Speaker
Description
The frequent occurrence of extreme weather events has brought serious impacts on cities under the background of climate change. For example, the increasingly frequent typhoon will cause strong wind disaster and extreme rainstorm disaster with the characteristics of rapid change and high uncertainty. These disasters cannot be responded quickly with the traditional methods of disaster prevention and mitigation planning such as improving defense standards and formulating disasters prevention plans. Therefore, it is urgent to explore more adaptive planning methods to meet climate change challenges and realize climate resilience. In recent years, urban residents publish real-time disaster information through social media, which can be identified and analyzed to provide timely information to support decision-making for disaster response.
Based on this situation, this paper selects Shanghai as the study area, takes Typhoon "Pulasan" and Typhoon "Bebinca" as an example (these typhoons consecutively influence Shanghai from September 14 2024 to September 22, 2024), and conducts disaster perception analysis based on the disaster information published on Sina microblog (one of the most famous social media in China). Firstly, this paper collects the text data of Sina microblog during the disaster, identifies the disaster information from the text, and analyzes disaster theme of public concern with the topic model of the Latent Dirichlet Allocation (LDA). Secondly, this paper calculates the public perception level, and evaluates the public emotion tendency and the public emotion intensity during the disaster based on the sentiment dictionary of China National Knowledge Infrastructure (HowNet). Thirdly, it analyzes the spatiotemporal distribution characteristics of the public perception level and the public emotion intensity.
The results are as follow: (1) 2734 pieces of disaster information are identified, and the disaster themes are mainly divided into three categories, including disaster disruption to daily activities, disaster-affected locations, and post-disaster public space; (2) In terms of time dimension, the distribution of the public perception and the emotional intensity presents obvious double peak characteristics. In detail, the public perception during Typhoon "Bebinca" is higher than that during Typhoon "Pulasan", while the negative emotional intensity during Typhoon "Bebinca" is lower than that during Typhoon "Pulasan". Meanwhile, the peak of the emotional intensity lags behind the peak of the public perception; (3) In terms of space dimension, the distribution of the public perception (Moran I=0.053, z=7.769, p=0) and the negative emotional intensity (Moran I=0.045, z=3.842, p=0) present obvious clustering characteristics, while the distribution of the positive emotional intensity (Moran I=0.016, z=1.301, p=0.193) presents random distribution characteristics. Furthermore, the public perception hotspots are distributed in the southern coastal area of Shanghai while the negative emotion hotspots are mainly concentrated within the middle ring road of Shanghai, which shows the spatial distribution differentiation between the public perception hotspots and the negative emotion hotspots.
To summarize, the public pays more attention to the disaster impact on public facilities and public spaces. Additionally, there is different disaster impact on urban residents in different areas, resulting in spatiotemporal differences between the public perception hotspots and emotion hotspots. Accordingly, corresponding planning strategies are proposed from three perspectives, including integrating disaster perception results into spatial risk assessment, optimizing emergency response based on the public perception analysis, and taking the distribution characteristics of the public emotion intensity as decision-making support for post-disaster recovery.
The significance of this paper is to explore the application of big data technology and social media in responding to climate change challenges, so as to provide a reference for the integration of new technologies into resilient urban planning.
| Keywords | climate change; resilient city; social media; spatiotemporal perception; typhoon |
|---|---|
| Best Congress Paper Award | Yes |
