Another Hurricane Coming?

Dean did head straight west, and it hit Mexico just as the last runs of the Global Forecasting System said it would. After that it turned really quiet. But now things are heating up again.

First of all there is Invest 95L. This popped up all at once recently, after a cold front came thorugh last week. I am wondering if it will develop into anything. If it does, it is likely to go to the east, but still I am watching it because it is close to here in Virginia and North Carolina.

Then there is Invest 94L. It is in the southern Atlantic Ocean, approaching the Caribbean. So I tried to run GFS and see what it says about Atlantic storms heading here. Here are the results of GFS runs:

2007 August 27
18Z - makes a beeline for the Outer Banks, like Isabel
2007 August 28
0Z - hits southern Florida, and probably would go into the Gulf, but can't tell as it goes out of range of the model.
6Z - misses the Outer Banks and heads for the Canadian Maritimes, striking there out of range on the 16th of September.
12Z - skirts up the east coast of Florida
18Z - heads straight for the Outer Banks, to strike there out of range on the 14th.
2007 August 29
0Z - Turns north and goes out to sea
6Z - Turns north and goes out to sea; much weaker than in previous runs
12Z - Turns north by end of run, but it seems largely to have disappeared from these runs.

This storm that I am describing above is not Invest 95L but a storm that comes after it. All the runs say 95L will strike the Leeward Islands as a weak storm and dissipate.

So it looks like it is turning quiet now, after some excitement. The other models show a tropical system developing in the Atlantic and hitting the middle of the Leeward Islands. I guess we will wait and see what happens.

Dean to Strike Mainland Mexico

It's been a while since I've posted about Hurricane Dean. But I did hear one interesting prognostication about where Hurricane Dean is going. He's going to hit Texas. Then California. Then Washington and Oregon. Then South Dakota. Then Michigan. Then right up to Washington, DC and into the White House!! YEAH!!!!! Right. Hurricane Dean. Howacane Dean. Howane Dean. Howard Dean.

Actually, it is pretty much following the paths of all those "0 Type" runs of the GFS that I have been talking about in this blog. It went through the Caribbean, making a mess out of Jamaica. Then it hit Yucatan this morning, making a direct hit with 155 mph winds on Chetumal. It will go into the Bay of Campeche and charge into mainland Mexico at about Tampico and be reduced immediately to a 20 mph wimp of a storm, but not before it causes catastrophic damage to some places. The people in Houston and the rest of the country are breathing easier that it did not hit any oil wells or refineries in or near coastal Texas. The Space Shuttle Atlantis came down a day early, but got some fantastic pictures of Dean and landed well, despite the presence of this Category 5 hurricane.

This is what the media have been saying. But they have not been saying something else. Not Channel 12 in Richmond. Not CNN. Not any news site on the Web. But the bloggers (and The Oil Drum)have been saying it all over the place. Hurricane Dean is endangering the world oil supply. It is headed straight for Cantarell oil field, where it will strike with hurricane force. If Dean destroys the oil drilling equipment at Cantarell, Pemex may decide it is not economical to rebuild it. That would cut off 1.7 million barrels a day from the world market, and would probably cause oil to increase in price. But as of yesterday and today, oil has actually been dropping in price. Don't the wizards on Wall Street and other places know about Cantarell?

There is another threat out there, Invest 92L. So far it is just a mass of clouds, and GFS doesn't do anything with it. For a while I thought it was this after storm that I had been tracking and giving a GFS string for. But that storm simply disappeared. I don't think much is going to happen to it. NOAA says conditions are not favorable for development. The Canadian model develops it, sends it to central Florida and then northwestern Florida and Alabama (a Type 4), but that's the only model that shows anything of the storm, except for a shadow on MM5. So right now Dean is the main game. He could have a substantial impact on the price of fuels.

I have also been watching the Arctic ice sheet. The rapid melting of this sheet continues, especially near Alaska and eastern Russia. There has actually been a slight increase in the cover on the other side, towards Europe. But this is the most the ice sheet has even shrunk. It just barely covers half the Arctic ocean. This shows that global warming is definitely happening, and it is probably due to human burning of fossil fuels.

Monster in the Caribbean

Hurricane Dean keeps chugging along, building up 150 mph winds with the help of abnormal seas. It may strike Jamaica as a Category 4, devastating the island, the Yucatan, including Cancún, as a Category 5, and Mexico, south of Brownsville, as a Category 3. Something that may help out with a US landing was pointed out by a CNN meteorologist. Erin dragged up some cold water from the depths, and she also piled rain on the mainland, causing cold water to flow into the Gulf. So the storm might not be as intense as once feared. The September 1 storm is still there, but it seems week.

Add four more zeroes to the GFS string of Dean. Add 9404 to the string for the September 1 storm, to give

5696654899059404

However, this storm is becoming weak in the model runs, and it could vanish altogether.

Dean Headed to South of Brownsville

The recent runs of GFS all make Dean a Type 0 Hurricane. It is headed for Mexico, south of Brownsville and Matamoros. However, some of the other models show it going farther north, such as GFDL and UKMET. GFDL now shows it going right into Houston, which would mean $6/gallon gasoline right then and there. The other models carry it to Mexico, and I suppose it will cause a catastrophe there, but it is apparently not going to affect the oil patch.

How about the after storm (September 1)? It's GFS string for today is 9905, so the entire string is now:

569665489905

It's all over the place. Further, some of the runs show two storms, with the second one going fish in the Atlantic. We will have to wait and see what happens to it.

The storms in Richmond were fantastic. As far as I am concerned, it was the second worst storm we have had. One-third of the Richmond area lost power, and many trees were down. Virginia Power can't restore everyone to power until tomorrow night. This storm was worse than Fran. It was worse than Floyd. It was worse than Ernesto and the storm before that. It even piled more rain than Isabel then, with 5.86 inches (a new record) as opposed to 3.4 inches with Isabel. Isabel was more severe because of its higher winds. It was about the same as Gaston, but this storm caused a lot of power outages, whereas Gaston created a lot of flooding. So not all severe storms are hurricanes.

Stability for Dean; Another one's coming

I missed a day on reporting Dean and other tropical weather. Something decisive happened on 2007 August 16 06Z, or about 6 days before the storm strikes. After the 00Z run showed a Type 2 Rita strike, all of the strikes of this system are in northern Mexico and southernmost Texas, or perhaps farther south. Further, the other models all point to this as well. Further, the storm has transmogrified from Invest 90L to Tropical Depression 4 to Tropical Storm Dean to its present appellation, Hurricane Dean, with 100 mph winds. It is expected to reach Category 4 by early next week, enough to cause a disaster somewhere. I think it may have minimal impact on Gulf oil rigs and refineries, unless it turns unexpectedly north.

Dean's GFS string is now

174979932645100000168651220000000

Why GFS went haywire and reported the 16865122, I have no idea.

But yet there's another one coming. The GFS has now been consistently showing a storm to strike the US or Mexico on or about September 1. The latest run of this has it missing Hatteras and heading to the Canadian maritimes. Past runs of this storm have come perilously close to the Hatteras coast. Since yesterday's 00Z run, the GFS string of this storm, which has no official name since it does not exist yet, is

56966548

To me there are too many 6s in the string, and the average value of the strike is 6. Does this mean we have a hurricane coming up the Carolinas or Virginia on September 1? Perhaps, but look at Dean's string. It went all over the place at first before settling on a Type 0 run. GFS forecasts are apt to go kerphluey until they are about 6 days before the event. Even then, they can goof; Ernesto last year was supposed to strike the Gulf about 4-5 days before landfall, but instead it came up Florida to Virginia.

In the meantime there is some mean weather going on right now in central Virginia. After the temperature hit 100 degrees today, a cold front produced a line of showers pushing through the Richmond area. The storms were severe. A check of area police reports at 8 pm showed trees down and traffic light malfunctions all over the place, and Virginia Power reported 33,000 customers without electricity. A check of Google weather stations revealed winds up to 29 mph, but in places the wind must have been 50 mph or greater to have knocked down all those trees, one of which piled onto two lanes of I-95 without causing a single red warning blip on the web edition of Virginia 511. Further, with trees down all over the place and power outages galore, WWBT-TV's web site had the gall to say that the top story right now is President Bush's daughters upcoming wedding. Another case of celebrity culture, this time under the purview of one of the worst Presidents in US history, and a case of gross misplacement of priorities.

Tropical Storm Dean

Today was a big one in the tropical weather world. The storm off the Africa coast that had been lighting up the GFS model the past week is now officially Tropical Storm Dean, beating Invest 91L in the Caribbean for that honor. That storm may become Tropical Storm Erin. Hurricane Flossie is coming closer to Hawaii and is stronger than predicted, causing the governor of Hawaii to declare a state of emergency. Hawaii has really had its problems. They had an R5.3 earthquake this morning, and there is a wildfire going on. The latter is something that Flossie can do something about.

So what do the models say? Here are today's GFS runs:

0Z - Skirts by the Carolina coasts, then suddenly turns left into them, charging up to Charlotte and western Virginia, where SUUSI is held every July.
6Z - Clobbers Miami head on then travels straight up the Florida Turnpike into Georgia and Tennessee.
12Z - Stays to the south again and hits Houston, causing a problem.
18Z - Stays to the south, this time hitting the LA/TX border, like Rita in 2005.

If one takes the type of each run (as defined in yesterday's Beyond the Wind) in a string, the result is 6512. So the overall string since August 8 is:

1749799326451000001686512

To me this is as volatile as the stock market has been lately. Now that the Gulf is threatened, oil prices are beginning to go up.

What do the other models say? NOGAPS is still struggling to make sense out of the storm. GFDL inexplicitly dropped the storm this morning, but now they show it again. The Canadians make it go fish, and UKMET takes it into the Caribbean. So the results are all over the place. The official forecast takes it to five days from now, when it strikes the middle of the Leeward islands and continues just south of the Caribbean islands. Where it goes from there is anybody's guess right now.

The Storm2k and other weather boards are heating up now. For some reason, someone insists on reporting on each GFS frame as it comes into view on the Web. If Dean strikes 7 days from now, that is 4x7x24 = 1,344 posts! They have disabled Search on that site, reducing its usefulness considerably. There are warnings about nonsense, chatty, or off-topic posts, because the managers of the forum fear that an onslaught of activity could slow it down to the point of inoperativeness. The meteorologists and media are picking the storm up, now, even calling it Dean before it became Dean. I hope they don't hype this storm.

Big Change

One of the things I don't like about computers, especially Microsoft products, is that they go from one side or extreme to the other. I call this "whomping". Excel is the biggest whomper. When you try to select a few hundred rows, it insists on going on and on and gives you a hundred or two rows more than you want. That's whomping.

Well, computer programs are not the only things that whomp. Invest 90L really whomped today. It became Tropical Depression 4, but it whomped from hitting Mexico and southern Texas to hitting Nova Scotia instead. Here are the GFS runs for today:

0Z - Well to the south, striking Mexico and southern Texas.
6Z - Hits Houston, causing high gasoline prices and shortages.
12Z - Goes by the Outer Banks, then hooks to the left like a bowling ball and hits the Tidewater area head on, then curves to the right and scores a strike by knocking down Maryland, Delaware, New Jersey, New York City, and southern New England including Boston. Hope this doesn't happen.
18Z - Completely misses the US, skirting by the Outer Banks by 200 miles. It then hits Nova Scotia

As you can see, TD4 whomped all the way from Mexico to Nova Scotia. What it's going to do now is anyone's guess.

I have come up with a scale that allows one to graph these results without using complicated map algorithms. I assign tropical system strikes with a scale from 0-9 as follows:

0 - Mexico and southern Texas (TMEX)
1 - Houston area (HOGA)
2 - Louisiana/Texas border (LATX)
3 - New Orleans and Mississippi (NOLA)
4 - Pensacola and Mobile (PEMO)
5 - Up the Florida peninsula (FLOP)
6 - Outer Banks (OUBA)
7 - New England (NYNE)
8 - Canadian Maritimes (MARI)
9 - Out to sea without hitting the North American continent (FISH)

This storm then went from 0 to 1 to 6 to 8 today. The past few days can be indexed thus:

08:---1
09:7497
10: 9932
11: 6451
12: 0000
13: 0168

This shows some wide swings. Other models' results can be indexed by this code, if they go far enough to show the result can be identified by the code.

Of course I don't like its wanting to go to the Outer Banks and Virginia. Some of the other models have hinted at this. The Canadian model wants to recurve this storm to the north at the Leeward Islands. The European model (ECMWF) wants to hit Florida. So we will have to wait and see what happens.

In the meantime we have some other threats. Invest 91L is between the Gulf of Mexico and the Caribbean; it may become Dean if it develops quickly enough. Hurricane Flossie is stronger and closer to the Hawaiian Islands than they had figured on before; the governor of that state has declared a state of emergency. And various storms threaten to come the US after TD4. To be continued tomorrow…

Invest 90L All Over the Place

Where is Invest 90L going to go? Is it going to turn northwards and become a fishspinner, going harmlessly out to sea? Maybe, because a strong cold front is due to exit the US at that time, tracking the storm to the north. Will it hit the Carolinas or Florida? Texas or Mexico? Or will it simply fizzle out and go poof? Answer: Any of the above. Or all of the above.

The GFS runs today say the following:

0Z - Mexico, near the Texas border.
6Z - Mexico, near the Texas border.
12Z - Mexico, near the Texas border.
18Z - Mexico, near the Texas border.

So it looks like Invest 90L is going to go to Mexico. But one person on Storm2k says "there is no way this thing is going to Mexico." He has support from some of the other models:

European (ECMWF): curve to the north and head towards central Florida.
Cold front coming off coast: Go fish.
NOGAPS - fizzle out and go poof.

At the beginning of the day I thought Invest 90L was going to go to Mexico. Now I am not sure. Right now it is still Invest 90L (which deprecates the fish theory), a few hundred miles west of the African coast. It has maximum winds of 30 mph. What will it do? The next few days may tell the story.

Invest 90L is for Real

Well, the moment of truth has arrived. The storm coming off the African coast passed it. A weather discussion said that we could have Hurricane Dean in a week. Jeff Masters' Weather Underground blog says it could become a threat by midweek. And this afternoon, NOAA gave it the name Invest 90L, meaning that it is something worth investigating.

Here are todays' GFS run results:

0Z - Looks like it's going to just miss the Outer Banks, then it turns westwards and heads towards central Virginia.
6Z - Hits Florida peninsula in the middle, continues in Gulf and hits Alabama and Florida panhandle.
12Z - Hits Miami, then goes right up the Florida peninsula into southern Georgia and Alabama. Another hurricane behind it seems to be headed for the Carolinas.
18Z - If this really happens, gasoline will be $12 a gallon and massive shortages may develop. A direct hit on the refineries and oil rigs near Houston.

These runs are consistent in that they hit somewhere along the US coastline. A fish (going out to sea) seems to be ruled out by the powerful high pressure ridge that is sending a tongue to the southeastern US coast. The other models are corroborating. The Canadian model has it, and in addition forecasts a sudden appearance of a tropical storm near Florida in the Gulf of Mexico soon. So do ECMWF, NOGAPS, and UKMET. MM5 does not show it, but shows instead a storm soon in the Gulf of Mexico.

Looks like we are going to have a storm here, probably Hurricane Dean. The storm will strike on or about August 22. NASA needs to keep watch of this system. That date is when Endeavour will come home if the mission is extended. Other interests on the Atlantic and Gulf coast should also monitor this situation.

Moment of Truth

The moment of truth is nearing for this storm. There is a lot of information pouring in right now that suggests that some sort of storm, probably a hurricane, will approach North America somewhere around 2007 August 23. Here are the latest GFS runs, all from 2007 August 10:

0Z - Misses Atlantic coast by about 500 miles and goes out to sea.
6Z - Misses Atlantic coast by about 200 miles and goes out to sea
12Z - Goes just south of the tip of Florida, into the Gulf of Mexico, and hits the Mississippi coast; similar to Katrina.
18Z - Hits Miami; then goes on to hit the Texas/Lousiana border, starting out like Georges but winding up more like Rita.

So now a wide variety of places have been hit by these far off GFS "projecticanes": New York City, central Virginia, the Carolinas, southern Florida, southern Mexico, Mississippi, New Orleans, and Beaumont, Texas area. The time is still about 348 hours away - or about two weeks. At this far date, one can't predict where it will hit, but it seems likely that somewhere on the North American coast will be hit by this storm. This is especially true because other models are developing the storm now, including the Canadian Meteorological Centre model, NOGAPS, UK Meteorologists' model, and the European Model. These models don't go out as far in time as the GFS, but they indicate some sort of storm will develop.

When should we see this storm on an actual weather map? About now is what the models say. There is a storm in western Africa, in Senegal and Gambia, about ready to leave the African coast. Some people are saying that when the storm leaves Africa, winds blowing sand from the Sahara Desert will kill the storm, so it may not even develop. But if it does develop, a high pressure ridge to the north of the storm (often called the Bermuda high) will guide the storm to wherever it is going to go, probably somewhere on the North American continent.

The moment of truth is nearing for this storm. Will it form off Africa, as the models say it will? Or will it fizzle out? The next few days will tell the answer. Interests on the Eastern and Gulf coasts should keep an eye on this developing situation. Since it does not really exist as a tropical storm or even as a storm on the ocean, the local meteorologists and NOAA are currently not saying anything about it.

Hot Times and Still a Possible Hurricane

It certainly has been hot here recently. Last August, the temperature hit 102 degrees F, tying for the warmest since I moved to Richmond in 1978. This year, on 2007 August 8, the temperature here hit 104 degrees F (40 degrees C), making it the warmest I have ever seen here. The next night it went down only to 80 degrees. Up to now, the temperature has gone into the single digits or teens, or maybe 30s or 50s or at least the 70s. Last night was the first night that it did not even get down into the 70s. Is this global warming? Maybe, but one can't attach "global warming" to any one particular weather phenomenon.

What I do see later on is another hurricane. Not Wow or Frontdragger. Those storms never materialized. We now have two new ones in the far distance:

GFS
2007 August 8
18Z Hurricane hits Carolinas and Virginia 23rd.
2007 August 9
00Z Hurricane brushes by Carolinas 23rd, then another one brushes by the Carolinas on the 25th.
06Z Hurricane hits Florida's midsection 22nd; sidekick vanishes.
12Z Hurricane goes fish in Atlantic 200 miles offshore; sidekick vanishes
18Z Hurricane hits New York City and Long Island 25th; sidekick hits Carolinas 27th, off model range.

It varies from run to run. An earlier run had the first hurricane going so far south that it hit the Yucatan and Central America. There is a lot of variation because it is so far out. My rule for deciding whether there is anything to these GFS hurricanes two weeks in the future is that if the storm currently exists (like Isabel on 2003 September 10) , then there is something to it. If the storm currently does not exist, it may never materialize.

Currently this storm does not exist. It is supposed to form when a storm comes off the Africa coast on August 10. If and when it does, then we may have to consider this storm to be a real possibility, but it is still a crapshoot where it's going to hit, or if it is going to stay at sea ("go fish"). However, some people at Storm2k have been saying that SAL (Saharan air layer; i.e., sand coming off the Africa coast) is too strong for a storm to develop. After it leaves Africa, it goes poof.

But I will continue to watch these two storms, especially since one is predicted to strike on my anniversary.