claims

Date-of-Loss Verification, Explained

Florida ties a storm claim's date of loss to NOAA verification, and the public record's own day boundaries make that date harder to pin than it looks.

The date is not a formality

In Florida, the date of loss on a weather claim is not a field somebody fills in from memory. It is defined by statute. Section 627.70132 of the Florida Statutes provides that for claims resulting from hurricanes, tornadoes, windstorms, severe rain, or other weather-related events, the date of loss is the date that the hurricane made landfall or the tornado, windstorm, severe rain, or other weather-related event is verified by the National Oceanic and Atmospheric Administration.

Here is why that matters more than it sounds. The same section says a claim or reopened claim is barred unless notice of the claim was given to the insurer in accordance with the terms of the policy within one year after the date of loss, and that a supplemental claim is barred unless notice was given within eighteen months after the date of loss. The date does not merely start a clock. It decides whether there is a clock left.

The section also contains a tolling provision. Those time limits are tolled during a deployment to a combat zone, or a combat support posting, that materially affects a named insured servicemember's ability to file. And subsection (5) states that the section does not affect any applicable limitation on civil actions provided in section 95.11. That separate statute lists, within five years, an action for breach of a property insurance contract, with the period running from the date of loss.

So in that one state, a single verified date feeds a notice window, a supplemental window, and the start of a suit limitation period. Three different clocks, one input.

This is one state, and you have to check yours

Everything above is Florida. Other states define the date of loss differently, or do not define it by statute at all and leave it to the policy wording. A policy can also impose its own notice requirements and its own contractual suit limitation on top of whatever the statute says.

Read your own state's statute, then read your own policy, then talk to someone licensed to advise you. Nothing here tells you what your deadline is, and no article can, because the answer depends on a state and a policy form that this page does not know.

Where "verified by NOAA" sends you

A statute can point at a data product without describing how that product behaves. The public record on a given storm comes in distinct pieces, and they do not arrive together.

The near real time piece is the daily storm report. The Storm Prediction Center compiles it from NWS Local Storm Reports, usually sent in near real time, and SPC labels those reports preliminary and shows them as is. The slower piece is Storm Data. The NWS states that both the Storm Data publication and the reports in the Storm Events Database are available 90 to 120 days after the event, and that certified, official copies must be requested from NCEI rather than from a local NWS office.

That means the date on an early file and the date supported by a citable record can be separated by months of elapsed time, even when they agree.

The day boundary problem

Now the mechanical trap. SPC groups its daily storm reports into a day that runs from 1200 UTC to 1159 UTC. That is not midnight to midnight local time. The current day's page updates every ten minutes starting at 6 AM CST or 7 AM CDT, and yesterday's page covers 6 AM to 6 AM local time.

A storm that hits at nine in the evening local time therefore sits on a report day that is labeled with one calendar date, while a homeowner describes the storm using the date they saw on their phone that night. Those two dates can differ. Nobody is lying. The report day and the calendar day are different units.

Local Storm Reports themselves carry near real time timestamps, which is the way out of the ambiguity. A timestamp resolves what a date label cannot. Radar-derived records are organized per day as well, so the same care applies when a file leans on them.

Two storms, one month, one roof

The failure this produces is specific and common: two separate events in the same market inside a few weeks, and a file that attaches the damage to the wrong one. Once that happens, every downstream argument inherits the error. Scope, causation, and whatever notice window applies all hang off a date that was assigned by convenience.

Severity bands help separate the events, because the public thresholds are fixed. The Storm Prediction Center defines severe hail as hail 1 inch in diameter or larger, significant severe hail as 2 inches or larger, and significant severe wind as 75 mph or greater. If one event cleared the 1 inch line and another reached 2 inches in the same market weeks apart, those are different events with different expected signatures, and the file should say which one it means.

The honest move, when the record cannot separate them, is to say the record cannot separate them. That is a smaller claim than picking one, and it survives scrutiny.

Why a whole season makes the date harder

There is a tempting assumption that more storms mean more evidence. For dating a single loss, the opposite is closer to true. Each additional event adds a candidate date and adds damage that has to be allocated to one of them.

Laboratory work on smaller hail sharpens the point. A study published September 29, 2025 in Frontiers in Materials, Sub-severe hail: the missing piece in assessing asphalt shingle risk in North America, fired ice spheres of 17.8 mm, about 0.7 inch, and 25.4 mm, about 1 inch, at asphalt shingle specimens. The concentration was 44 impacts per square foot, chosen from available field data as a realistic high-concentration scenario, and the authors note that no standard threshold exists. Each specimen took 1,000 impacts across two series.

The results cut against the older assumption. Earlier work had suggested that hailstones smaller than 1 inch rarely cause functional damage, which the paper defines as punctures or fractures that compromise water shedding, or damage that otherwise reduces service life. In this test, cumulative granule loss from the sub-severe impacts in the first series exceeded the total granule loss from a single 50.8 mm, or 2 inch, impact. Table 6 reports 9.5 square mm per sub-severe hit for the control group and 10.2 square mm for the experimental group, against 20.0 square mm for a single 2 inch impact on new product.

Then the weathering effect. On naturally weathered specimens, granule loss from a 50.8 mm impact averaged 192.2 square mm per impact, against 57.2 square mm for the control and 20.0 square mm for the new baseline. Average performance decline was 47%, with a range of 39% to 67%.

Read the limits with the findings. The authors tested only six products, two specimens each, and call the sample small while saying more testing is needed. This is laboratory work on a handful of products, not a field survey. What it supports is a cautious inference rather than a conclusion: a roof's condition on the day of the event everyone is arguing about may already reflect earlier, smaller events. Which is exactly why the date has to be established on its own evidence instead of inferred from how bad the roof looks.

What a defensible date looks like in a file

A date that holds up has four things attached to it.

A timestamp, not just a calendar date, taken from the Local Storm Report or the equivalent, with the time zone or UTC offset written out. A named source, meaning the specific product you pulled from rather than the agency that runs it. The date you made the pull, because the early record changes as reports come in and the 90 to 120 day record can read differently. And a saved copy of the pull itself.

Storm Stream is an API over the same public feeds: its exact-date history path reads NOAA NCEI radar-estimated hail for one day at one point and labels the result verified, while its multi-year sweep is labeled modeled and unverified.