Your Hurricane Date of Loss Isn't the Landfall Date

Hurricane Ian's page on the National Hurricane Center archive holds 208 separate text products, and 68 of them are public advisories, 38 numbered and 30 intermediate ³. Those 68 carry the watch and warning language a Florida file has to rely on. The first went out at 0900 UTC on 23 September 2022, the last at 0300 UTC on 2 October. The final three came from the Weather Prediction Center, not NHC, which signed off after Advisory 35 on the morning of 1 October ³. An adjuster arguing whether a Lee County roof sits inside the hurricane deductible has to find the right advisories in that stack, attribute them to the office that actually issued them, then produce them in a form a court will accept.

What § 627.4025(2)(c) requires after the 2023 amendment

The statute sets one statewide occurrence window. The window begins when the National Hurricane Center issues a hurricane warning for any part of Florida, and it ends 72 hours after the last hurricane watch or hurricane warning for any part of the state terminates ¹. Landfall does not appear in the statute.

The pre-2023 version read differently in ways that still matter on older files. It opened on a watch or a warning, and it kept the window open while hurricane conditions existed anywhere in Florida ². Section 13 of chapter 2023-130 pulled the watch out of the front-end trigger, deleted the continuing-conditions clause, and left the 72-hour tail running from termination of the last watch or warning ¹. A 2021 or 2022 date of loss is governed by the text in force then, so pull the correct annual version before writing the coverage position ².

Statewide reach is the part that surprises desk adjusters. A Polk County loss can fall inside the window with nothing resembling hurricane conditions on the ground there, provided a warning was up somewhere in Florida and the loss predates the 72-hour tail. The proof of both endpoints lives in the contemporaneous advisory text, which is why the archive matters more than the storm track.

Which copy survives a coverage fight?

NOAA's own guidance says NCEI is "the only place that you can obtain certified copies of U.S. weather information," and that free pulls from NDBC, NHC, and NWS pages carry no certification ⁴. A Department of Commerce certified copy of a public record self-authenticates under Federal Rule of Evidence 902(4) ⁶, so no records custodian has to be produced. That rule governs federal proceedings; state evidence codes carry their own parallel self-authentication provisions, so check the local rule rather than assuming the federal one travels. Order the certified copy from NCEI at the point the file turns adversarial, not the week of mediation.

In wind-versus-water disputes the certification question sits on top of a burden question, and the case adjusters most often cite for it gets characterized wrong. In Broussard v. State Farm Fire & Casualty Co., 523 F.3d 618 (5th Cir. 2008), the Fifth Circuit reversed the judgment as a matter of law the district court had entered for the homeowners and remanded for a new trial ⁵. What survives for daily file work is the allocation the court applied: under an open-peril policy the insured carries the burden of proving direct physical loss, and the carrier carries the burden of proving that an exclusion, including the water exclusion, bars the claim ⁵. Broussard is Fifth Circuit law applying Mississippi law, so a Florida file uses it as persuasive framing rather than controlling authority. Calling Broussard a policyholder win misreads the disposition, and defense counsel will correct it in front of whoever you were trying to persuade.

Build the record in the order a reviewer will read it. Contemporaneous advisories come from the NHC data archive, with the post-storm Tropical Cyclone Report arriving months later as supporting best-track analysis rather than a replacement for the contemporaneous text. Ground-truth observations are a separate pull: the live NWS Local Storm Report service only keeps a short rolling window before reports age off, so any older date of loss has to come from the Iowa Environmental Mesonet LSR archive or the NCEI Storm Events Database.