HomeRiskIQ three-city storm-surge land study
THREE-CITY RESEARCH PILOT — Published 2026-10-13
Analysis: 2026-10-11

Modeled Category 3 high-tide, near-worst-case MOM; not a current event

NoData does not distinguish dry land from masking or unsupported areas. It does not mean dry or safe.
Mapped class-area fractions are not probabilities, forecasts, evacuation zones, parcel conclusions or HomeRiskIQ scores.
A MOM combines maxima from many hypothetical storms; no single storm produces its entire footprint.
Depth-class bins are feet above ground, not continuous depths or water elevations.
The roughly 10 m output is finer than the underlying hydrodynamic modeling; coastlines and small features have uncertainty.
This source does not measure future sea-level rise, erosion, subsidence, future construction, large waves or all inland/freshwater flooding. Local defenses may be incomplete.
Census municipal boundaries and land/water definitions determine the denominator. Results should not be read as neighborhood or household exposure.

Method
Full 2026 Census municipal land geometry, including land with NoData.
Same-vintage TIGER/Line county FACES, selected PLACEFP, LWFLAG=L. These cities contain only L/P faces; L/P ATOTAL sums equal PLACE ALAND/AWATER.
Native-class raster row runs intersected with each original Census land face. Geographic edges densified to at most 0.00005 degrees, transformed to EPSG:5070, areas summed in square metres.
All positive-area intersections retained. No area threshold removes canals or sliver cells.
Class polygons clipped to identical land mask; topology-preserving 1 m display simplification only. Maps use local NAD83 UTM grid north. White is outside the municipality; blue is Census water. No basemap requests.
UI shares rounded to 0.1 percentage point, areas to 0.01 km². Rounded shares may not add to 100%. Exports retain calculation precision, not a claim of model accuracy.
Two independently implemented decoders and two intersection methods checked. Convergence halving densification step changed class areas by <0.00069 m². This is computational validation, not external scientific peer review or NOAA endorsement.

Source release: NOAA v4 June 2026 reprocessing, publicly updated June 18; archive folder June 17; internal processing metadata June 14.
Census TIGER/Line 2026 PLACE and county FACES, boundaries generally January 1, posted October 6. Destin land changed −1.5992% from 2025, so vintage differences are not risk trends.

Files
study.json: full class areas, grouped areas, thresholds, validation, sensitivity, source URLs and SHA256 hashes.
city-summary.csv: three city comparisons, each with scenario and NoData caveat.
depth-classes.csv: every class (including zero-area classes, 99 and 255) for each city.
City SVG/PNG maps: identical native class mask and source dates; grid north up.

No percentages measure current flooding, likelihood or the share of households affected. Existing HomeRiskIQ scores are separate. Retain scenario, dates, denominator and NoData caveat with any permitted reuse.