Investigating How the Public’s Understanding of Coastal Flooding Decision Support Systems Differs from the Experts: Results from an Eye-Tracking Study

Abstract

Disaster preparedness efforts often fail not because forecasts are inaccurate, but because warnings do not reach or are not understood by those at risk. Hurricane Katrina exemplifies how communication breakdowns can turn accurate forecasts into catastrophic losses, while the 2025 flash flooding in Texas demonstrates that monitoring systems alone are insufficient without timely, comprehensible information. This study evaluates the usability of Delaware’s Coastal Flood Monitoring System (CFMS) in a framed field economic experiment that integrates eye-tracking technology. We compare the comprehension and performance of experts and the public across eight incentivized tasks, analyzing both traditional usability metrics (accuracy) and process-level eye-tracking indicators (fixations, saccades, fixation duration). The public had a lower accuracy rate and required more fixations, longer fixation durations, and more saccades to complete tasks, whereas experts achieved higher accuracy and processed information more efficiently with significantly fewer fixations, shorter durations, and fewer saccades. Heatmap analysis revealed the largest usability gap in flood map tasks involving submerged roads (a potential life-threatening risk): the public’s visual attention was scattered, whereas experts’ fixations were concentrated, indicating clearer task comprehension. These findings show that some CFMS elements remain difficult for the public to interpret, highlighting the need for user-centered design alongside technical accuracy. Strengthening accessibility is essential if early warning systems are to achieve their life-saving potential as flooding risks intensify.