Authors: A. Marcela Suarez*, University of California - Santa Barbara
Topics: Spatial Analysis & Modeling, Hazards, Risks, and Disasters
Keywords: Twitter, disaster, emergency management, Hurricane Harvey, reliability
Session Type: Paper
Start / End Time: 9:55 AM / 11:35 AM
Room: Cleveland 1, Marriott, Mezzanine Level
Presentation File: No File Uploaded
When a weather-related disaster occurs, social media platforms such as Twitter become a useful tool for the general public to report real-time information. Content shared on Twitter usually contains timely, updated, and sometimes on-site reports. Because of this, and the speed and ease with which information is shared through this micro-blogging platform, forecasters and emergency managers have tried to leverage tweets as an on-the-ground reference for both confirming the occurrence of weather-related disaster events and tracking their local impacts and damage. However, since Twitter data is massive, semantically varied, noisy and spatio-temporally uncertain, little of the information it might provide is actually useful, and when it is, it usually requires exhaustive manual filtering to identify reliable posts. This study explores how citizen reports validated by the National Weather Service and authoritative geographic information provided by FEMA can be used to assess the reliability of tweets about weather-related disaster events. By using Hurricane Harvey as a case study, we propose a framework to assess the reliability of tweets that consists in analyzing their spatial distribution and estimating both the proximity to validated reports and the spatial correspondence with areas at risk. Results are validated using the hurricane Harvey damage map produced by FEMA.