Emerging Geospatial Big Data and Disaster Resilience


Date: 2/28/2019

Author: Cassie Oswood

         Dr. Michelle Meyer and Dr. Lei Zou recently presenting their research on big data and how social media can be applied towards disaster resilience.
         Dr. Michelle Meyer is an Assistant Professor at Texas A&M University with expertise in sociology of disasters and environmental sociology. She is also the IfSC’s Community Resilience Discovery Lead as well as the Executive Director of the Hazard Reduction and Recovery Center. Her research partner, Dr. Lei Zou is a recently appointed Assistant Professor at Texas A&M University with research focusing on community resilience and social media data mining.
         Dr. Meyer and Dr. Zu have teamed up to focus their research on the application of big data for natural disaster resilience. Natural hazards and the damage that results from them are increasing because of climate change. As reported in the presentation, there are 50,000 deaths, 169,000,000 people affected, and $71,800,000,000 lost per year from natural hazards. To compensate for the increasing hazards, the monitoring and response systems have had to advance with the changing technology and society. The biggest challenge in monitoring natural hazards (before, during, and after the event) is a lack of data. The research conducted by Dr. Zou and Dr. Meyer presents social media as the solution to this insufficient data availability.
         In 2008, 24% of people in the United States were regularly using social media platforms. As of 2018, that number has risen to 81%. The advantage of using social media for data mining is that most platforms provide location monitoring and human behavior in real time. Previous traditional geospatial data collection methods involved remote sensing, field work, and surveys. The emerging geospatial big data is now made available by social media such as Instagram, Twitter, and Facebook. Transportation data is also provided by ridesharing apps like Uber and Lyft.
           But how can all of that raw data be analyzed?
          The research in this project is aimed at producing logarithms that import, archive, analyze, and store that data permanently in a big database. Data from social media, GIS/RS, census, and surveys will be inputted into a hybrid analysis system to produce platforms and products, applications, and algorithms and framework for use.
          The importance of this data analysis is its application for natural hazard and disaster resilience. For example, it was found that communities that suffer more damage from disasters exhibit more discussion for longer periods of time. By investigating different twitter use, it can be determined why there is higher or lower media coverage during different stages of disasters.
By producing logarithms that automatically extract useful information, social media can be used to monitor safety and rescue necessity and location, post-disaster damage estimations, and situational awareness. Tweets can be fed into a logarithm that deciphers the need for help of an individual, their location, and the number of victims. That information can then be fed directly to emergency services to respond for rescue.
          A survey of residents living in Houston during Hurricane Harvey reported that 86% of the 1,0002 respondents use social media. The most used platform for receiving information regarding the storm was the television (58.3%) and the second most used was social media (10.7%). Of the social media platforms used by these respondents, the most popular sites were Facebook (85% use), YouTube (58% use), Instagram (52% use), Snapchat (43% use), and Twitter (41% use). These platforms were used most commonly to find information regarding the location of the storm, the predicted flooding, the predicted landfall location, and the Flood Rescue number.
          Dr. Meyer reports that technology being used to assist in disaster rescue is “being used now, because we have the technology now.” This was seen when volunteer groups such as the ‘Cajun Navy’ formed in response to Hurricane Harvey. As the research suggests, these groups started following the flood of Baton Rouge. However, they came to prominence during Hurricane Harvey. These groups have 3 levels to their rescue strategy:

  1. The boaters in the water performing the rescues
  2. The dispatchers usually in distant location scanning social media sites to find people in need of rescue
  3. The coordinators behind the scenes that transfer the information from the dispatchers to the boasters and keep track of who has been rescued and where the people in need of rescue are.
         From the research of social media use in disasters and how it can be used by volunteer rescue groups, it is clear that technology is advancing in response to the increasing frequency and magnitude of natural disasters. The research being conducted by Dr. Meyer and Dr. Zou is not only riveting but a necessity for a sustainable and resilient future. As Dr. Philip Berke said,  “It was one of the best presentations heard on big data use.”


Michelle-and-Lei-Pic-2.jpg