地理研究 ›› 2021, Vol. 40 ›› Issue (8): 2314-2330.doi: 10.11821/dlyj020200894

• 研究论文 • 上一篇    下一篇

多灾种时空耦合网络构建:从多维网到单顶点网

何锦屏(), 李双双()   

  1. 陕西师范大学地理科学与旅游学院,西安 710119
  • 收稿日期:2020-09-18 接受日期:2020-11-20 出版日期:2021-08-10 发布日期:2021-10-10
  • 通讯作者: 李双双(1988-),男,陕西潼关人,博士,讲师,硕士生导师,主要研究方向为全球变化与区域灾害防治。E-mail: lss40609010@126.com
  • 作者简介:何锦屏(1998-),女,湖南平江人,硕士研究生,主要研究方向为全球变化与区域灾害防治。E-mail: JPH1116@126.com
  • 基金资助:
    国家自然科学基金项目(41701592);国家自然科学基金项目(41877519);陕西师范大学研究生创新团队项目课题(TD2020035Y)

Spatiotemporal network modeling of multi-hazard: From bipartite to single point network

HE Jinping(), LI Shuangshuang()   

  1. School of Geography and Tourism, Shaanxi Normal University, Xi'an 710119, China
  • Received:2020-09-18 Accepted:2020-11-20 Published:2021-08-10 Online:2021-10-10

摘要:

基于1970—2017年秦岭南北72个站点气象数据,以“地理时空分析-小波相干分析-时空耦合网络”为方法框架,对秦岭南北干旱-热浪时空耦合特征进行分析;进而以干旱-热浪时空耦合网络为基础,完善时空网络连边规则,拓展单顶点网分析方法,再认识多灾种时空耦合的群聚群发效应。结果表明:① 全球变暖背景下,秦岭南北降水模态逐渐由20世纪80年代雨涝主导向干旱主导转变,同时热浪在2010年前后经历第2个谷值期后快速增加,加剧了区域干旱-热浪耦合风险。② 秦岭南北干旱-热浪变化具有同步性,但是不同时段显著周期存在差异。其中,在20世纪70—80年代初,秦岭南北干旱-热浪4~8 a周期同步减弱,并向低频2~4 a周期转变;后期同步耦合增强时段有2个,分别是1995—2002和2012—2017年。在空间格局上,秦岭以北和汉江谷地为秦岭南北干旱-热浪耦合影响关键区域,而丹江口水库附近、嘉陵江流域、秦岭南坡中段为干旱-热浪耦合波动区。③ 在研究方法上,地理时空分析为秦岭南北干旱-热浪时空耦合提供基本事实判断,小波相干可定量干旱-热浪多时间尺度耦合关系,多灾种时空耦合网络可解释多灾种“平静-爆发”现象,识别干旱-热浪耦合稳定区和波动区,三种方法相辅相成,初步形成面向多灾种时空耦合分析方法体系。

关键词: 多灾种, 干旱, 热浪, 复杂网络, 秦岭南北

Abstract:

Modeling the processes of multi-hazard interaction could provide a pathway for a better understanding of the question how to identify and characterize the interaction of natural hazards and what is the possible cumulative or amplification effect where a hazard triggered or combined secondary hazards, which would help scientists to understand multi-hazard cause even damage to the social-ecological system. Based on the daily observations of 72 meteorological stations released by the National Meteorological Information Center of China, this study focuses on the spatiotemporal variation in concurrences of droughts and heat waves in the south and north of Qinling Mountains region for the period of 1960-2016 by using geographical analysis, wavelet conference, complex networks and other statistical techniques. Specifically, we finished the prospective transform from bipartite to single-point network and the modification in the spatiotemporal network rules for multi-hazards, which enriched the traditional research paradigm to investigate changes in concurrent of multi-hazards under the global warming. The results showed that: (1) The wet-dry pattern in the south and north of Qinling Mountains region has switched from the dominant flooding in the 1980s to continuous droughts. Meanwhile, the increasing shift of heat waves occurred after the depression or hiatus process during 1998-2010. Severe droughts and heat waves concurrences have become frequent and higher risk of multi-hazards have increased more than single hazard in recent years. (2) Investigation reveals there is synchronous variation between the concurrent droughts and heat waves, and the relationship shows different periodicity in different periods between 1970 and 2017. A statistically significant high-power region is evidently found from the 1970s to the 1980s at the shortest timescales, i.e. 4-8 years, with droughts and heat waves decreasing simultaneously, and then the weaker correlation is observed until the early 1990s. High-power regions are also observed during 1995-2002 and 2012-2017 from quasi-decadal to the interannual timescales, i.e. 2-4 years, which show a substantial increase in concurrent droughts and heatwaves across the study region. (3) Spatially, Guanzhong basin and Hanjiang river basin are two specific regions prone to concurrent of droughts and heat waves, while the cold spots of multi-hazard are observed in the Jialing river basin, the middle part of south piedmont of Qinling Mountains and the region near Danjiangkou reservoir. (4) Here, the multi-hazard approaches framework is provided to analyse the effects of concurrent and compound extremes. In particular, spatiotemporal analysis can be used to understand the background of multi-hazard and provide some basic evidences about the interaction of natural hazards. To quantify the presence of scale-dependency in the relationship of droughts and heat waves, wavelet coherence is used from a nonstationary perspective. Finally, the spatiotemporal network for multi-hazard could explain the ‘burst’ phenomena of multi-hazard, i.e. the spatiotemporal distribution of concurrent droughts and heat waves is best described by power lows, which are neither regular nor completely random.

Key words: multi-hazard, drought, heat wave, complex network, south and north of Qinling Mountains region