Alt text: U.S. county-level map showing a bivariate color scale comparing net building payment amounts (y-dimension) and total building insurance coverage (x-dimension). Darker purples indicate counties with both high coverage and high payments. Lighter colors show lower values or missing data. Insets include Alaska, Hawaii, Puerto Rico, and U.S. territories.
A dynamic map showing the Top 10 Costliest Flood Events in U.S. history based on total NFIP claim payments, adjusted to 2025 dollars.
Each frame highlights one event, mapping both:
- Total claim payments, and
- Insurance coverage,
at the county level using a bivariate choropleth.
Alt text: Animated U.S. county maps showing the geographical distribution of NFIP claim payments and insurance coverage for each of the top 10 flood events by adjusted payouts.
Alt text: Geographical distribution of NFIP flood insurance claim payments and total building insurance coverage across U.S. counties for the 10 costliest flood events (adjusted to 2025 dollars). This map highlights spatial patterns of financial risk and protection during major flood disasters.
- NFIP Redacted Claims Dataset:
https://www.fema.gov/openfema-data-page/fima-nfip-redacted-claims-v2 - NFIP Policies Dataset:
https://www.fema.gov/openfema-data-page/fima-nfip-redacted-policies-v2 - U.S. County Geometries:
https://www2.census.gov/geo/tiger/TIGER2024/COUNTY/
📁 figures/
└── coverage_vs_payments_figure.png # Final bivariate choropleth map
└── top_10_flood_events_animated.mp4 # Top 10 costliest flood events animated
└── top_10_flood_events_animated.gif # Top 10 costliest flood events animated
└── top_10_flood_events_animated.png, event_*.png # Top 10 costliest flood events static images
📁 scripts/
└── download_from_openFEMA.py # Downloads NFIP datasets from FEMA
└── plot_bivariate_choropleth.py # Processes data and creates the map
└── static_plots_bivariate_choropleth_top_10_flood_events.py # Generate Animation of top 10 costliest flood events
This work was inspired by @Marc's deep dive into NFIP datasets that got me thinking about how coverage and claims vary spatially.