An initial heads up on Matthew: we have a long way to go though!

Good afternoon.

It has started to filter in on social media that what was then a tropical wave – and now Hurricane Matthew – could threaten or even hit the Northeast or Mid-Atlantic with great force. That is supported by some – but certainly not all – models, but is more than a week away (possibly farther away) if such does happen. The possibilities are virtually endless as there are multiple systems aloft which will likely interfere with Matthew. In addition, models beyond 7 or 8 days out are notoriously unreliable.

Current situation
Short wave infrared image of Tropical Storm Matthew at 6:15 pm EDT September 28, 2016. Matthew is currently located near the Windward Islands.

Short wave infrared image of Tropical Storm Matthew at 6:15 pm EDT September 28, 2016. Matthew is currently located near the Windward Islands.

Hurricane Matthew is currently tracking westward across the southern Caribbean Sea. It is located just west of the southern Windward Islands and should continue to track west over the next couple days. Despite not having a strong circulation, sustained winds are already near 60 mph. It will likely continue to strengthen and should become a hurricane by tomorrow or Friday. This weekend, it may strengthen into a major hurricane as supported by some models.

The track is quite certain through Saturday. Most likely, Matthew will continue to move westward (in general) to near or just north of Aruba by Friday and south of Haiti by Saturday. From there, the confidence level decreases enormously and a wide range of possibilities are on the table. The next section will explain the possibilities, since what happens in the short term will have a profound impact on what happens down the road.

Middle to long term

Confidence in the forecast decreases markedly beyond the weekend. The first crucial factor will be timing the initial turn. Models are unanimous in a hard turn to the north (similar to 1954’s Hurricane Hazel) somewhere between about 73°W and 82°W. The exact location will also have a great impact on the islands in the Greater Antilles directly affected. This could be an extremely dangerous storm for the western and central Caribbean.

Model differences

The GFS has been consistent on a turn near 73-75°W, taking Matthew over either western Haiti or eastern Cuba. It is also the fastest model with landfall there by Monday. That would then open up a possibility that Matthew avoids the east coast, but a trough may try to approach the east coast from the west by late next week. Depending on the exact timing, Matthew may phase with that trough and sling to the west inland (not unlike Sandy), or remain its own entity and move towards Atlantic Canada (or out to sea), or ride up the coast around the ridge to the east. Either way, it is FAR too soon to speculate on possibilities more than a week away. Ensembles are in good agreement as well, which would normally boost confidence.

The ECMWF – historically the more reliable tropical model – appears confused. The operational run is much slower than the GFS, only clearing Cuba by the middle of next week. That is about two full days slower than the GFS, even though the tracks are similar. Since it wants to rebuild the ridge, Matthew would likely stall out near the Bahamas or east of Florida. It would then be a long waiting game as Matthew waits for the trough (or misses it altogether and waits for the next one) or for the ridge to break down. The ensembles are all over the map, with some suggesting that Matthew ends up in the central or western Gulf of Mexico blocked by a ridge over the southeastern US, into Central America and stalled out like the operational. Given the enormous spread of the ECMWF ensembles, confidence in the medium to long range is low.

Important points to make

1) Confidence beyond this weekend is very low. The turning point will be one of the key factors in determining the long-range outcome and we won’t know until Saturday or Sunday most likely the area(s) – if any – at risk.

2) Even after this weekend, there remain many variables that are uncertain. First, the upper-level low which will be the initial steering mechanism is expected to weaken into next week. Second, it is very uncertain where Matthew will come Monday. Third, the impact that the low will have on the ridge to the east is very uncertain. Finally, the timing of the next trough late next week (and its strength) is subject to considerable change.

3) Much of the energy which will be involved in features beyond 7 days is based on relatively limited data in computer models. The accuracy of models beyond 7 days is often quite low. As we get into the weekend and next week, we will start to get a better sample of the atmosphere through additional observations. It is likely there will be more weather balloon launches next week and more upper air sampling if Matthew is still a significant land threat.

4) Even if we had a decent idea, it is far, far too soon to accurately predict impacts for any one location. I have seen people on social media already ask what Matthew will do to their area (and even beg people). That is not usually known with any confidence until within 3 to 5 days of impact, sometimes even less. Certainly not a question we should be asking at this time with the overall picture still muddy.

5) Regardless, it is likely that Matthew will be a very severe storm for the Caribbean. Models are unanimous in bringing a strong hurricane – most likely Category 3 or 4 – to one or more of Cuba, Jamaica or Hispaniola. Those with interests in those islands should be extra vigilant as such impacts will be 4 to 6 days away from now, but the exact island(s) where landfall may occur is not yet certain.

6) I am not posting any models at this time as they are beyond 7-8 days. Please use extra caution with models beyond that range and don’t let them become viral. Trust local meteorologists on this as they have the most accurate information available. Spreading models in “fantasyland” is irresponsible due to the hype such may create gtcd.

Forecaster Craig Ceecee – @Ceecee_Wx

Leave a Reply