1. Social media and early warning systems for natural disasters: A case study of Typhoon Etau in Japan.
- Author
-
Kitazawa, Katsushige and Hale, Scott A.
- Abstract
Due to the significant improvement of disaster-related information, further reduction of disaster risks re-quires not only governments to provide more scientifically accurate information but also the public to take appropriate action at the correct times. Meanwhile, there is ongoing work to integrate social media into Early Warning Systems (EWSs), but the ecology of social media information during crises remains poorly understood. This study seeks to understand public responses on social media to EWSs using the case of a 2015 typhoon (the Kanto-Tohoku Heavy Rain) in Japan. Using a corpus of 35 million tweets, computational methods such as topic modelling, and geospatial analysis we find that: 1) emergency warnings are likely to have people be more attentive to the warnings but this does not translate to an increased discussion of actions such as evacuation; 2) the expected shift of public attention (from awareness to preparation and then action) seems to happen on social media. Overall, we show that analysis of social media data can compliment tra-ditional survey-based approaches to understand how the public respond to information from Early Warning Systems. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF