Filters parcels & buildings (z10+); county/hex aggregates are unaffected.
Zoom in to drill down: counties → H3 hexes → parcels → individual buildings. Aggregates shade by the selected peril's elevated share, so they move with the toggle. Click a county to focus the panel on it.
Read before quoting.
· yearbuilt is construction year, not roof age.
· Hail is radar-detected density, not modelled severity.
· Tornado / wind / winter are tract-resolution, not per-building.
· Hazard is centroid-sampled for 98.24% of buildings.
· Assessor $ values are popup-only reference (≈63% coverage, county-dependent); never used for styling or ranking.
· NULL road/station distance = furthest (no station within 25 km), not missing —
on a building, and on a parcel that has buildings. A parcel with no buildings
reads “no building to measure”: nothing was measured there, which is not the
same claim as remoteness.
· Access drive time is a modelled estimate in 5-minute bands — straight-line
distance × a terrain factor calibrated against HERE routes, not a per-building route.
· Aerial imagery is locational context only, including at partial opacity —
blending it under the footprints changes what you can SEE, never what was measured. Scores are screening indicators from
modelled and tract-resolution inputs — never derived from imagery. A building shown
under aerial is not individually assessed.
Statewide exposure
Counties by exposure
Assessor $ coverage — why not dollars
Improvement-value coverage is ~63% and county-dependent. The largest exposures often publish the least — El Paso (the state's biggest, 290,630 buildings) reports $0. A dollar ramp maps who reports, not where risk is.