FloodCast: A user-centered flood monitoring system to enhance user engagement

Flooding is a major natural hazard whose frequency and severity have increased due to climate change. Although technological advances have improved flood monitoring systems, their usability remains a critical but often overlooked aspect. FloodCast is a sensor-based flood monitoring system that enables users to track real-time water levels in flood-prone areas and predict flood trends for next three hours. User requirements were gathered from individuals with firsthand experience of flooding. FloodCast's output aims to support decision-making, particularly in identifying safe routes during heavy rainfall. Data were collected through two focus group discussions with eight participants each and a survey of 76 respondents. Through iterative prototyping—informed by the user needs, an initial interface was designed. Initial testing that involved 17 participants revealed user confusion stemming from unclear visual cues and navigation, 35 % of which did not attain the benchmark score of System Usability Scale (SUS). After refining the design (e.g., simplifying data visualization, improving accessibility), the final prototype was tested with 21 participants where 100 % passed the SUS benchmark score. Users provided favorable evaluations of the redesigned interface, highlighting the clarity of the flood prediction graph, the usefulness of the reference images, and the value of the 15-min update feature. • Users rely on familiarity with locations when interpreting flood risk showing that context-specific data is more valuable than general alerts. • Clear visual design is essential so that users do not misinterpret colors, graphs, and trend indicators. • Color conventions in flood warnings require careful implementation due to variability in user interpretation. • Applying user-centered design (UCD) principles to flood monitoring app significantly enhance its effectiveness and adoption by aligning functionality with user needs, and behaviors.