Examining Community Vulnerabilities through multi-scale geospatial analysis of social media activity during Hurricane Irma

Abstract

Natural disasters may cause a substantial but spatially uneven increase in online activity among users in disaster zones seeking food, shelter, and medical assistance through social media. However, the presence of a digital divide affects social media participation among users. Therefore, the social and spatiotemporal inequality in the usage of social media data should be fully considered before such data can be leveraged to predict damage, investigate impacted populations, and prioritize activities during disaster management. Using Hurricane Irma as the case study, this research integrates Twitter data from September 2017 with other public data sets to examine spatial variation in social media activity during disasters. To investigate the social and spatiotemporal inequality in the usage of social media and examine the association between social media activity and flood rate at census tracts level, Multiscale geographically weighted regression was conducted. Our results show counties should not be seen homogeneous during disasters, specifically when using social media as a platform to disseminate information or direct help to affected people. It is clear from this work that in the Tampa Bay area, coastal, affluent communities have considerable representation on social media while poorer, inland areas are absent. It is important to note the limitations of online social media and microblogs in terms of access. Social media use varies across demographics and geographic location. Disaster management plans must consider the specific demographic, physical, and socioeconomic characteristics of each area before taking advantage of social media.