Roughly a quarter of US roads and critical infrastructure face operational flood risk; askamerica's own bridge data shows structural scour risk edging down even as flood-map/inundation risk is understood to be growing

Summary

Roughly a quarter of US critical infrastructure and road mileage already sits in areas with meaningful flood risk. The most complete national estimate—First Street Foundation's physically-based flood model, the only US-wide study that scores operational flood risk for roads and critical facilities rather than just insurance-zone membership—found that 23% of US road segments (nearly 2 million miles) and 25% of critical infrastructure facilities (~36,000: utilities, airports, ports, hospitals, police/fire) face flood risk severe enough to render them impassable or inoperable today. Over the next 30 years, climate change is projected to add roughly 63,000 more miles of at-risk road and 2,000 more at-risk critical-infrastructure facilities.

askamerica's own data adds one directly-measured, bridge-specific trend that complicates a simple 'exposure is growing' story: FHWA's own scour-critical rating for the 516,338 US bridges that cross a waterway (82.6% of all 624,857 NBI bridges) has edged down, from 4.20% scour-critical in 2017 to 3.84% in 2025—meaning targeted federal/state remediation has been outpacing new structural deterioration. That is a narrower, engineering measure of foundation vulnerability to scour, not the broader overland/inundation flood risk First Street measures, and the two should not be conflated.

On the map-driven growth question specifically—how much exposure has grown as FEMA flood maps get updated—askamerica has no ingested spatial flood-hazard layer to compute this directly (see Data Coverage below), and no single national dataset publishes a clean 'newly-mapped-into-SFHA' infrastructure count. The best available evidence is qualitative but consistent: FEMA's own maps are widely reported as outdated and not climate-adjusted (FEMA's own director has said so publicly), and First Street's 2020 comparison found roughly 1.7x more properties nationally face substantial flood risk than FEMA's Special Flood Hazard Area designation captures—implying that as maps and models are updated toward reality, mapped exposure tends to rise, not fall.

What 'meaningful flood risk' means here

First Street Foundation's 2021 report, "The 3rd National Risk Assessment: Infrastructure on the Brink", is the only nationwide study found that scores flood risk for roads and critical infrastructure by an operational threshold—the flood depth that would render a road impassable to cars or a hospital inoperable—rather than by FEMA's binary Special Flood Hazard Area (SFHA) boundary, which is known to undercount real risk (see below). Headline county/city-level findings: highest concentration of infrastructure risk in Louisiana, Florida, Kentucky, and West Virginia, with 17 of the top 20 most at-risk counties (85%) in those four states; Cameron Parish, LA ranked most at-risk county nationally.

askamerica direct measurement: bridge scour-critical trend, 2017-2025

askamerica's transport.bridges table (FHWA National Bridge Inventory) carries Item 113, the scour-vulnerability rating FHWA assigns to every bridge crossing a waterway—the closest thing in this corpus to a bridge-specific flood-risk indicator, and the only flood-adjacent transportation-infrastructure trend computable end-to-end from askamerica's own tables.

SQL used: SELECT "year", scour_critical, COUNT(*) FROM transport.bridges WHERE record_type='1' GROUP BY "year", scour_critical, with codes 0-3 (unstable/failure-imminent through scour-critical-monitor) summed as the numerator and rows excluding code 'N' (not over a waterway) as the over-water denominator, per FHWA's own Item 113 coding guide. The declining share reflects the same story ASCE's 2025 Infrastructure Report Card tells nationally—bridges graded C, with 6.8% in poor condition—of remediation funding (including IIJA/BIL dollars) reducing the worst structural flood-vulnerability faster than new bridges are aging into it. This does NOT mean overall flood risk to bridges is falling: scour-critical measures foundation stability under a calculated flood, not the broader risk of a bridge or its approach roads being submerged/closed by rainfall or storm surge, which is the risk First Street's model and NOAA precipitation trends both show increasing.

NFIP claims: volatile, event-driven, no clean trend

Paid NFIP flood claims (askamerica disasters.nfip_claims, aggregated by county-year) do not show a clean secular growth trend at this grain—they spike sharply in major hurricane/storm years (2005 Katrina ~277,000 claims nationally aggregated, 2012 Sandy ~176,000, 2017 Harvey ~146,000, 2024 Helene/Milton ~204,000) and fall to a few tens of thousands in quieter years. This dataset measures realized insured losses, not infrastructure exposure, and is included here as supporting context for the disaster-severity backdrop against which transportation flood risk sits, not as a direct infrastructure-exposure measure.

Data coverage gap: no ingested FEMA flood-hazard/SFHA spatial layer

askamerica's disasters schema ingests FEMA's National Risk Index for earthquake hazard only (disasters.fema_nri_earthquake_by_county); the source NRI layer's 17 other hazard blocks, including flood, are confirmed not yet ingested (a sourcing gap, not a coverage-window issue). Separately, no FEMA Special Flood Hazard Area (SFHA) polygon/zone layer is ingested at all—disasters.nfip_claims and disasters.nfip_policies carry a flood_zone column, but only for locations with an actual NFIP claim or policy, not a full spatial hazard map, and neither joins to bridge/road geometry. This means the specific ask—'how much of X infrastructure sits inside a mapped flood zone, and how has that grown release-over-release as FEMA remaps'—cannot be computed directly from this corpus. The First Street Foundation figures above are the best available substitute: an independent, physically-based national flood model that (unlike FEMA's SFHA) is explicitly built to capture risk FEMA's own maps miss.

What This Report Does Not Answer

• How much that exposure has grown as flood maps get updated specifically (map-vintage-driven growth): No FEMA SFHA spatial layer or flood-map-revision-history dataset is ingested in askamerica, and no public dataset was found publishing a clean national 'infrastructure newly mapped into flood risk after a remap' figure. Reported instead: (1) askamerica's own directly-measured bridge scour-critical trend 2017-2025 (a related but narrower structural-vulnerability measure, which declined), (2) qualitative evidence that FEMA maps undercount current risk and First Street's model, updated for climate, finds materially higher exposure than FEMA's SFHA -- both cited as the best available substitute for a map-vintage comparison.

Every query behind this report

3 warehouse calls ran in this session, in order. Each is reproducible against the same snapshot.

SELECT "year", scour_critical, COUNT(*) AS n
FROM transport.bridges
WHERE record_type = '1'
GROUP BY "year", scour_critical
ORDER BY "year", scour_critical

SELECT "year", COUNT(*) AS claims, SUM(total_paid) AS total_paid
FROM disasters.disasters_nfip_claims_by_county
GROUP BY "year"
ORDER BY "year"

SELECT "year", COUNT(DISTINCT county_fips) AS counties_with_claims, SUM(claim_count) AS claims, SUM(total_paid) AS total_paid
FROM disasters.disasters_nfip_claims_by_county
WHERE "year" BETWEEN '2000' AND '2025'
GROUP BY "year"
ORDER BY "year"