periods were 3.78–6.56 years. The highest values of moderate, severe, and extreme droughts occurred in northern and southwestern areas of Anhui Province, and the lowest values occurred in the north-central and south-eastern areas.