A new perspective to understand public response to the Typhoon Doksuri from coastal and inland regions

Abstract

Massive amounts of data from social media possess the potential to rapidly identify the primary issues of concern in emergency disaster management. In summer 2023, Super Typhoon Doksuri which was an exceptionally special typhoon disaster that caused severe damage to China's coastal areas and disastrous impacts in inland regions, particularly triggered the most severe rainstorm in Beijing area in over a century. To enhance typhoon hazard reduction in both coastal and interior locations, it is crucial to examine public response to these events. This study uses microblog text data from July 27 to August 3 of 2023 to map the public response to Typhoon Doksuri. The Support Vector Machine (SVM) algorithm was used to classify the microblog text in combination with the typhoon path to analyze the spatial and temporal variations of the emotions of the affected individuals. The relationship between changes in public opinion, the distribution of topics, and the major disasters triggered by the residual circulation of Typhoon Doksuri in the Beijing-Tianjin-Hebei region is discussed. The Mentougou mega-storm in Beijing area that occurred in July 2023 is a typical case. The findings demonstrate that during the typhoon event, the focus of public attention changed with the movement of the typhoon path, and various public opinion topics exhibited temporal synchronization. Public sentiment indicates that the overall supportive sentiment is higher than is fearful sentiment. Based on this, it is crucial to strengthen the Beijing-Tianjin-Hebei cooperative emergency response, and response measures were proposed related to urban flood control and drainage construction, public awareness, backward areas, secondary disasters, resident relocation, and social media technology.

Keywords

• Typhoon disaster
• Data mining
• Disaster prevention and mitigation
• Public opinion