Back-to-Back Colorado Hail, July 16–17: What the NOAA Radar Actually Recorded

I check NOAA radar before I check my phone most mornings during hail season. Last week was a good reminder of why.

Colorado took back-to-back hail hits on July 16 and July 17. When I pulled the MRMS data on the morning of the 18th, the pattern was worth paying attention to: the radar flagged hail 111 separate times on July 17, concentrated in a small cluster of grid cells, with a max stone size of 1.34 inches — and it was roughly the same ground that got hit the day before. That's not a metro-wide sweep, but 1.34 inches is well above the size where functional damage starts showing up on older roofs, and two consecutive days of hits on the same area is exactly the pattern that gets missed.

I'm going to walk through exactly what the data shows, what that stone size means in practical terms, and what steps are worth taking this week if your address was in the storm path.

What NOAA MRMS Recorded

NOAA's Multi-Radar Multi-Sensor (MRMS) system uses a network of Doppler radar stations to estimate hail size and location in near real time. It's the same data backbone that HailScore uses to build its database — currently 24.5M+ verified hail records across all sources, including 6.2M+ NOAA NEXRAD radar records going back to 2015.

Here's what MRMS logged for July 16–19 in Colorado:

A detection is one radar observation of hail-producing conditions in a grid cell — the radar re-checks constantly, so an area getting hit repeatedly racks up detections through the day. Here's the part worth knowing: on July 16 and July 17, every detection landed in the same nine grid cells (each roughly a square kilometer). Not similar areas — the identical nine cells, two days in a row. That's not a wide swath; it's a small patch of Colorado that took a repeated pounding.

A quick note on what this data actually means: MRMS measures radar-estimated hail exposure — where the atmosphere produced hailstones at that size and location. It does not confirm that every property in those grid cells has visible damage. Ground-truth confirmation requires a physical inspection. That distinction matters, and I'll come back to it.

What 1.34-Inch Hail Does to a Roof

I've been in residential roofing in Colorado for about three years. I've stood on a lot of roofs after storms in this size range, and 1.34 inches is right at the inflection point where you need to look closely rather than assume either way.

For context on stone size:

1.00 inch (quarter-sized): Functional damage is possible on 3-tab shingles and architectural shingles with less than 10 years of UV exposure. Look for spatter marks, granule loss concentrated in circular patterns.

1.25 inches (half-dollar-sized): Damage is probable on standard architectural shingles, especially in the 10–15 year range. At this size, the shingle mat itself can fracture without any visible surface bruising — which is why visual-only inspections miss things.

1.34 inches: You're in the damage-likely zone for mid-life shingles. Metal, tile, and newer Class 4 impact-resistant asphalt handle this size reasonably well. Standard 25- or 30-year architectural shingles from the 2000s–2015 era are the ones to check.

Wood shake roofs are the most vulnerable at any size above 1 inch. If you have a shake roof that took this storm, an inspection is worth prioritizing.

Season-to-Date: Where 2026 Stands

Colorado has now logged 57 hail days so far in 2026, with a season-to-date maximum stone size of 4.94 inches. That max number is significant — stones approaching 5 inches are well into the structural-damage range, and that reading somewhere in Colorado this year tells you it's been an active season. The full Colorado hail statistics page tracks these numbers year-round.

For comparison, a "typical" Colorado hail season runs April through September, with peak frequency in May through early July. We're now in the back half of July, which means we're transitioning from peak season toward the tail end — but the data clearly shows we're not done. Mid-July afternoon convective storms are a regular feature of Colorado summers, and this past week proved it.

How to Use HailScore to Check Your Address

HailScore pulls from that same MRMS data — plus NOAA Storm Events, CoCoRaHS volunteer reports, and NEXRAD radar records — and scores any US address on a 0–100 scale based on hail exposure history. The score reflects how much radar-documented hail activity has fallen near your property over time, weighted by stone size, frequency, and recency. It does not confirm damage, and it doesn't replace an on-roof inspection.

What it does do is give you a data-backed starting point. If your score comes back in the orange or red range and your roof is more than 10 years old, that's a meaningful signal to get eyes on it. If your score is low, that context is worth knowing too.

You can check whether hail hit your address for free — the score, storm map, and event timeline are always free, no email required to see the basics.

What to Do This Week

If your roof is 10+ years old and you were in the July 16–17 footprint:

Do a quick visual from the ground. Look at soft metals first — gutters, downspout elbows, AC unit fins, windowsill caps. These dent readily and leave a clear record. If you see spatter patterns there, the roof took a similar hit.

Check your skylight and ridge cap flashing. Hail marks on metal flashing are easier to see and photograph than bruising on shingles.

Look at your shingles from the edge of the roof (safely — don't walk it yourself unless you're trained). You're looking for granule pooling in gutters and circular scuff marks on the shingle surface.

Don't call the first contractor who knocks on your door. Colorado gets a wave of out-of-state storm chasers after every significant event. Hire someone with a local license who will be around to stand behind the work.

Document before anything gets touched. Photos with timestamps matter, and if you need the weather record itself as a citable document, the Official Hail History Report is a one-off PDF listing the NOAA Storm Events and NEXRAD radar records within a mile of your address over the past decade, each with its government source ID.

The Honest Framing

HailScore's database — 24.5M+ records and counting — is built on NOAA radar data. That data is the most consistent, verifiable record of where hailstorms have been. But radar estimation has limits. Stone size estimates can run 10–15% high or low. Coverage gaps exist in complex terrain. And radar can't see through your shingles to tell you if the mat is cracked.

What the data is good for is pattern recognition at scale: knowing that a neighborhood has taken repeated 1-inch-plus events over five years, for example, is meaningful context whether you're a homeowner, an adjuster, or a contractor trying to prioritize where to canvass.

The July 16–17 event is now part of that record. If your address was in the footprint, that history is already in the database. What it means for your specific roof still requires someone standing on it.

---

Alex Chicilo is a Colorado-based residential roofer with about three years of field experience and the founder of HailScore — a free tool that scores any US address for hail exposure risk using NOAA radar data. Questions? Reach out at hello@myhailscore.com.