Information Dissemination, Diffusion, and Response during Hurricane Harvey: Analysis of Evolving Forecast and Warning Imagery Posted Online

This article aims to build interdisciplinary understanding about modern hazard communication by investigating visual information dissemination, diffusion, and response leading up to and during a weather-related disaster. The study analyzes data from online social media posts by authoritative sources during Hurricane Harvey, focusing on forecast and warning tweets containing hurricane risk imagery. The research integrates quantitative and qualitative analysis of tweets, retweets, and replies to explore how the roles of different information content and sources evolved with Harvey's threat. Building on the work of Mileti and other warning scholars, the results illustrate the complexity of dynamic multisource, multimessage, and multihazard forecast and warning situations, including hurricanes. In such situations, people can engage in different hazard response processes simultaneously and near-continuously, as they are exposed to, attend to, and make sense of an evolving, heterogeneous collection of available information. The analysis also finds that during Harvey, authoritative sources used a mix of tweeting strategies to disseminate and amplify emerging information. Different types of sources led the creation of forecast and warning content at different times, with other sources playing complementary roles in communicating potentially important or salient content to broader audiences. Overall, the study provides updated models of hazard warning communication and response and associated processes such as milling, along with new methodological approaches for utilizing social media and other online data to understand these processes. In addition to these theoretical and methodological contributions, the analysis points to opportunities for the National Weather Service and others to improve tropical cyclone risk communication.