Assessing Urban Flood Risk in Lagos using IoT and GIS

Thesis Abstract

Urban flooding poses a persistent threat to Lagos, exacerbated by rapid urbanization, inadequate drainage, and climate variability, challenging resilience and infrastructure performance. This study aims to assess urban flood risk in Lagos through an integrated Internet of Things (IoT) sensor network and Geographic Information System (GIS) framework to inform risk-informed urban planning and disaster management. The specific objectives are to (i) map hydrometeorological and drainage-conditioning variables for selected districts, (ii) develop a real-time flood susceptibility index by fusing IoT-derived rainfall and water level data with land-use, elevation, and surface permeability, (iii) calibrate and validate the index against historical flood events using probabilistic approaches, (iv) evaluate community and infrastructural exposure across Lagos megacity zones, and (v) formulate data-driven recommendations for adaptive drainage design and early-warning operations. The study adopts a convergent mixed-methods design, combining quantitative sensor-based measurements with qualitative stakeholder insights to ensure triangulation and relevance for policy. The population comprises all flood-prone districts within Lagos Metropolitan Area, with a purposive sample of 12 districts selected for dense sensor deployment based on historical inundation records, population density, and drainage capacity. A total of 60 IoT nodes (including rain gauges and water level sensors) will be installed across five representative districts, complemented by archived rainfall, hydrological, and drainage data from Lagos State Water Corporation and Nigerian Meteorological Agency. Instruments include calibrated IoT sensors, high-resolution LiDAR-derived digital elevation models, satellite-derived land-use classifications, and structured interviews and focus group guides for urban planners, engineers, and community leaders. Validity and reliability will be ensured through sensor calibration, cross-validation with rainfall-runoff records, and test-retest procedures for interview protocols. Data analysis employs a multi-step methodology (i) preprocessing and quality control of IoT streams; (ii) development of a flood susceptibility index using a weighted logistic regression model integrating rainfall intensity, water depth, drainage capacity, elevation, and land-use factors, with model selection guided by Akaike Information Criterion (AIC) and Receiver Operating Characteristic (ROC) analysis; (iii) spatiotemporal analysis of flood risk using GIS-based hotspot detection, flood-inundation mapping, and time-series analysis of sensor data; (iv) validation against historical flood extents from Sentinel-2 imagery and documented flood events, employing confusion matrices and Cohen’s kappa; and (v) stakeholder perception analysis through thematic coding of interview transcripts guided by the theoretical lens of the Resilience Theory and the Capability Approach. Anticipated findings include a high-resolution flood risk map with district-level risk scores, identification of critical drainage bottlenecks, and quantified exposure of vulnerable populations and assets. The study expects IoT-informed models to outperform traditional GIS-only approaches in short-term flood forecasting and scenario assessment, with robust validation indicating practical applicability for municipal early-warning and infrastructure planning. The contribution to knowledge lies in (i) operationalizing an integrated IoT-GIS framework for urban flood risk assessment in a mega-city context, (ii) developing a replicable flood susceptibility index that blends surface hydrology, infrastructure capacity, and urban form, and (iii) providing empirically grounded recommendations for adaptive drainage design, land-use zoning, and community engagement to enhance urban resilience. The main conclusion envisaged is that real-time sensor data, when fused with high-resolution spatial data, substantially improves flood risk understanding and decision-making capacity in Lagos. Policy recommendations include prioritizing sensor expansion in high-risk corridors, implementing nature-based drainage solutions in identified hotspots, updating zoning regulations to reflect dynamic flood risk, and establishing an integrated, city-wide early-warning system linked to municipal emergency response protocols.