RAPID: When Floods Return After a Year: How Recent Experience and Layered Warnings Shape Protective Action

This project addresses an urgent national priority in disaster resilience by investigating how communities respond to rapid-onset flash floods in the Texas Hill Country. In July 2025, and again in July 2026, catastrophic floods struck the same river systems and populations in Texas, with profoundly different outcomes. Based on interviews and surveys of residents, tourists, camp operators, small businesses, and others from affected Texas counties, this research will focus on improving public safety, informing emergency management practices, and supporting access to life-saving information for all Americans. The project extends warning fatigue research to an environment where the redundancy provided by overlapping alerts received from individual devices, local government, and national government, is helpful versus when it may trigger “opting out” by the affected population. Findings from this research will directly benefit local governments, technology companies, and the agencies responsible for development and operation of alerting systems, by helping clarify when layered warnings enhance protective action versus risk creating warning fatigue that may lead to opting out by the targeted population. The research leverages the unprecedented, though unfortunate, opportunity presented by recurring flooding events in Texas to understand how factors, including prior disaster experience, evolving warning technologies, and incomplete recovery from a previous flash flooding event, shapes preparedness and protective action. The project will conduct comparative, longitudinal interviews and focus groups with affected residents, emergency managers, and camp operators, alongside a survey of 800 warning recipients across 30 counties in Texas. Hydrodynamic modeling will be used to reconstruct site-specific lead times and flood progression, enabling the separation of warning-system effects from physical flood conditions. The project will analyze how warning messages delivered via sirens, opt-in text alerts, device-generated imminent-threat alerts, and Wireless Emergency Alerts, are interpreted and normalized across repeat events. Grounded in the notion that warnings succeed only when recipients personalize risk before acting, the study will also advance understanding of the potential use of transient visitors in these circumstances and identify the small businesses that host them as an overlooked warning conduit. By integrating qualitative and quantitative data from sirens, sensors, mobile alerts, and person-to-person communication the study provides one of the first integrations of warning technology with human behavior across two major disaster events. The study will develop a framework for community learning and warning performance assessment, producing actionable evidence to help improve disaster warning infrastructure and communication. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.