关键词: 暴雨灾害, 灾害网络演化, 协同救援效果, 社会网络分析

Abstract: In recent years, rainstorm disasters in China have exhibited increasing trends of suddenness and recurrence, posing significant threats to economic development, the safety of human lives, and property. Furthermore, extreme rainstorm events often exhibit spatiotemporal dynamic coupling relationships with social and economic activities, leading to cascading effects that are challenging for single disaster response organizations to manage effectively. To recognize the systematic requirements for addressing rainstorm disasters, it is essential to establish an integrated response model encompassing “multiple disaster events, multiple stages, multiple emergency tasks, and multiple response entities,” as well as to develop a rational emergency task allocation mechanism to enhance the adaptability of collaborative networks within emergency organizations and improve the precision of disaster prevention and relief measures.
This article focuses on the 2023 Shulan rainstorm as a case study. Data are collected from news reports disclosed by mainstream media, laws, regulations, and emergency plans published on government websites at various levels. Based on these materials, the study first identifies representative disaster events frequently triggered by typhoons or rainstorms and constructs a rainstorm disaster network topology diagram using the interrelationships among these events. The overall network attributes, node characteristics, and edge vulnerabilities are analyzed in depth to summarize the primary features of this rainstorm disaster. Subsequently, the content of emergency tasks is extracted. If an emergency task co-occurs with a specific disaster event in the same source material, a connection is established between them, and a “disaster event - emergency task” two mode network is constructed. A “core-periphery” structural analysis is then performed to identify key disaster events and emergency task contents. Next, “emergency task - response entity” two mode networks are constructed for statistical units T1 through T4. Connections are established if a response entity executes a specific emergency task during the corresponding period. Finally, the collaborative rescue effectiveness of this rainstorm disaster is analyzed from the perspectives of disaster events, emergency tasks, and response entities.
The main findings of this paper are as follows: (1)The control windows for different disaster events vary. For instance, typhoons and rainstorms are primarily managed during the warning response, personnel search, rescue, clearance, and resettlement stages, whereas crop damage is addressed during the clearance, resettlement, recovery, and reconstruction stages. During these stages, the number of disaster events, the chains controlled by each response entity, the reduction in vulnerability of disaster chain edges, and the clustering coefficient of the disaster network can serve as indicators to evaluate their collaborative performance. (2)Across the stages of rainstorm disaster response, there are both similarities and differences in the prioritization of emergency tasks. It is crucial to ensure the stable execution of cross-stage emergency tasks, such as consultation, judgment, command, and coordination. (3)In these stages, response entities with larger effective scales and greater opportunities for “structural holes” are predominantly government organizations that align with the key emergency tasks of the stage. However, to enhance societal resilience and cohesion, it is necessary to further leverage the flexibility advantages of enterprise and social organizations and address issues related to insufficient participation. Additionally, the structural characteristics of the collaborative network exhibit temporal variations: during the warning response stage, the nature of disaster response entities is relatively homogeneous; during the personnel search and rescue stage, information barriers exist among multiple entities; during the clearance and resettlement stage, entities gradually adapt to complex environments; and during the recovery and reconstruction stage, the focus of emergency tasks requiring collaborative execution becomes more prominent.
Based on the findings of this study, the criteria for evaluating the collaborative rescue effectiveness of multiple entities demonstrate diverse and intricate features. However, relying solely on a single case study to assess this effectiveness has certain limitations. Future research could consider incorporating additional samples of extreme disaster emergency management and conducting more in-depth analyses from the perspective of the dynamic evolution of disaster networks to further validate the generalizability of the conclusions presented in this paper.

Key words: rainstorm, evolution of disaster networks, cooperative rescue effect, social network analysis