Interesting analysis from Dr. Gray

[BonesXL]

Very Interesting.....

Tropical Weather Discussion


by Gary Gray


--------------------------------------------------------------------------------




...Frances Next Up [8/29 4PM EDT]...






Well, Gaston has made landfall sooner than expected. While there is some bad news to that, in that some folks were hit earlier than expected and may have been caught off guard, it is largely good news. As Gaston was making a last-minute surge, this early landfall kept him from reaching hurricane strength and kept him reasonably in line with the intensity forecast. He also made landfall near the expected landfall point. So, largely, he stayed in line... his timing was the biggest problem. But, I'll be frank, with Frances churning through the Atlantic, I'm grateful for the opportunity to drop Gaston NOW and move on. So, let's do just that...


Thankfully, we've got a number of days to deal with Frances before any possible landfall in the U.S. This is not only good news for folks needing to prepare, but it is good news for forecasters trying to get a handle on the eventual landfall point. Today (Sunday), we are faced with a shift in the models. The shift is very modest in the grand scheme... most models both yesterday and today showed a turn to the north around the western periphery of the upper level ridge somewhere between 70 and 90 degrees west longitude. Yes, that's a wide range, but that was the range amongst ALL models, and most were focused around 80-85 west and, therefore, took Frances into southern or central Florida, with some taking her across Florida and then northward to a second landfall in the northeastern Gulf of Mexico. Today, the shift is really quite small, to about 75-80 degrees west longitude. But that is JUST enough for many of these models to miss landfall in Florida (though, note that MANY still provide significant impact to Florida) and push Frances further up the coast. So, the shift in the solutions is trivial in terms of the overall flow and pattern, but is major in terms of the eventual landfall point and impact.


Before moving on lt me point out that, as I write this, Tropical Depression #8 may be forming. This was the system WSW of Bermuda. It is not expected to have much impact on land... It is moving away from Bermuda and should not reach the U.S. So, I'm not going to give it much/any attention herein. Folks in North Carolina should keep an eye on it, just in case, though... It is expected to reach tropical storm strength and could come quite close to the Outer Banks before turning out to sea. So, it is largely a nuisance. So, on to Frances...


Initial Conditions: Frances is not moving rather steadily off to the WNW. She made this turn yesterday (Saturday) afternoon or early evening. It has continued, with some of the standard wobbling... she has occasionally edged more westward and then earlier today angled around slightly more towards the NW. The long-term motion, though, is pretty steady off to the WNW. She's up nearing 19N (or just a hint south of there) and 55W as I write this mid-afternoon Sunday. At that latitude, Frances is already clear of the Virgin Islands and Puerto Rico... not that they will see no impact - they very well might, but they will see no direct landfall.


In terms of intensity... as I write this recon is enroute. So, I may know more by the time I finish this. However, I would give a rough estimate that Frances is probably around 120kts. Dvorak estimates from 12Z Sunday were down around 100-105kts. But that just happened to coincide with a convective minima. The convection has since deepened and held. Even as I write this, the 18Z fix from SAB is in and it is unchanged. So, maybe a 120kt estimate on my part is high. However, over the last few hours prior to this post Frances' eye has been shrinking. So, I still would not be surprised if she is up near 120kts. She should at least be on the high end of SAB's Dvorak 5.5 estimate (5.5 equates to a rough range of 95-110kts)... so, Frances should at least be in the 105-110kt range... at least.


Model Discussion: As discussed up front, the models have shifted to the east somewhat. I should note that they have not ALL shifted... some of the easternmost models, like the NOGAPS, have held or even shifted a bit to the WEST. So, really, we've seen some decent convergence of the model solutions. I'm still not confident and comfortable with their subtleties and certainly the changes from yesterday to today give some idea what I'm talking about. While I'd hesitate to say that the models are "inconsistent", given that in the overall pattern the shift was modest, the subleties and impact changed quite a bit. And with the eventual/likely landfall still 6-10 days away, there is plenty of time for more such changes in the model solutions. So, while the models are probably generally very much on track, they should be viewed with some skepticism. At any rate, here's how their various solutions add up...


The 12Z Sun GFS has Frances stairstepping a bit as the ridge to the north "pulses". Eventually, by Sat/Sun it has Frances heading up through the eastern tier of the Bahamas. The more NW motion becomes established as the ridge breaks down and a shortwave moves through the central U.S. But, as the shortwave pulls up over the ridge, the ridge never gives way entirely, and Frances maintains significant westward component... coming very close to the central and northern Florda and Georgia coastlines early next week, before finally coming ashore near Savannah or southern SC and continue moving NW from there. The 12Z Global Canadian model is very similar. The primary, but minor difference is that the ridge holds a bit more. It doesn't drive Frances much more westward, but she does drift into the SW Bahamas and then turns up the FL coast RIGHT along the central FL coastline. The 12Z UKMET is nearly identical to the Canadian, but has just enough westward component up the FL coast to kick Frances inland around Vero Beach. The 12Z ECMWF is a step further in that direction, as it has Frances cutting across the Bahamas and then making landfall late Sat or early Sun somewhere between Ft. Lauderdale and Palm Beach, FL. The NOGAPS remains the easternmost model, turning Frances up east of the Bahamas. Still, with ridging building back in to her north by next Saturday, the NOGAPS begins bending her back towards the NW, aiming towards the Carolinas.


As for the tropical models... The 12Z GFDL is well east of most of the "normal" models above. It is in very good agreement with the NOGAPS, pulling Frances up east of the Bahamas. It only goes to five days, so we can't see if it would then bend her back more to the west a bit. However, with the GFDL showing slight deceleration as Frances moves east of the Bahamas, it seems certain that she's missing the trof and will, therefore, likely bend more to the west thereafter... towards the Carolinas, like the NOGAPS. Of the 18Z tropical model suite members, three of the four are in superb agreement. By day five (Friday afternoon) the BAMD, BAMM and LBAR all have Frances in the SE Bahamas with motion generally continuing to the WNW. The A98E is the outlier, turning Frances WSW and putting her near the NE Cuba coast in five days. In terms of intensity, both SHIPS and the GFDL drop Frances down into the 80-90kt range by Fri.


Millennium Weather Forecast: Let me address Frances' intensity forecast first, because I'm simply confused at the model predictions. What could cause Frances to weaken dramatically (30-40kts over the next five days as the GFDL and SHIPS show)? Shear? The complexities of the flow COULD lead to some shear. Likely, however, it would not be severe shear. In fact, with a ridge to the north for much of the next few days, shear should be held in check. Further evidence that shear is lacking is the agreement between the BAMD and BAMM tracks. They use different steering layer flows to guide Frances, yet they both move her to nearly identical points after five days of motion. Cold water? The latest SST analyses show that Frances has, just over the past several hours, moves into warmer water and has no cooler water ahead of her (with the exception of a subtle decrease near 73W). And it wouldn't be due to upwelling either as the models don't stall Frances... some slow her for a while near/in the Bahamas, but never stall her. Dry air? Maybe, but I doubt it. There is a pool of dry air to Frances' west, but it is shrinking and moderating. I will give the models the benefit of the doubt that they're seeing something I'm not. For that reason, I won't intensify her much... except in the very near term, I'll put her to 120kts. But, I also won't decrease her. I just don't see it. So, I'll keep her at 120kts through the forecast period.


Now, as for Frances' track... As I implied earlier, it's too soon to commit. But, generally speaking, I'm leaning towards the more southern solutions in the models. There are several reasons that I suspect that the ridge north of Frances, as she approaches the Bahamas, will be on the strong side. First, models in the medium range have a general tendency to be too weak with ridges. That bias is neither extreme nor a guarantee, but I believe other factors support the notion. Second, Frances is large and powerful. She is not huge, but generally good sized. Large, powerful hurricanes can modify their flow beyond model predictions. How/why? The mass transport in the outflow out of the hurricane and into the surrounding area (esp. into surrounding ridges) is quite extreme. As such, the ridge to the north ends up stronger than anticipated and the storm holds to a slightly more westward track. Finally, hurricanes act much like a cork in a stream. As such, they don't evolve with the flow as an extratropical system does. Rather, they basically move like a cork in the stream. As such, there is sometime some inertia in getting hurricanes to turn. So, what this all means is that I'm anticipating a track most similar to the ECMWF, with Frances making landfall over southern Florida. One concern I have with this forecast is the old saying in model analysis, "the trend is your friend". There is a scientific basis for this saying in that the models use data from previous runs as their background field. As such, models can sometimes be slow to change when there is real data suggesting such a change. Mind you, not ALL the models have trended north (neither the ECMWF nor the NOGAPS have), nor are most of the models showing their northernmost solutions (a few old runs of the GFS kept Frances offshore completely). But, in general, the latest model runs have shown a modest northward shift. If this IS a trend, and if it holds, my forecast may not hold. So, folks all the way up through the Carolinas should pay close attention to Frances. But my thinking right now is for a more southern landfall... down in the general vicinity of southern Florida.


That's about it for now. As a recap... basically, we're expecting Frances to hold at her current intensity and move through the southern Bahamas late this week. Thereafter, I'm anticipating impact over souther Florida, but we certainly need to keep our options open further north. In closing, let me note that as I finish this up, the first recon fix is in. They measured flight level winds at merely 97kts. However, this pass was made through the NW and SW quadrants. We should reserve judgment until recon hits the NE quadrant. However, the recon did get 949mb as the central pressure, in good agreement with TPC/NHC's running estimate. So, likely, Frances is still right at 115kts. We'll need more recon data to verify that she's that strong, but based on the normal pressure fields, 949mb should equate nicely with 115kts. That's further reason to figure that recon simply hasn't measured the highest winds yet..... and we'll know soon enough.