Developing Community-Based Early Warning Systems for Flood Disaster Management Using Mobile Technology

Abstract

Flooding has become increasingly frequent and severe in Akwa Ibom State, largely driven by rising rainfall variability, rapid urbanization, and limited early warning infrastructure. This study assessed long-term rainfall trends, flood incidence patterns, community risk perception, institutional capacity, and determinants of household preparedness between 2005 and 2026. Secondary data were obtained from NEMA and Akwa Ibom State Emergency Management Agency, while primary data were collected through structured questionnaires administered across selected Local Government Areas. The Mann–Kendall trend test revealed a statistically significant upward trend in annual rainfall (τ = 0.74, p < 0.001), with Sen’s slope estimator indicating an average increase of 38 mm per year. Pearson correlation analysis showed a strong positive relationship (r = 0.87, p < 0.001) between annual rainfall and flood events. Logistic regression results identified access to early warning systems (OR = 4.14), mobile phone ownership (OR = 3.16), education level (OR = 2.09), prior flood experience (OR = 1.86), and income level (OR = 1.78) as significant predictors of household preparedness (Model accuracy = 76.3%; Nagelkerke R² = 0.48). Despite high levels of flood experience (87.4%) and growing risk perception (81.6%), more than half of respondents (54.3%) exhibited low preparedness, while institutional assessments revealed major gaps in real-time monitoring, last-mile communication, and mobile-based alert systems. The findings underscore the urgent need for integrated, technology-driven, and community-centered early warning systems to enhance resilience and reduce flood-related losses in the state.