Colorado's August 24–28 Hail Outbreak: 2.67-Inch Hail Over the South Metro

I've been roofing in Colorado for about three years, and the last week of August is usually when homeowners have already mentally filed the season away. This year it did not cooperate. From Monday August 24 through Friday August 28, NOAA's MRMS radar network estimated hail somewhere in Colorado every single day — and the middle of that stretch was the widest radar footprint I've seen in a Colorado week this season.

Here's what the data actually shows, what it doesn't, and what I'd tell a homeowner on the Front Range to do about it.

What the Radar Recorded, Day by Day

MRMS (Multi-Radar Multi-Sensor) is a NOAA system that stitches several weather radars together and flags grid cells — each roughly one square kilometer — where it estimates hail is falling. It is a model estimate, not a physical measurement, but it is the most consistent address-level record we have of which neighborhoods were under a storm.

August 26 and 27 are the story. Going from under 1,700 flagged cells on Tuesday to more than 9,200 on Thursday is not an isolated afternoon cell — that is a multi-county system parked over the state for two days.

Why Two of Those Dates Don't Count as Hail Days

This is the part most storm coverage gets wrong, so it's worth being precise.

MRMS publishes a trailing 24-hour maximum. A snapshot taken just after midnight re-reports the previous day's cells under the new date. So a date can show hundreds of flagged cells and still be, in the strict sense, the same hail that fell the day before.

That is what happened on both ends of this window. On August 24 and August 28, every cell of 1 inch or larger had already been reported the previous day. Colorado's season counter — the one on the Colorado hail statistics page — excludes both dates for exactly that reason. So this five-date stretch contributed three new hail days to the season, not five.

The flagged-cell counts in the table above are still real radar returns, and I've left them in. But "the radar reported hail on this date" and "this was a distinct hail day" are two different questions, and conflating them is how a busy week gets inflated into a record one.

What 2.67 Inches Means on a Roof

At 2.67 inches, you are just under baseball size. For context from the jobsite: around 1.5 inches (golf ball) asphalt shingles typically start bruising — the granule matrix compresses and fractures, and it isn't always visible from the ground. By 2 inches you're into more consistent bruising, cracked tab corners, and accelerated granule loss. Approaching 2.75 inches, an older asphalt roof frequently doesn't come through without meaningful damage.

Two honest caveats on that number, though.

First, 2.67 inches was the single strongest cell in the state, not what fell on any particular street. Second, and more important: MESH is tuned so that roughly 75% of measured hail comes in at or below the radar estimate; read a size as a class (within about one tier), not as a measurement. It is a modeled figure with a known lean, not a caliper reading.

And the thing I say on every one of these posts, because it stays true: HailScore™ measures hail exposure recorded in NOAA data. It does not confirm damage. Two houses on the same block can come out of the same storm completely differently. Only someone physically on your roof can tell you what your shingles actually look like.

Where It Hit: the South Metro

The August 26 storm tracked over the south Denver metro, and it produced the two Colorado storm records from this window with enough activity to publish:

That footprint covers a lot of rooftops. If you're in Parker, Lone Tree, Highlands Ranch, Centennial, or Englewood and you haven't looked at your address since the 24th, that's the check worth doing this week.

This also landed on top of an already-heavy August. The August 10–17 stretch put hail over Colorado on eight consecutive days earlier in the month. Roofs don't reset between storms — a roof that took moderate hail two weeks ago and took more last Wednesday is carrying both.

What I'd Actually Do This Week

1. Look up your address. Run it at myhailscore.com. The free report shows radar-detected hail events near your property with dates, estimated sizes, and distances — going back years, not just this week. No signup for the basic score.

2. Check from the ground before you climb anything. Dents in soft metal are the most reliable cheap signal: gutters, downspouts, aluminum flashing, window screens, and the fins on your AC condenser. Shredded garden foliage is another tell.

3. Don't sign anything on the spot. After a storm like this, affected neighborhoods get worked hard by contractors — the good ones and the ones passing through. A legitimate inspection documents what's actually there and tells you honestly if the answer is "nothing." There is no 48-hour deadline that requires you to decide in your driveway.

4. Know your own policy terms. Claim deadlines and coverage rules vary by carrier and policy; confirm yours with your insurer or the Colorado Division of Insurance (doi.colorado.gov). If you're weighing a replacement anyway, the cost and payback math on impact-resistant Class 4 shingles is worth reading before you pick a product.

The Data Behind This

Everything above comes from publicly available NOAA sources — MRMS radar estimates, NEXRAD, the NOAA Storm Events Database, SPC local storm reports, and CoCoRaHS volunteer observations. HailScore aggregates and geocodes that record so it's searchable by street address.

The database now holds 38.1M+ NOAA-linked hail records across all sources, including 8.7M+ NOAA NEXRAD radar records going back to 2015. That depth is the point: your address's history isn't just last Wednesday, it's a decade of accumulated exposure, which is the context a single storm report can never give you.

The August 24–28 outbreak is in that database now. If you're anywhere on the Front Range, it takes about a minute to see where you stand.