Upcoming week - September 20-26

[Andrew92]

Evaluating last week:

This week was certainly a crazy one. Two category 4 hurricanes at once for the first time since 1926! There was also a landfalling hurricane in Mexico, and proof of just why I’m really glad I put up my own disclaimers. We’ll take a look at how I did with each storm, then figure out a decent grade for this week though.

Starting with Igor, this was a very good forecast, especially in terms of track and its timing. Even the intensity forecast was very close! For the track, I predicted a west-northwestward movement until Wednesday, with a gradual northwestward turn commencing that day, a comfortable distance from the Lesser Antilles. That’s pretty much what happened those first three days. I predicted that northwestward movement to continue all the way through Sunday, when it would start re-curving near Bermuda. That also was executed about as well as possible! So the track and timing certainly gets an A. Now for intensity, I predicted no more than 155-mph winds, and operationally that did happen. However, I predicted this on Monday, and it happened Tuesday night. But I did correctly predict an eyewall replacement cycle would happen late Monday and early Tuesday, weakening the storm to a borderline C3/C4 hurricane. While the strengthening on Tuesday was not expected, it did weaken again overnight into Wednesday before strengthening again, so Tuesday was at least acceptable in this area. I predicted on this intensification trend that maximum winds would reach 145-150 on Wednesday and Thursday, and they reached 145 on Thursday early morning (shortly after midnight). I then predicted a weakening trend to a C3 on Friday, which happened. Then, I started falling off a bit on the intensity portion as I predicted the storm to weaken to a C2 on Saturday, but it happened Friday night. That said, the storm was a C2 all day on Saturday, so that was still a very good day. But I did not foresee the weakening to barely a hurricane today, predicting still a C2 storm as it passed by Bermuda. Lastly, I did predict Igor would pass west of Bermuda and bring tropical storm winds, but never mentioned hurricane winds. As of the last possible advisory for the week, 74-mph sustained winds were reported in Bermuda, so I just barely missed this. Still, all things considered, the intensity was pretty good until late, and even then wasn’t terrible. I give myself an A- for intensity, and a solid A for Igor this previous week, when also considering this was tricky with Bermuda thrown in.

Looking at Julia, there is no way even with 20/20 hindsight that I would have predicted a C4 hurricane. That said, the first two days were good, with maximum winds indeed reaching 50 mph as it passed the Cape Verdes, and then becoming a hurricane on Tuesday while heading west-northwest. The northwestward turn was also correctly anticipated, and I also predicted that heading would last until Saturday. That day, I predicted a northward turn well east of Bermuda, followed by a north-northeast turn on Sunday. It turned a little sharper than I thought on Sunday, but this is a very slight error due to its location over open waters, well east of Bermuda. So the track gets an A, but the intensity is something else. I predicted Julia would be a minimal hurricane for about 24 hours. Well, it became a C4 hurricane within that time. That said, a weakening trend did begin on Wednesday, but no way by then would it have weakened to a tropical storm. That happened on Friday night, instead. During the time Julia was to move northwest and even north, I predicted a steady intensity of about 60-70 mph. It did fall to about 50 mph actually, but when you look at the FTP file, it did fluctuate back to 65 mph on Saturday night – so the idea was at least executed, just a little outside the window. The intensity was sure not good on Wednesday through most of Friday, but very good at all other times. I give myself a C+ for Julia’s intensity, and a B overall for this storm.

Igor and Julia were pretty good forecasts overall, even during this tough week. The same cannot be said, however, for Invest 92L/Karl. I predicted no development in the Caribbean, and that it would happen after emerging in the Bay of Campeche on Thursday. Well, development did happen in the Caribbean. Karl reached the Bay of Campeche on Thursday indeed, but then I predicted a landfall near Tampico as a 65-70 mph tropical storm. It ended up making landfall near Veracruz instead, quite a bit further south, and as a major hurricane (though I suspect it will be lowered to a C2 when the official report comes out). Even the timing after emerging into the Bay of Campeche was off by about 24 hours. I predicted the landfall for Saturday, and dissipation on Sunday. The landfall came on Friday, and dissipation Saturday instead. Besides the timing of reaching the Bay of Campeche, the only other aspect I nailed with Karl was time over water, about 36 hours. Yes, this was a difficult forecast. OK, the general westward track was well-executed. I also got timing of reaching the Bay of Campeche and how much over water right. That’s about all I can say for this forecast, however. I did put up disclaimers for this prediction to be prepared for those in Mexico, but I still give myself a D overall for Karl – only proving the point of the disclaimers a little more.

Lastly, I didn’t actually say it by accident, but I did mean to predict no other developments. Invest 94L is on the verge of doing so…but hasn’t done it yet! I may be being a little generous to myself for not actively saying it, but I will assume it was implied and give myself an A here – with only little weight though.

So two good forecasts, one bad one, and no other developments. The bad forecast was probably the most important one, unfortunately. But with another important forecast, Igor, that was executed almost to perfection, and a decent Julia forecast all things considered, this week was still a decent one, but not great. It is also one I can take some things and learn from for the rest of this season, and you will probably start hearing stuff like the “2010 factor” in light of especially Julia. I give myself a B- for this week.

How will this week go for me? Let’s take a look…

Current situation and models

Two storms are heading out to sea, Igor and Julia. Both are on the way down, though Igor could stir up some rough conditions in especially Newfoundland, but rough seas could happen throughout the East Coast and Maritime Provinces from a storm this size. Julia is expected to be either absorbed or sheared apart by Igor – either way, it will dissipate in a couple days. Those two forecasts will be, therefore, very easy to predict, and I will not even bother with recent history for these two.

But what else is there? This is 2010, and as of late, the storms have been popping like popcorn (OK, bad joke, I know). The trend isn’t going away either, with Invest 94L threatening to develop at any time into a tropical depression. Most models think this will become Lisa soon too. Fortunately, almost all models predict a re-curve; and the ones that don’t do not appear to be analyzing the storm that well, or aren’t very good at predicting tracks in general. Intensity is quite another matter when factoring in that this is, after all, 2010. It should go over waters churned up by at least Julia, so this storm probably won’t be as strong as either Igor or Julia, but I’m ruling nothing out for now.

Oh my, did I go on and on about a lock for a fish cyclone? What does that mean when I bring up what is a hot topic, a potential Caribbean development this week? The Canadian and NOGAPS models have been sniffing one area for development early this week, but no other model is buying it. The Euro and GFS do, however, think something will develop soon in this area, potentially as soon as the weekend. That something appears to be Pouch PGI46L, currently near the Windward Islands. There are diverging track scenarios right now, so all bets are off there. The Euro takes it to the Bay of Campeche, the Canadian into the northern Gulf, and the GFS into southern Florida. I will take a brief look at this below, but this appears to be a heavier topic for next week.

Recent history:

As promised, I won’t bother you about recent history for Igor and Julia; we all know what they’re going to do. Let’s first jump to Invest 94L, and look at some recent storms that have developed off the coast of Africa during this upcoming week. How about these storms:

Noel in 1995
Isidore in 1996
Ivan in 1998
Jeanne in 1998
Isaac in 2000
Joyce in 2000
Kate in 2003
Lisa in 2004
Karen in 2007

Noticing something? Out of these nine storms, only two did not re-curve far from land. The two that didn’t, Joyce and Karen, also dissipated just east of the Lesser Antilles. Also, all of these storms became hurricanes, though not always during this upcoming week and most were C1’s. However, Isidore and Kate did prove that a C3 can develop this late in the season in this part of the world. That said, Invest 94L will be coming off of the heels of Igor and Julia, two very powerful storms. Joyce and Lisa (ironic) both fit that mold the closest, and both were lucky to be hurricanes (Joyce was right behind Isaac, and Lisa right behind Karl). Also, Lisa in 2004 was the one of these two to re-curve. Hmmm…is this a Lisa thing?

Now to the Caribbean. The proposed development is likely to happen in the western part per the models, so what has happened there for development since 1995 this upcoming week? Nothing, absolutely nothing. OK, that’s a little harsh, so let’s look at dates closer to this weekend. Now, we find that Opal’s incipient depression in 1995 formed near the Yucatan on September 27, and Keith’s from 2000 formed on September 28. Opal’s ran into land shortly after developing, while Keith developed into a powerful storm a few days later. As for tracks, Opal re-curved northeastward, while Keith continued west into Mexico. A couple of potential scenarios to work with, and I know it’s a popular topic, along with intensity. But assuming this happens during the weekend, I’ll talk more about the future track next week. We should know a little more about the steering currents by then, too.

So what does this all tell us?

Invest 94L is likely to become Lisa very soon, when adding the 2010 factor. It also appears it is a re-curver, but per recent history and with this being 2010, will probably become a hurricane for at least a brief period.

It is 2010 and Karl proved it can happen, but I’m not buying the Caribbean development prior to the weekend. There is too much model support to go against it later, but the disturbance that eventually became Karl also took its time, and I expect the same with PGI46L.

Back to looking ahead

First, let’s quickly do Igor. I predict Igor will weaken to a tropical storm on Monday and become extratropical on Tuesday, while passing just south of Newfoundland; close enough to produce very rough gale-force winds in that province. Confidence is 90%.

And Julia: I predict Julia will weaken to a tropical depression on Monday morning, and a remnant low later that night, while continuing on its track east-northeast to northeast. Confidence is near 100%.

OK, that was easy. Now, let’s take a look at Invest 94L. The track forecast is very cut-and-dry. This is a quick re-curving storm. It’s the intensity that will be a little tricky, as one might think coming off the heels of Igor and Julia, that this would remain no more than a minimal tropical storm. Most models do more with 94L than that, and recent history gives this a decent chance at becoming a minimal hurricane, even if just briefly. I predict Invest 94L to become a tropical depression late Monday morning, and Tropical Storm Lisa that afternoon or early evening. After initially moving west-northwest, a northwestward turn will begin early Tuesday morning. Lisa will also slowly intensify and become a minimal hurricane on Wednesday, with maximum sustained winds of 75-80 mph. Also that day, a northward turn will commence between 35 and 40 degrees longitude. Lisa will weaken to a tropical storm late on Thursday near latitude 25, and speed up from there to the north and north-northeast. Weakening will be slow, and Lisa may hold on to maximum sustained winds of 50-60 mph for most of Friday and early Saturday morning. But Lisa will weaken to a tropical depression on Sunday near the Azores, and dissipate before the week is up. Confidence is 75%.

Now let’s assess the Caribbean. While the Canadian and NOGAPS have supported development by Wednesday near the coast of Central America, no other model predicts that. The NOGAPS has also been notorious this year for developments there this year. By the weekend, that’s a different story. History suggests the weekend is when this area becomes more favorable, and Karl has proven that it already is. Most reliable models also support something developing around that time, so it’s probably safest to predict something by week’s end. The timing will be everything, which makes it a little tricky, but here’s my best shot. I predict no development in the Caribbean until Sunday, and it will take place that day between Jamaica and Honduras. A tropical depression will develop on Sunday during the late morning or afternoon hours, and become Tropical Storm Matthew at the 11 PM advisory while heading west-northwest towards the Yucatan Peninsula. Maximum sustained winds at that time will be 40-45 mph, but it will look impressive and very well-organized. Confidence is 80%.

Elsewhere, I predict no tropical cyclone development this week. Confidence is 90%.

-Andrew92

[jinftl]

Any early thoughts on what October will bring....any storms like Mitch or Wilma in the cards?

[Andrew92]

I do think there will be a storm like Mitch or Wilma sometime in October. It could even end up being PGI46L. The models are just too consistent in shifting the pattern, too many systems are popping in favorable areas, and we have to remember that 1998 (Mitch) and 2005 (Wilma) are well-established analogs for this season.

I also wouldn't rule out one more hit for mainland Mexico before this year is up. Again, could be PGI46L, could be something else. Just too soon to tell for sure.

-Andrew92

[JonathanBelles]

I've gotten to look forward to when these statements come out. lol. I don't agree with everything all the time, but it's interesting to see another persons thoughts at that time. You are also quite accurate most of the time. Keep up the great work!

[Macrocane]

Another interesting discussion, thanks Andrew92. By the way I do think Karl was a major hurricane it looked too good to be a cat 2, maybe the peak intensity will be lowered to 100 kt but no less than that.

[Andrew92]

Well with Lisa and Matthew gone, and nothing else active or imminent, might as well evaluate this week and get it over with.

In short, not good. The one major area where I was remotely OK was with Lisa, and even that was off. The development was well-anticipated, but timing was off by about 12 hours. OK, that’s not exactly earth-shattering, but the track sure was. I predicted a northwesterly track, which was a far cry from the easterly track that did happen. I also predicted a re-curve between 35 and 40 degrees, and it didn’t even make it that far west! As for intensity, I was correct in predicting a hurricane of no higher intensity of 75-80 mph, and 80 mph was reached late Friday. However, I predicted that would happen on Wednesday instead, with a gradual weakening from there on. The northward motion was correctly anticipated late in the week, but in the wrong location for obvious reasons, and Lisa did speed up as it weakened from hurricane intensity. I predicted that would happen at about 25 degrees, but it came a bit close to 20 instead. Lisa did then continue weakening, and has de-generated into a remnant low, which was anticipated. Basically, there were snippets of this forecast that were good, such as the maximum overall intensity, the general northward track, and the spot-on timing of dissipation for Sunday. But the events leading up to a lot of this were also way wrong. That said, Lisa was a fish, which was the general prediction, and not terribly important to nail all these aspects right on the nose. Therefore, I give myself a B- for this part of the forecast. OK, not quite as bad as I thought there looking back, but still nothing to brag about too much.

I was also correct in the timing of Julia’s weakening to a remnant low, so that gets an A. Also getting an A is the correct prediction of no developments besides that of Lisa in the eastern Atlantic and Matthew in the Caribbean.

However, the last 24 hours of the Igor forecast were even a train wreck! I predicted Igor would weaken to a tropical storm on Monday, pass just south of Newfoundland but bring tropical storm force winds there, and then become extratropica on Tuesday. Igor actually held its hurricane intensity and made landfall in Newfoundland instead on Monday. That said, it did still become extratropical on Tuesday, but not before I was left with a blown forecast. The weight isn’t too significant, but the grade here is still a D-.

And then there’s Matthew. Well, I guess I was correct in predicting something in the Caribbean, about between Jamaica and Honduras. The timing was a sordid story, however. I predicted the development for Sunday, and it came on Thursday instead. By missing the timing that much, the forecast was guaranteed to be blown. There was no way from there that I would get the landfall in Honduras/Nicaragua for Friday with 60-mph maximum sustained winds and then dissipating over Central America. I have no choice but to give myself an F for this part of the forecast.

The not-as-important forecasts, overall, were pretty good. The more crucial Igor and Matthew forecasts, however, were horrific to say the least. The bad forecasts average to an F and the good forecasts average to probably an A-. With more weight to the bad forecasts because they were overall unfortunately more important, I have to give myself a D- for the week overall.


This upcoming week coming up in just a couple!

-Andrew92