Skip to contentCALCHIS

Case study

Hurricane Milton, replayed: three triggers, three outcomes

A replay, not a record. Calchis was not monitoring triggers during Milton. This runs NOAA’s post-season best track through the same arithmetic the live evaluator uses. Live, the evaluator reads the Hurricane Center’s operational advisories, which the best track later revises, so the figures a desk saw that night can differ. The three triggers are illustrations, not any bond’s terms.

On the morning of 8 October 2024 Milton was a Category 5 hurricane in the southern Gulf: 145 kt sustained winds and a central pressure of 902 mb. It weakened on the way in and made landfall near Siesta Key at Oct 10 00:30Z as a Category 3, then crossed the peninsula to the Atlantic. We placed three parametric triggers around Tampa, the kind a fund might hold against a Florida wind position, and replayed the storm through them point by point.

Sustained wind ≥ 96 kt (Category 3), center within 100 km of Tampa

Breached at Oct 10 00:00Z: 105 kt, 76 km from Tampa (109.4%).

Best-track timeFrom TampaWind (kt)Of thresholdStatus
Oct 8 00:00Z1,053 km145—outside radius
Oct 8 06:00Z978 km125—outside radius
Oct 8 12:00Z879 km125—outside radius
Oct 8 18:00Z823 km140—outside radius
Oct 8 22:05Z745 km145—outside radius
Oct 9 06:00Z576 km140—outside radius
Oct 9 12:00Z402 km135—outside radius
Oct 9 15:00Z316 km125—outside radius
Oct 9 18:00Z239 km115—outside radius
Oct 9 21:00Z158 km110—outside radius
Oct 10 00:00Z76 km105109.4%breached
Oct 10 00:30Z63 km100104.2%breached
Oct 10 03:00Z39 km8689.6%approaching
Oct 10 06:00Z115 km70—outside radius
Oct 10 09:00Z214 km68—outside radius
Oct 10 12:00Z312 km65—outside radius

Sustained wind ≥ 113 kt (Category 4), center within 150 km of Tampa

Never fired. Closest: 92.9% at Oct 10 00:00Z (105 kt), an approaching alert and no breach.

Best-track timeFrom TampaWind (kt)Of thresholdStatus
Oct 8 00:00Z1,053 km145—outside radius
Oct 8 06:00Z978 km125—outside radius
Oct 8 12:00Z879 km125—outside radius
Oct 8 18:00Z823 km140—outside radius
Oct 8 22:05Z745 km145—outside radius
Oct 9 06:00Z576 km140—outside radius
Oct 9 12:00Z402 km135—outside radius
Oct 9 15:00Z316 km125—outside radius
Oct 9 18:00Z239 km115—outside radius
Oct 9 21:00Z158 km110—outside radius
Oct 10 00:00Z76 km10592.9%approaching
Oct 10 00:30Z63 km10088.5%approaching
Oct 10 03:00Z39 km8676.1%monitoring
Oct 10 06:00Z115 km7061.9%monitoring
Oct 10 09:00Z214 km68—outside radius
Oct 10 12:00Z312 km65—outside radius

Central pressure ≤ 950 mb, center within 100 km of Tampa

Never fired. Closest: 90.5% at Oct 10 00:00Z (956 mb), an approaching alert and no breach.

Best-track timeFrom TampaPressure (mb)Of thresholdStatus
Oct 8 00:00Z1,053 km902—outside radius
Oct 8 06:00Z978 km925—outside radius
Oct 8 12:00Z879 km929—outside radius
Oct 8 18:00Z823 km916—outside radius
Oct 8 22:05Z745 km902—outside radius
Oct 9 06:00Z576 km909—outside radius
Oct 9 12:00Z402 km919—outside radius
Oct 9 15:00Z316 km932—outside radius
Oct 9 18:00Z239 km945—outside radius
Oct 9 21:00Z158 km951—outside radius
Oct 10 00:00Z76 km95690.5%approaching
Oct 10 00:30Z63 km95887.4%approaching
Oct 10 03:00Z39 km96674.7%monitoring
Oct 10 06:00Z115 km975—outside radius
Oct 10 09:00Z214 km979—outside radius
Oct 10 12:00Z312 km983—outside radius

What the replay shows

The wording decides. One storm, three triggers, three different answers. The Category 3 trigger breached the first time the center came inside its radius. The Category 4 trigger never fired, though Milton was a 140 kt storm a day earlier: it weakened before its center reached the circle, and a trigger is judged where it is written, not where the storm was strongest. The pressure trigger came within 6 mb and stopped.

Near misses are information. Two of the three would have sent an approaching alert and never breached. For a desk deciding whether to hedge before the next advisory, “90% and weakening” is a different decision from “90% and strengthening”, and both are different from silence.

Every reading is on the record. Live, each of these rows would be an evaluation with the source’s own figures, the geography test used and the trigger’s terms, named by a hash of its content so a counterparty can check it without trusting us. None of that settles a bond: the calculation agent’s determination does.

Method and source

Positions, winds and pressures: IBTrACS v04 (NOAA NCEI), best track, storm MILTON 2024 (NOAA NCEI), public domain. Post-season best track, not the operational advisories a live evaluation reads. Each point is judged as the live evaluator judges a point-and-radius trigger: the storm center inside the radius, wind at or above its threshold, pressure at or below its threshold, closeness as a plain ratio for wind and as the pressure deficit below 1013.25 mb for pressure, approaching at 80%. The full method is on How Calchis determines a trigger; the record format is the Determination Standard.