College of Geographical Sciences, Shanxi Normal University, Taiyuan 030031, China.
Key Laboratory of Geographical Processes and Ecological Security of Changbai Mountains, Ministry of Education, School of Geographical Sciences, Northeast Normal University, Changchun 130024, China.
Int J Environ Res Public Health. 2023 Feb 24;20(5):4074. doi: 10.3390/ijerph20054074.
Droughts are widespread in China and have brought considerable losses to the economy and society. Droughts are intricate, stochastic processes with multi-attributes (e.g., duration, severity, intensity, and return period). However, most drought assessments tend to focus on univariate drought characteristics, which are inadequate to describe the intrinsic characteristics of droughts due to the existence of correlations between drought attributes. In this study, we employed the standardized precipitation index to identify drought events using China's monthly gridded precipitation dataset from 1961 to 2020. Univariate and copula-based bivariate methods were then used to examine drought duration and severity on 3-, 6-, and 12-month time scales. Finally, we used the hierarchical cluster method to identify drought-prone regions in mainland China at various return periods. Results revealed that time scale played an essential role in the spatial heterogeneity of drought behaviors, such as average characteristics, joint probability, and risk regionalization. The main findings were as follows: (1) 3- and 6-month time scales yielded comparable regional drought features, but not 12-month time scales; (2) higher drought severity was associated with longer drought duration; (3) drought risk was higher in the northern Xinjiang, western Qinghai, southern Tibet, southwest China, and the middle and lower reaches of the Yangtze River, and lower in the southeastern coastal areas of China, the Changbai Mountains, and the Greater Khingan Mountains; (4) mainland China was divided into six subregions according to joint probabilities of drought duration and severity. Our study is expected to contribute to better drought risk assessment in mainland China.
干旱在中国广泛发生，给经济和社会带来了相当大的损失。干旱是复杂的随机过程，具有多属性（如持续时间、严重程度、强度和重现期）。然而，大多数干旱评估往往侧重于单变量干旱特征，由于干旱属性之间存在相关性，这些特征不足以描述干旱的内在特征。在本研究中，我们使用标准化降水指数，利用 1961 年至 2020 年中国的月格点降水数据集来识别干旱事件。然后使用单变量和基于 copula 的双变量方法来研究 3 个月、6 个月和 12 个月时间尺度上的干旱持续时间和严重程度。最后，我们使用层次聚类方法来识别中国大陆在不同重现期的干旱多发地区。结果表明，时间尺度在干旱行为的空间异质性中起着重要作用，如平均特征、联合概率和风险区域化。主要发现如下：（1）3 个月和 6 个月时间尺度产生了可比的区域干旱特征，但 12 个月时间尺度则不然；（2）较高的干旱严重程度与较长的干旱持续时间相关；（3）干旱风险在新疆北部、青海西部、西藏南部、中国西南部和长江中下游较高，而在中国东南部沿海地区、长白山和大兴安岭则较低；（4）根据干旱持续时间和严重程度的联合概率，将中国大陆分为六个亚区。我们的研究有望促进中国大陆更好的干旱风险评估。
