中文摘要：

基于中国0.5°×0.5°逐日气温和降水格网数据,利用复杂网络分析方法,对2008年南方低温雨雪冰冻灾害综合致灾过程进行再认识,综合分析低温雨雪冰冻灾害在时空维度的网络特性。结果表明：2008年南方低温雨雪冰冻灾害是典型的多灾种叠加事件,低温与雨雪灾害叠加放大了致灾因子的危险性;基础设施设防水平低与春运高峰叠加增大了承灾体的脆弱性;低山丘陵区与人口聚集区叠加降低了孕灾环境的稳定性。低温冰冻雨雪灾害具有小世界特征和核心—边缘结构,具体表现为：在空间打击上具有集聚性,影响区域相对集中;在时间打击上具有连续性,间隔1天事件相对较少。在研究方法上,复杂网络是一种有效分析多灾种叠加的方法,可以进一步挖掘自然灾害的时空演化信息。

英文摘要：

In recent years, some new areas of development have been proposed for natural disaster science. It is suggested that the discipline should try to attach more importance to recognizing disaster loss accumulation and amplification trends, re-evaluate the mechanisms of major natural disasters, and facilitate interdisciplinary research between natural disaster and complexity science. The 2008 Chinese ice storm struck the most populated and economically developed region of China, and direct economic losses exceeded $22.3 billion（indirect losses may have been even greater）. Recognition of the formation mechanism of the 2008 Chinese ice storm will be very important for developing integrated disaster risk reduction strategies. This study employed daily temperature and precipitation data from a 0.5° × 0.5° gridded dataset released by China＇s National Meteorological Information Center to evaluate the synthetic hazard process of the severe cold surge, ice-snow events, and corresponding frozen disasters in southern China in 2008. More specifically, we analyzed the spatial-temporal structure and complexity of the catastrophes by using bipartite networks. Results indicated that the large freezing rain and snow disaster was a typical multi-disaster overlaying event. The overlaying of freezing rain and snow amplified the risks of other hazards such as critical infrastructure failures. Moreover, the timing of the event, which occurred during the Chinese New Year,increased the vulnerability of people to exposures（i.e., they were more likely to be traveling）.Low mountain-hilly terrain and high population density reduced the stability of the fostering environment. The freezing rain and snow catastrophes in 2008 showed the mathematical characteristics of small-world features. In terms of spatial data, the occurrence of hazards was clustered when the unprecedented ice struck the region; in terms of temporal data, the intervals between successive waves were too short for ice to melt or be broken, thus disaster losses were