Where Flood Risk Outruns Insurance: Evidence from Two Decades of Expanding Combined Risk Hotspots across the Contiguous United States

This repository contains the complete dataset and analytical codebase supporting the study of Combined Risk Hotspot (CRH) zones, urban land use change, flood exposure, economic damage, and social vulnerability across the Contiguous United States (CONUS), with focused analysis on five major metropolitan areas: Baton Rouge (LA), Houston (TX),Jacksonville (FL), Orlando (FL), and Tampa (FL). The study spans from 2000 to 2023, enabling multi-decadal assessment of combined effect of urban expansion, social vulnerability and flood risk. This study also detect the total economic damage with underinsurance patterns within and outside designated CRH boundaries at both national and metro scales. Combined Risk Hotspots (CRH) are defined as areas experiencing simultaneous convergence of high flood exposure, rapid urban development, and elevated social vulnerability — representing zones of compounded and disproportionate disaster risk across CONUS. The dataset integrates multiple geospatial and tabular data sources including Land Use/Land Cover (LULC) raster data derived from national coverage, FEMA flood exposure layers, U.S. Census demographic records, Social Vulnerability Index (SoVI) data, administrative boundary shapefiles, and NFIP economic loss records — all referenced to the CONUS extent. These are complemented by Python-based Jupyter Notebook scripts for spatial analysis, statistical modeling, risk assessment, and data visualization. The repository is organized into the following thematic categories: 1. Combined Risk Hotspot (CRH) Zone Data: Spatial boundaries and area statistics for Combined Risk Hotspot zones across CONUS, aggregated at both the state level and across the five selected metro areas, including comparative CRH rankings to identify metros with the highest concentration of compounded risk. 2. Boundary Shapefiles: Administrative and metropolitan boundary shapefiles for each study city, state boundaries across CONUS, and broader metro boundary extents used as spatial reference layers throughout the analysis. 3. Land Use / Land Cover (LULC): National-scale raster data (.tif) for years 2000 and 2023 capturing urban and non-urban land cover classifications across CONUS. Includes derived layers showing developed land change and net development transitions, used to assess urban encroachment into Combined Risk Hotspot zones. 4. LULC Transition Data (Sankey): Aggregated CSV files per metro area summarizing land use class transitions between 2000 and 2023, used to generate Sankey diagram visualizations illustrating urban growth patterns within and around CRH boundaries. 5. Census Data: U.S. Census demographic datasets from 2000 and 2023 capturing population characteristics, housing, and socioeconomic indicators used to assess community exposure within Combined Risk Hotspot zones across CONUS. 6. Social Vulnerability Index (SoVI): Spatial SoVI datasets for 2000 and 2023 quantifying community-level vulnerability to environmental hazards, used to identify socially vulnerable populations residing within Combined Risk Hotspot areas and track vulnerability shifts over time. 7. FEMA Flood Exposure: National geospatial flood exposure layers including high flood risk zones, serving as one of the core input layers for delineating Combined Risk Hotspot boundaries and assessing population exposure across CONUS. 8. Economic Damage: Tabular records of flood-related economic losses from 2000 to 2023 across CONUS from NFIP dataset, disaggregated by CRH and Non-CRH zones, capturing trends in insured and uninsured flood losses to quantify the economic consequences of residing within CRH areas. 9. Metro Area & Population Exposure: Summary statistics on metro-level CRH area coverage and population exposure estimates for the five study metros, providing a comparative picture of risk concentrat within the broader CONUS context. 10. Bivariate & Cluster/Hotspot Analysis: Derived spatial analysis outputs including bivariate mapping, cluster analysis, and hotspot analysis results identifying areas of compounded flood risk and social vulnerability, used to validate and contextualize Combined Risk Hotspot delineations across CONUS. 11. Analytical Code: Jupyter Notebook scripts covering Economic Damage analysis, Relative Risk Ratio (RRR) analysis, Sankey diagram generation, and spatial distribution of underinsurance patterns across the CONUS. All spatial data are referenced to standard CONUS coordinate systems and are compatible with QGIS and ArcGIS. Raster files are provided in GeoTIFF format and vector files in ESRI Shapefile format. Tabular data are provided in CSV and XLSX formats. All code scripts are written in Python and are executable via Jupyter Notebook.