NHC 2006 forecast review

The National Hurricane Center released it’s annual forecast verification report today. The report reviews the accuracy of the NHC’s track and intensity forecasts, as well as those of the forecast models, for both the Atlantic and eastern Pacific basins. Here’s some highlights for the Atlantic basin:

- The track forecasts for the time periods of 12-72 hours set new records for accuracy, breaking records set in 2004. The errors for 24, 48, and 72 hours were 50.8, 97.0, and 148.7 nautical miles, respectively.

- Of the models routinely available to the forecaster’s in time to be considered for their forecast, only the consenus models beat the NHC forecast across multiple time periods.  The Florida State Super-Ensemble was second-best among these models; it was slightly edged out by the GUNA consensus, which is a simple average of the GFDL, UKMET, NOGAPS, and GFS models.

- The report also includes performance of models that are not available to the NHC before forecast time. The model from the European Center for Medium Range Weather Forecasts, commonly referred to as “The European”, displayed superior performance for the 72-120 hour time periods (it’s error at 5 days was a mere 162.7 nautical miles, while the GUNA’s  was 227.1).

- As far as overall accuracy goes, the story was the same as in previous years for the NHC’s intensity forecasts, the average error was in line with that of the past five years.  While there was an overall tendency to under-forecast storms in 2005, the reverse was true in 2006.

- As far as performance relative to their statistical baseline model goes, however, the story was (more) discouraging for the forecasters.  The forecast generated by the simple SHIFOR model was better than the NHC’s from 24 hours onwards. At the extended periods, the NHC forecast was 30-45% worse than that of SHIFOR.

- The forecast models were similarly embarassed. On average, after 24 hours, none of the state of the art models beat out the model that was designed in 1979.

Tropical Storm Florence update 071900L

There’s not much terribly new to say about Florence. She’s continued to move to the west-northwest and is now roughly 1000 miles southeast of Bermuda.

Despite recent difficulties against the usual culprits of shear and dry air the intensity forecast isn’t much different from that of a couple of days ago. Florence is forecast to be a hurricane by Friday evening with further intensification thereafter to near category three status.

The most interesting thing about Florence is her large size. The very broad circulation covers about 12° of longitude; nearly 700 nautical miles. A pain for the forecasters over the past few days has been trying to determine an exact center. Ordinary visual and infrared satellite imagery has not been particularly helpful in this task as it’ll show one or more vortices spining inside the circulation only to die out later. The three agencies tasked with providing position estimates to the National Hurricane Center have tended to pick different vortices to call “the center” and as such, there have been substantial discrepencies in their position estimates, which in turn affects their intensity estimates, (which had gotten as high as 63 mph they’ve since gone down a notch). Tomorrow afternoon the hurricane hunter will be making its first flight  to Florence to give more definitive answers to these questions.

Florence has been moving a little bit slower than forecast. As such the storm should enter the gap in the mid-Atlantic ridge sooner than previously expected. As such, it appears the only land mass seriously under the gun is Bermuda, with the Canadian maritimes possibly being affected later. Owing to the very large size of Florence, it’s about a given that Bermuda will feel some effects, but due to the small size of Bermuda, it remains highly unclear whether Bermuda would get the worst. The National Hurricane Center apparently views the forecast as straight forward; the normally loquacious Stacy Stewart’s 11 AM forecast discussion was a mere two (brief) paragraphs long.

Tropical Storm Florence

The big picture for Tropical Storm Florence remains similar to how I outlined it on Sunday. The storm is currently in less than favorable conditions, but that is expected to change in the storm’s favor in the future. It remains in a position that could threaten the US, but it will be the position of the Bermuda high that determines whether Florence affects land.

On Sunday I stated that guidance was indicating a hurricane in two to four days and opted for the latter part of the spectrum, which implied a hurricane by Thursday afternoon. It now appears that estimate was a bit too soon. One reason being that a shear generating trough to the north of the storm had been forecast by models to simply lift out. Instead the trough fractured, which leaves an upper level low near the path of Florence. That low should impede Florence’s intensification to a hurricane until Saturday, in my opinion. The Hurricane Center’s forecast brings Florence up to hurricane strength by Friday, which is slightly slower than what model guidance suggests.

Another obstacle for Florence was a disturbance to the southwest of it, which before the birth of Florence was being monitored for signs of development. The disturbance did not develop, but it did interefere with the circulation of the developing tropical storm. That obstacle is now out of the way as Florence’s circulation has subsumed the disturbance. The result though, is another hindrance for Florence; the lack of a compact circulation is another reason for intensification to be in check during the next couple of days.

One more potential wrench for Florence’s intensification prospects is a disturbance to her east. Dr Masters devotes a paragraph of his discussion. The interaction he describes is not necessarily a given. Earlier this season, in the eastern Pacific, hurricanes Bud and Carlota developed about 700 nautical miles away from each other without Fujiwhara interaction, and without the lead storm (Bud) being affected. As such, I don’t think Florence will be affected by the disturbance if the current distance between the two (about 800 nautical miles) is not narrowed substantially.

As far as the path of Florence goes, the near to mid term forecast is unchanged, with a west-northwest/northwest course persisting. At the latter end of the forecast period, most models turn Florence to the north in response to the Bermuda High being shifted to the east, thereby opening up an alley for Florence to turn out to sea without affecting the United States (though if such a turn were to occur as quickly as the GFDL model is forecasting, it could affect Bermuda).

As my climatology post below shows, that is the most likely scenario, however the possibility of the Bermuda High being more resurgent than forecast or the development of a blocking high over the Canadian maritimes and thereby forcing Florence into North Carolina or the northeastern states remains plausible. We still have plenty of time to watch the situation unfold before having to be concerned with that prospect.

A little bit of climatology

Not much new to say about Tropical Depression Six; just wanted to share the history of tropical depressions and storms in a similar position to that of 6.

August and September Tropical Depressions and storms that passed within 100 nautical miles of TD6’s 11 PM position.

As you can see, it is difficult for storms that start as far north as 6 did to make it to the east coast; most go up the gap that opens when the Bermuda high is in an easterly position.   There are, though, some notorious storms that have started to go up such a gap only to have the high strengthen or shift to the west, which forces the storm ashore, such as Gloria and the 1938 storm are two such examples with Diane perhaps being the most dramatic example of a high trapping a storm that would have been headed out to sea.

Tropical Depression Six

An area of low in the mid-Atlantic has developed enough for the National Hurricane Center to classify it as Tropical Depression Six. Advisories will begin at 5PM EDT. To give you a feel for where it’s at, here’s a satellite shot of the central Atlantic. The disturbance is a little to below and to the right of center, near 15°N 40°W:

Intensity guidance brings this up to hurricane strength in 2-4 days. I’m inclined to lean with the later end of that spectrum owing to the depression being around a high shear enviroment at the moment. As it moves west however, it should find a more favorable upper-air enviroment and it will certainly find warmer sea surface temperatures.

The depression is in a position that threatens the east coast, though it is far too soon to speculate whether that threat will be realized. Preliminary model guidance takes it on a northwest/west-northwest track through the forecast period. The $64 question is what the strength and positioning of the Bermuda High will be. The long range GFS forecast is for it to be centered well to the east of the United States in a fashion that would allow a storm to remain at sea. As the system is a good 8-10 days away from land we’ll have plenty of time to see how the steering pattern evolves in the days ahead.

The end of Ernesto and other tidbits (updated)

- The National Hurricane Center issued its last advisory on Ernesto this morning. The Hydrometeorological Prediction Center will issue advisories highlighting the rain threat from what’s left of the tropical depression starting at 5 PM.
The Associated Press reports that parts of Virginia have received six inches of rain from Ernesto and there are more than 200,000 people without power.

A flash flood warning covers Hampton Roads and flash flood watches are in effect for the rest of eastern and central Virginia, all of Maryland and New Jersey, almost all of Pennsylvania and southern New York. People in those areas may want to check their local National Weather Service office for further details.

- With Ernesto ashore, there are no active storms in the Atlantic. There are, however, two storms in the eastern Pacific, with Hurricane John threatening the southern half of Baja California. A hurricane hunter is currently in John and just found flight level winds of 102 knots in the northeast quadrant, which adjusts to 100 mph at the surface. That is about 15 mph lower than what the 8 AM PDT advisory rated it at.

The observations put John on equal footing with Henriette of 1995, which hit Cabo San Lucas with 100 mph winds before turning westard to open waters. The NHC forecast puts John on a similar path to that taken by Henriette, albeit closer to the coast after passing Cabo San Lucas.

- The Colorado State University forecast team has issued their updated hurricane season forecast , which includes a verification of their forecast for the month of August.

From the standpoint of named storms, their forecast for August was close. The forecast was four and three formed. Beyond that, however, it was a bust. They forecast three of those to be hurricanes and one of them to be major. Were it not for the hurricane hunter being in Ernesto at the right time, we would have had a hurricane-less August. And Ernesto of course, was not a major hurricane.

As the forecast notes, however, he likely would have been had he gone into the Gulf of Mexico and not interacted with Haiti and Cuba. Overall, dry and stable air were pinned as the culprits for the tropical cyclone activity being lower than forecast.

The forecast also points to El-Niño like conditions and increased hurricane activity in the Pacific as being indicative of reduced activity in the Atlantic. Indeed, from a named storm standpoint, the eastern Pacific has the most named storms at this point in the season that it’s had since 2000. Also, four of those storms were major ones, which puts it on par with 2002, which was an El Niño season. And don’t forget Ioke, which was the first category five hurricane to form in the Central Pacific.

As a result, their forecast totals for the season are adjusted downward again to 13 named storms, 5 hurricanes, and 2 major hurricanes.

A COUPLE OF UPDATES

- The 11 AM PDT advisory for John rates him at 110 mph, a little bit higher than justified by the recon plane, which seems to be a compromise between the recon observations which suggested 100 mph surface winds, and satellite estimates of 115 mph. As mentioned above, a 100 mph storm hit Cabo San Lucas in 1995. However, while the center line of the National Hurricane Center forecast from this morning put CSL in the worst of the storm, Los Cabos radar and satellite imagery indicate that John has more of a northward component of motion, which suggests that area will miss the worst of the storm as it would either enter the Gulf of California or cross the Baja peninsula to the north.

- Just after posting this, a tropical wave about 1200 miles east of the windward Islands was classified as an invest area; an area that will be monitored for development. As one can see in the image below (credit: NOAA’s Satellite Services Division), the wave is just south of dust associated with the Saharan Air Layer, which runs uninterrupted from Africa. As such, any northward movement would severely impair development. Otherwise, it would have good prospects for development.

The value of the upper-air data in the Ernesto forecast

A few days ago I warned to not take much stock in the models until the Gulfstream-IV performed its first mission to collect upper-air observations to be fed to the models. Here’s a follow-up to show why that was the case, with a focus on the four models that make up the GUNA consensus, which was the best performing guidance last year. The members are the GFDL, UKMET, NOGAPS, and AVN models.

Let’s start with the last run of models without the data. As one can see in the graphic there was a good bit of spread in the models overall, with the AVN forecasting a south Florida hit, the GFS and NOGAPS forecasting west Florida, and the GFDL taking the storm out to the panhandle.

The Sunday night run with the upper air data did not feature such a spread. It isn’t until late in the forecast period that there’s an appreciable difference.

On the subsequent Monday morning run, some of the spread opened up again, with the UKMET putting its forecast back to the west some and though you can’t really see it on the graphic, NOGAPS moved Ernesto considerably faster to the north than the others.

Here’s how the 48 hour forecasts for these model runs turned out.

Time	GFDL error	UKMET error	NOGAPS error	GFS error	Avg error
12Z 27 Aug	86 nm	107 nm	68 nm	21 nm	70.5 nm
00Z 28 Aug	56	55	49	28	47
12Z 29 Aug	37	71	115	17	60

The upper-air data provided a 25-33% improvement to the average error of the four models.

As one can see though, the GFS was a significantly better performer than the other three members of the consensus and indeed as early as Saturday morning, its forecast was closer in the long term, though it was viewed as a “dubious solution”.

Tropical Storm Ernesto update 281500L

After the 5 AM advisory was released, the hurricane hunter flew into the northeast quadrant of Ernesto and found that winds were a little bit lower than supposed, because of that his strength was lowered to 40 mph at the 11 AM advisory. The plane stayed in the area making numerous fixes of the center as Ernesto went ashore in Cuba just after 7 AM.

A plane is now flying around the northern coast of Cuba waiting for Ernesto to go feet wet again. (It won’t fly into the center as long as it’s over land). The observations it has collected up to this point show that Ernesto has mantained his strength up to this point.

Without the fixes from the plane, the NHC estimated the 11 AM and 2 PM positions using satellite imagery, which is always a bit tricky when the storm isn’t a full fledged hurricane (when the presence of an eye makes the task simple). That has proven to especially be the case with Ernesto, so one should treat those positions with some uncertainty until Ernesto is over water again.

As always, refer to the National Hurricane Center for the latest on Ernesto.

Tropical Storm Ernesto update 28 Aug 0530L

A hurricane watch is now up for southern Florida from Deerfield Beach on the east coast around to Chokolostkee on the west coast. The Florida Keys are also under a hurricane watch.

While the strength of Ernesto as rated by the National Hurricane Center (50 mph) has not changed it is likely that in actuality it has increased (in the sense that it was probably under 50 mph at 11 PM last night, but there was not a recon plane in the storm to confirm that). Pressure has fallen to 1002 millibars and the reconnaisance plane currently in the storm should be able to squeeze in one last past in the northwest quadrant to see what Ernesto’s strongest winds before he goes ashore in Cuba.

My previous post discusses what the forecast models output last night so I won’t discuss that here, but I will tie in some ideas from that post, so one may want to scroll down and read it before continuing here.

The NHC forecast brings Ernesto into the United States as a category one hurricane. It gives Ernesto a little less time over Cuba than the models were forecasting, perhaps because of yet another reformation of the center that moved Ernesto a little further north than its northwest movement would have taken it. The forecast discussion contains a caveat similar to the one in my previous post and notes that if Ernesto clears Cuba faster than expected, it could become a category two or three hurricane, owing to extremely favorable atmospheric conditions.

Even though the forecast models were fed with the upper air data, which generally improves the quality of the forecast (and certainly helped reduce the spread in the models), there are still things that are beyond the ability of the models to predict (such as the aforementioned center reformation).  Because of that, one could reasonably expect a track error of say 25 miles over a 24 hour period. Normally that would not be a big deal, but owing to the narrowness of Cuba, that is the difference between being over land and being over water. This forecast is an extremely close run thing in the near term.

As always, refer to the National Hurricane Center for the latest and greatest official updates on Ernesto. Residents of south Florida should also check the page of Miami National Weather Service Office for local statements).

Models go with best case scenario for Ernesto

The 00Z runs of the dynamic forecast models as hoped, brought forth near perfect agreement on the US landfall of Ernesto. The forecast features about 12 hours of time over water before going ashore in Cuba, followed by 24 hours of Cuba and then 12-24 hours over water before coming ashore in the southern-most part of mainland Florida, in the Everglades or just east thereof.

It is in line with the best case scenario described in the previous post. The added bonus is his severely weakened state. The reconaissance plane found that pressure has increased to 1008 millibars. That combined with the fact that Ernesto only has another 6 hours or so before entering Cuba strongly suggests that Ernesto would be quite fortunate to come out of Cuba at tropical storm strength. Having only 24 hours of less would provide too little time for for Ernesto to reintensify into anything more significant than a rain event.

The big caveat to this is that Cuba is not much more than 50 miles wide, and is narrower than that at many points. As such, it would take only a small northward bump to put Ernesto on a track that runs along the north coast of Cuba, instead of down the middle. Such a track would maintain provide far more opportunity to be a threat to cause damage in the middle to upper Keys. The position of Ernesto at 8 AM will give a decent clue as to whether such a scenario will play out.