StormProof unlimited NWS storm verification · for pros

StormProofhail seasonsDenver → 2004

Denver hail season 2004

48 NWS-recorded hail reports ≥1″ within 35 miles, across 9 storm days, max 2.00". Every one of those reports is a dated, located, citable official record — the context behind this market's 2004 claim volume.

Biggest storm days (2004, final record)

Date≥1″ reportsMax hailAreas named in the record
August 10, 2004262.00"WELD, BOULDER, LARIMER, ARAPAHOE
June 9, 200491.75"ADAMS, WELD, DENVER, JEFFERSON
August 30, 200431.50"ARAPAHOE
July 22, 200421.25"ADAMS
June 8, 200421.50"JEFFERSON

“Large hail broke car windows, damaged roofs and stripped trees and at least 3,000 acres of crops. At Eaton's high school, middle school and two elementary schools, at least 60 skylights were shattered by hail. Water flooded the hallways and classrooms. Damage to the schools alone was estimated at $2 million.”

— NWS event narrative, August 10, 2004 (NCEI Storm Events)

When it fell

May 5 · Jun 11 · Jul 3 · Aug 29

Wind context: the record also holds 10 thunderstorm-wind events ≥50 kt (≈58 mph, the NWS severe criterion) in this radius for 2004 — relevant where the dispute is wind vs hail causation.

Working a Denver claim from 2004?

These are aggregates. A claim file needs the per-address record: every recorded event within 1, 3 and 10 miles of the property, distances, official narratives, and citations an adjuster can check line by line. That's the report — generated in seconds, hosted on HailEvidence (the neutral evidence surface), formatted as an insurance-appeal attachment.

Unlimited reports — Pro $99/mo Single report $29

Provenance

Final counts: NCEI Storm Events Database, file vintage c20260527, hail events with recorded magnitude ≥1.00″ and point coordinates within 35 miles of the Denver anchor. NWS records are point and path observations. The absence of a nearby report does NOT prove that no hail fell at this address — it means no observation was logged nearby. A report of nearby hail documents the event; it does not by itself prove damage to a specific structure. Spotted an error? Email the address on our terms page and we correct against the source.