What the record shows about hail and roofing work in your area
Two NOAA products and one city permit series, each named, each with the period it covers. The two hail products do not agree with each other, and the page shows you both rather than picking the more dramatic one.
This tool has not seen your roof. It cannot tell you whether yours is damaged, whether it needs replacing, or when — only what was recorded over the area around you.
Which area?
Enter a ZIP and we will say which county it sits in and put that county's reading first. Both metros are on this page either way. Nothing is stored and nothing is sent anywhere.
Five digits. Austin and San Antonio metros.
Enter a ZIP above and we will name its county. Until then, both metros are below.
Narrowing to your county needs JavaScript, because it is a lookup on something you type. Everything else on this page is already in the page: both metros' confirmed-hail counts by county, both permit readings, and the licensing position are all below, whether scripts run or not.
Radar hail signatures
NOAA's Severe Weather Data Inventory records the storm cells NEXRAD flagged as probably producing hail. A signature is a radar detection over a box on the map. It is not a report that hail fell, and it is not a report about any address.
Radar hail signatures over the Austin box
268
SWDI nx3hail signatures recorded Aug 10, 2026 to Sep 12, 2026, over a box 0.5° of latitude and longitude either side of 30.2672, -97.7431 — one reference point for the Austin metro. That is a rectangle on a map, not a county and not a city limit.
LIVERadar hail signatures over the San Antonio box
190
SWDI nx3hail signatures recorded Aug 27, 2026 to Sep 12, 2026, over a box 0.5° of latitude and longitude either side of 29.4241, -98.4936 — one reference point for the San Antonio metro. That is a rectangle on a map, not a county and not a city limit.
LIVEConfirmed hail reports
NOAA's Storm Events Database records hail that a person reported and the National Weather Service accepted. It is filed by county, it runs months behind the weather because a human confirms each entry, and it is a much smaller number than the radar one. Both are true.
Confirmed hail reports, Austin metro counties
20
Recorded Oct 24, 2025 to Jun 15, 2026, filed by county. The newest record is 98 days old: NCEI publishes this series on a lag because a person confirms every entry, so that gap is the publisher's and not this site's.
| County | Confirmed hail | Other events on record |
|---|---|---|
| Williamson | 9 | 21 |
| Travis | 5 | 20 |
| Burnet | 4 | 12 |
| Blanco | 1 | 6 |
| Hays | 1 | 4 |
| Bastrop | 0 | 1 |
| Caldwell | 0 | 4 |
A county showing 0 in the first column and a number in the second is one NOAA files events for and which reported no hail in this window. That is a reading, not a gap.
LIVEConfirmed hail reports, San Antonio metro counties
11
Recorded Aug 31, 2025 to Jun 20, 2026, filed by county. The newest record is 93 days old: NCEI publishes this series on a lag because a person confirms every entry, so that gap is the publisher's and not this site's.
| County | Confirmed hail | Other events on record |
|---|---|---|
| Atascosa | 3 | 3 |
| Bandera | 3 | 6 |
| Comal | 3 | 10 |
| Guadalupe | 1 | 6 |
| Kendall | 1 | 6 |
| Bexar | 0 | 46 |
| Medina | 0 | 5 |
| Wilson | 0 | 1 |
A county showing 0 in the first column and a number in the second is one NOAA files events for and which reported no hail in this window. That is a reading, not a gap.
FEED DOWNWhy the two hail numbers do not match, and why we show both
- One is a radar detection, the other is a person. A radar hail signature is an algorithm's read of a storm cell's reflectivity profile — NEXRAD saying this cell probably contains hail. A confirmed report is someone who saw hail, told the National Weather Service, and had it accepted into the record. A storm can throw dozens of signatures across a metro and produce no confirmed report at all, because nobody was standing where it fell or nobody called it in.
- They are not counted over the same shape. Storm Events files by county. The radar product returns no county at all, so its records are pulled over a box drawn around a single reference point for each metro — which is not a county, does not follow any boundary, and includes and excludes places a county line would not. Nothing on this site converts one shape into the other.
- Neither one is a statement about a house. A signature over a metro box and a report filed for a county are both area readings. Neither says hail struck a particular roof, and no arithmetic on this page can get from one to the other. What they are good for is knowing whether this was a season worth looking at your roof after.
Re-roofing work on the public record
Each city is counted by its own mechanism, and the two mechanisms are not the same instrument. San Antonio publishes a dedicated Re-Roof Permit class and we count it. Austin publishes no roofing permit type at all, so its figure is a text match on the permit description — which sweeps in rooftop solar, rooftop equipment and anything else whose wording happens to contain the word. Read each number against its own city over time. Setting one against the other would measure the difference between two permit vocabularies, not a difference between roofs.
Re-roof permits on record, Austin
1,860
Across 12 complete months, September 2025 to August 2026. Classified by matching text in the permit description, because this city publishes no roofing permit type.
What that sweeps in is measured, not guessed at. Across the 2,037 roof-matched permits in the archive behind our own Austin roof-permit page, 650 descriptions — 32% — mention solar, photovoltaic or PV, and 402 — 20% — use wording that explicitly describes replacing a roof covering. Rooftop solar is the single largest contributor to the total.
Read this as roof-related construction activity, which is what it measures, rather than as roof replacement volume, which it is broader than. Treat the replacement share as a floor: some jobs describe a replacement in phrasing the match does not catch. And do not set this total against another Texas city's roofing count. A city that publishes a dedicated re-roof permit type is measuring something narrower by a different mechanism, so the difference between the two numbers would be a fact about permit vocabularies rather than about roofs.
This is work filed with one city, counted by Austin's own mechanism. It is not a count of roofs replaced, it says nothing about any particular address, and it is not comparable with the other metro's number on this page.
LIVERe-roof permits on record, San Antonio
4,896
Across 12 complete months, September 2025 to August 2026. Classified by the city's own permit type — "Re-Roof Permit".
This is work filed with one city, counted by San Antonio's own mechanism. It is not a count of roofs replaced, it says nothing about any particular address, and it is not comparable with the other metro's number on this page.
LIVEWho is allowed to do the work
Texas licenses HVAC, electrical and mold work. It does not license roofing.
The Texas Department of Licensing and Regulation publishes the list of occupations it regulates. Air Conditioning and Refrigeration Contractors, Electricians, and Mold Assessors and Remediators are on it. Roofing is not — so there is no TDLR licence number to ask a roofer for, the way there is for the trades either side of it. That does not make a roofer unqualified; it means a licence number is not the thing to check. What is checkable: liability and workers' compensation insurance carried in the company's own name, how long the business has traded under that name, and the manufacturer certification behind whatever workmanship warranty is offered.
What this tool cannot see
- Whether your roof is damaged. We have not seen it, and nothing in these feeds is about your address. Hail recorded a mile away tells you the season was worth a look; it does not tell you what it did to your roof. That is what a person on a ladder is for.
- How big the hail was. The radar product's size column has no published unit. NOAA's own REST usage documentation for this service does not define the columns and does not mention that field at all, and the values are as consistent with inches as with anything else — and inches and millimetres are a factor of twenty-five apart. So this page publishes counts and no sizes. A number whose unit we would be guessing at is worse than no number.
- Your roof's age, area, or what a replacement would cost. There is no route to any of the three. Travis Central Appraisal District publishes no street address in its improvement export, so nothing can be looked up by address, and Bexar publishes no bulk export at all. Building-footprint data that would give a roof area was set aside on licensing grounds. And the declared valuation on a permit is an applicant's fee-basis statement to the city, not a paid invoice, so it is not a cost.
- When to replace it. We are not going to guess. Age alone does not decide it, a hail season does not decide it, and neither does a permit filed down the street. A tool that turned an area's storm record into a replacement date for your house would be selling something, not measuring anything.
Where these numbers come from
- Austin confirmed hail. NOAA Storm Events Database — https://www.ncei.noaa.gov/pub/data/swdi/stormevents/csvfiles/
- San Antonio confirmed hail. NOAA Storm Events Database — https://www.ncei.noaa.gov/pub/data/swdi/stormevents/csvfiles/
- Radar hail signatures. NOAA Severe Weather Data Inventory, nx3hail — https://www.ncdc.noaa.gov/swdiws/. Built and registered; no dataset committed yet, so no count is published above.
- Austin re-roof permits. City of Austin Issued Construction Permits (Socrata) — https://data.austintexas.gov/
- San Antonio re-roof permits. City of San Antonio Permits Open Data — https://data.sanantonio.gov/
- Licensing. Texas Department of Licensing and Regulation — Programs Licensed and Regulated by TDLR — https://www.tdlr.texas.gov/licenses.htm