NHC model cycle 12Z Fri 9 Oct 2026 · north-central Gulf of Mexico
Physics-based, statistical and AI models are all forecasting the same Gulf hurricane. This page shows where they agree, where they split, and whether the split is unusual. It uses NOAA's public model file and three seasons of verification.
Isaias is at its peak now. NHC's forecast holds it near 100 kt until landfall and weakens it fast over land: the center is just offshore at +12 h, 7 PM Friday CDT, and inland by +24 h. NHC's discussion points to "a potential landfall along the extreme western Florida Panhandle coast this evening." The models disagree on two numbers: how strong the storm is at the coast, and how fast it falls apart over land.
Isaias strengthened from 90 kt at 06Z to 105 kt at 12Z, a major hurricane, after an Air Force Reserve Hurricane Hunter aircraft measured a 120 kt flight-level wind. The early aids all start from that 105 kt. NHC's discussion says shear of 30–40 kt "has definitely increased as forecast," but is "nearly aligned with the storm's motion, which limits its adverse effects," and that HAFS-A, HAFS-B, HWRF, SHIPS and LGEM "all suggest that Isaias could maintain major hurricane intensity up until its landfall this evening." In the public model file, the interpolated versions of those aids are at 81–89 kt (HAFS, HWRF) and 97–102 kt (SHIPS, LGEM) at +12 h.
Isaias sat nearly stationary in the Bay of Campeche for about two days, first as a weak disturbance and then as a depression. It then turned east-northeast and intensified rapidly, from 35 kt at 06Z Wednesday to 70 kt at 06Z Thursday. That is 35 kt in 24 hours. In NHC's operational best track it reached 90 kt at 06Z Friday and 105 kt at 12Z Friday. NHC has revised earlier points since they were first posted (12Z Thursday from 70 to 75 kt, 18Z Thursday from 85 to 80 kt); the scores below use the latest best track. Early on, the models disagreed about both the turn and the intensification. Some moved the system east quickly and strengthened it. Others kept it slow, weak and farther south. Those forecasts can now be checked against NHC's best track.
The 12Z Wednesday runs, made 48 h before Friday's 12Z position, show who handled the turn. DeepMind's raw and interpolated runs both placed the center 25 nmi from where it was, the closest of any aid with its own track; NHC's official forecast was 36 nmi off. Scored against the latest best track, raw DeepMind has been the closest aid with its own track at two days, or tied for it, in each of the five runs from 12Z Tuesday to 12Z Wednesday (6, 13, 21, 19 and 25 nmi). The raw HAFS-A and HAFS-B runs from 12Z Wednesday missed by 89 and 112 nmi. Single runs vary, which is why the averages below matter more. Over every forecast so far that all the aids below share, the interpolated DeepMind aid has the smallest average track error, 16 nmi, and NHC is next at 18 nmi. NHC's discussion says its track is "closest to the latest HCCA and GDMI guidance."
The overnight intensification caught most aids low. The raw 12Z Wednesday COAMPS-TC run came closest, forecasting 104 kt for 12Z Friday against 105 kt in the best track, though it missed the track by 164 nmi. The raw DeepMind run had 94 kt, NHC 95 kt and the raw HWRF 92 kt; the raw HAFS-A and HAFS-B runs had 71 and 84 kt. Across the shared forecasts, the smallest intensity errors belong to the corrected consensus (HCCA), NHC, DeepMind and the neural-network consensus (NNIC), between 5.0 and 7.1 kt.
Nearly every aid was too slow to intensify Isaias. Every aid in the table except COAMPS-TC has a low bias. The global models ran the lowest, as expected at their resolution, and the statistical LGEM was 16 kt low. Two days out, the raw GFS and its ensemble mean put the center 79 and 154 nmi from where it was. COAMPS-TC had the largest track error of the regional models.
| Aid | Type | Intensity error | Intensity bias | Track error |
|---|---|---|---|---|
| HCCA | Consensus | 5.0 kt | −4.1 | 26 nmi |
| OFCL | NHC official | 5.5 kt | −3.8 | 18 nmi |
| GDMI | AI (DeepMind) | 6.5 kt | −5.0 | 16 nmi |
| NNIC | Machine learning | 7.1 kt | −6.2 | — |
| CTCI | Regional physics | 7.8 kt | +2.9 | 89 nmi |
| HMNI | Regional physics | 8.4 kt | −7.4 | 56 nmi |
| HWFI | Regional physics | 9.0 kt | −8.4 | 47 nmi |
| SHIP / DSHP | Statistical | 10.6 kt | −10.3 | — |
| HFBI | Regional physics | 11.3 kt | −8.4 | 57 nmi |
| HFAI | Regional physics | 11.6 kt | −10.8 | 59 nmi |
| AVNI | Global physics | 15.9 kt | −15.9 | 51 nmi |
| LGEM | Statistical | 16.1 kt | −15.7 | — |
| AEMI | Global physics | 24.7 kt | −24.7 | 76 nmi |
Mean absolute errors at +12 to +48 h, over the 28 forecast cycle and lead time pairs that every listed aid shares (a homogeneous sample), scored against NHC's operational best track through 12Z Friday. Statistical and ML aids borrow another model's track, so their track errors are not listed. This covers two to three days of one storm against a preliminary best track. It shows which models handled this storm's genesis and rapid intensification, not which models are better overall.
At +12 h, with the center just offshore in NHC's track, the aids span 81–103 kt. The Canadian ensemble mean and deterministic run (102 and 103 kt) and the statistical LGEM (102 kt) are highest. The regional physics models are at 81–89 kt and the interpolated DeepMind aid at 84 kt. NHC's 100 kt is near the top of the range.
By +24 h, with every plotted aid's center inland, the regional models are at 40–47 kt and the DeepMind aid at 48 kt. The Canadian ensemble mean (86 kt), the UK global model (74 kt) and the GEFS mean (69 kt) are much higher. NHC is at 55 kt.
At +24 h SHIPS, a statistical model with no land in it by design, holds 70 kt. Its land-aware version, Decay-SHIPS, is at 53 kt. The 17 kt between those two is the land effect as that statistical scheme models it, which makes the pair a useful control. By +36 h the DeepMind aid is at 12 kt, its last forecast point, and the regional models at 19–31 kt. The UK model and Canadian ensemble mean are still at 63 and 74 kt.
As in the last cycle, the spread is not mainly track spread. At +24 h every plotted aid has the center within 32.0–32.7°N and 87.1–88.0°W, yet their intensities range from 40 to 86 kt. Much of the disagreement is in how each model weakens a hurricane over land.
The DeepMind line on most spaghetti plots is GDMI, an interpolated aid. The raw model output (GDMN) arrives too late for the forecast cycle it would serve. NHC therefore takes the previous run and adjusts it to the storm's intensity at the start time.
For Isaias, the raw 06Z DeepMind run forecast 96 kt for 12Z, against 105 kt in the best track: it was too low at the start, because Isaias strengthened faster than it predicted. The aids were adjusted to 105 kt. After the adjustment, the 12Z aid runs 9 kt above the raw run at 0 and +12 h and 18 kt above it at +24 h. This is the opposite of the first three cycles shown on this page, when the raw run started too high and the adjustment pulled the aid down. The adjustment follows the observed intensity in either direction.
"AI is less aggressive" can be partly an artifact of this adjustment, as it was in earlier cycles. Compare raw and interpolated output before drawing conclusions about any model class.
Intensity forecasts over land depend on when and where the center crosses the coast. The tracks below cluster tightly toward the Alabama and western Florida Panhandle coast. At +12 h every plotted aid has the center between 29.4°N and 30.1°N and between 86.8°W and 87.5°W, near the coast; by +24 h all are inland. NHC's discussion says aircraft fixes show Isaias "has drifted a bit east of the previous forecast track."
Not among the core models. The split comes from the models outside that core.
We took eight aids that ran in every season from 2023 to 2025: HAFS-A, HAFS-B, HWRF, HMON, COAMPS-TC, GFS, Decay-SHIPS and LGEM. For every Atlantic forecast cycle that started at hurricane strength, we computed their spread, the standard deviation across models. By that measure, Isaias's disagreement is ordinary at +24 h (47th percentile) and on the low side at +48 h (18th), when the storm is well inland. Fewer than six of the eight core aids reach +72 h this cycle, so no +72 h spread is shown.
Isaias stands out once the other models come back in. Adding the UK global model, the Canadian ensemble mean and SHIPS raises the +48 h spread from 5.4 kt to 16.0 kt. (The DeepMind aid is normally in this comparison too, but it has no +48 h value this cycle.) That figure uses a different set of models, so it can't be compared to the histograms. It does show where the disagreement comes from: models that treat land very differently.
NOAA's global ensemble (GEFS) runs 31 versions of the GFS, each with slightly perturbed starting conditions and model physics. At this resolution its intensities are much too weak. At 06Z the members started between 55 and 84 kt against an observed 90 kt, so read the plume for timing, not strength. On the 12Z clock, 18 of the 31 members weaken fastest between +6 and +18 h, and 30 of the 31 lose at least 15 kt in that window. It spans 18Z Friday to 06Z Saturday (1 PM Friday to 1 AM Saturday CDT), consistent with NHC's forecast landfall this evening. That is independent evidence that the timing of the move inland is fairly well constrained.
The European (ECMWF) model and ensemble are not in NOAA's public model file, so they are not shown. Google's model is itself an ensemble, but the public file carries one forecast per cycle.
Each version of this page, one per NHC model cycle, is committed to a public repository before landfall, and the commit history timestamps it. After the storm, each aid will be scored against NHC's best track from +12 to +72 h, on the questions this storm raises:
One storm can't settle which model is better. NHC's annual verification reports are where models get judged. This page is a worked example of reading the guidance.
This page is updated as new NHC model cycles arrive. Each earlier version is preserved unchanged.