As of 11:00 PM, the National Hurricane Center has bumped Dean’s winds up to 50 mph — and the discussion says this estimate is “possibly conservative.” It also notes that the shear around Dean is “beginning to decrease,” so I think it’s fair to say that, after struggling to hold itself together this afternoon, Dean has now begun its intensification phase. The official forecast brings it to Category 3 status in five days — at which time the forecast track has it just offshore of Hispaniola. But of course, that’s subject to change. The computer models continue to be all over the map, as would be expected for a storm this far out at sea. The intensity models also disagree about just how fast it will intensify. The best advice is to continue to stay tuned.

Meanwhile, also at 11:00 PM, the NHC has designated Tropical Depression Five in the Gulf of Mexico. It is expected to become Tropical Storm Erin tomorrow, then make landfall along the Texas coast on Thursday. Tropical Storm Watches have been issued from Freeport, Texas south to Rio San Fernando, Mexico. Here’s a look at the track models and intensity models for it.

Last but not least, Category 2 Hurricane Flossie is passing due south of Hawaii as we speak:

I would be remiss if I didn’t point out that, four years ago today, the lights went out in New York City, and across much of the eastern U.S. and Canada. And BrendanLoy.com was there. :)

Also, 27 years ago today, Sean Vivier was born. Happy Birthday, Sean!

To tie it all together, here’s a column Sean wrote about the Great Birthday Blackout of 2003.

Oh, I shot my husband in the back with a shotgun as he slept. I took off in the family minivan to the Gulf Coast. I went to trial for killing him, and got convicted a few months ago.

I got out of jail today.

Three suicide bombings in a northern Iraqi town left at least 100 people dead and 150 wounded, police officials said.

Visit CNN for the latest.

The GFS computer model, one of the most commonly looked-at global models, can’t decide where it wants Tropical Storm Dean to go. That’s hardly unusual, given how far away it is; we’re talking about projected landfalls that are almost two weeks away. Besides, as Alan Sullivan points out, Dean is struggling so mightily against wind shear right now, the computers “can hardly even see such a weak system.” Even so, it’s entertaining to watch the GFS’s predicted landfall points bounce around the Eastern Seaboard and Gulf Coast. On his “Model Hype” blog, AccuWeather’s Jesse Farrell has posted an animated GIF showing the last two days’ GFS landfall predictions, which “started in Texas, moved up to the Canadian Maritimes, then back to Texas.” See for yourself:

And that, in a nutshell, is why you shouldn’t put any stock in individual computer model forecasts at this early date. Heh.

In my post below debunking the false notion that meteorologists have developed a habit in recent years of repeatedly “crying wolf” about the number of hurricanes that will form each season, several commenters raised an argument that I’ve heard several times before, and which, like the “cry wolf” perception, deserves some additional scrutiny. Their question is essentially this: If scientists can’t even reliably predict how many tropical storms will form each season (for instance, Dr. Gray’s forecasts were off by between 1 and 3 storms every year from 2000-2004, by 13 storms in 2005, and by 7 storms in 2006), how can we possibly trust their long-term predictions about climate change?

This question focuses on one obvious difference between seasonal hurricane projections and global-warming forecasts (namely, the much longer-term nature of the latter) but ignores other crucial differences — not least the difference between “climate” and “weather,” a distinction that both sides have a bad habit of conveniently ignoring when it suits their purposes. Most importantly, the question assumes that forecasts will always become less reliable as the forecasted time period becomes more remote, which while perhaps intuitively sensible, is a vast oversimplification.

You have to look at what is being forecasted, not just when the forecast is for. How precise, specific, and subject to random, unpredictable variations is the phenomenon at the heart of the forecast? For instance, a prediction for whether it will rain at some point tomorrow is probably going to be more accurate than a prediction for precisely what the rain rate will be at 3:47 PM today, even though the latter forecast deals with a more imminent event. Why? Because predicting whether something will happen during the course of 24 hours is easier than predicting exactly what will be happening at a particular moment in time. The latter is largely determined not just by big-picture meteorological phenomena (cold fronts, low pressure systems, etc.) but also by near-random localized events and tiny variations in timing, which are unpredictable and insignificant in the grand scheme of things, but can make all the difference when it comes to what the weather will be in a particular place at a particular time.

Global-warming forecasts are largely focused on determining the average temperature of the Earth at various points in the future. The average temperature of the Earth is about the broadest, most nonspecific measuring stick possible in meteorology and climatology. It isn’t subject to the sort of random variables that arise when predicting individual weather events, because the random events will generally even themselves out over the broad swaths of time we’re talking about. Thus, the computer models have a better chance of getting it right, because they only need to look at the big-picture factors. They don’t have to “sweat the small stuff,” if you will.

(more…)

It’s official. Tropical Depression 4 has become Tropical Storm Dean as of 11:00 AM EDT.

Hopefully “his” comment yesterday won’t come true: “I’m gonna go to Puerto Rico, and Hispaniola, and the Bahamas! And then I’m gonna go to Florida, and Alabama, and Mississippi, and Louisiana! And then my remnants are gonna recurve inland and rain on the White House! YARRRRH!!!”

Anyway, here’s what the discussion says, with a couple of key points boldfaced by yours truly:

SATELLITE IMAGES…MICROWAVE DATA AND QUIKSCAT MEASUREMENTS INDICATE THAT THE DEPRESSION HAS REACHED TROPICAL STORM STATUS AND ALSO THAT THE CENTER IS LOCATED FARTHER TO THE SOUTH AND WEST THAN PREVIOUSLY INDICATED. DEAN HAS DEVELOPED A SMALL CENTRAL DENSE OVERCAST OVER THE CENTER BUT MOST OF THE CONVECTION IS STILL LOCATED TO THE WEST AND SOUTH OF THE CENTER. THE OUTFLOW IS CONFINED TO THE WEST AND SOUTHWEST SINCE THE CYCLONE IS STILL EMBEDDED WITHIN EASTERLY SHEAR. THE SHEAR IS FORECAST TO DECREASE AND MOST OF THE INTENSITY GUIDANCE GRADUALLY STRENGTHENS DEAN. IN FACT…DYNAMICAL GUIDANCE SUGGESTS THAT DEAN COULD BE NEAR THE LESSER ANTILLES AS A HURRICANE WITH 100 TO 110 KNOTS IN FIVE DAYS. [That translates to 115 to 125 mph — a Category 3 major hurricane. -ed.]

DEAN APPEARS TO BE MOVING WESTWARD OR 265 DEGREES AT 20 KNOTS AROUND THE PERIPHERY OF A STRONG SUBTROPICAL RIDGE WHICH IS EXPANDING WESTWARD. IN THE SHORT TERM…2 TO 3 DAYS…DEAN SHOULD CONTINUE TO MOVE ON A GENERAL WESTWARD TRACK WITH A GRADUAL DECREASE IN FORWARD SPEED. THEREAFTER…THE FORECAST BECOMES HIGHLY UNCERTAIN SINCE SOME RELIABLE MODELS MAINTAIN A STRONG RIDGE TO THE NORTH AND KEEP THE CYCLONE MOVING WESTWARD ACROSS THE CARIBBEAN AND OTHERS SHOW A WEAKER RIDGE RESULTING IN A GRADUAL TURN TO THE WEST-NORTHWEST. THE OFFICIAL FORECAST FOLLOWS THE CONSENSUS AND IS A LITTLE BIT TO THE SOUTH OF THE PREVIOUS ONE…DUE TO THE REPOSITIONING OF THE CENTER…AND NOT BECAUSE WE HAVE CHANGED THE FORECAST REASONING.

Alan Sullivan, in South Florida, is more hopeful today than he was yesterday that Dean will steer clear of his area:

My birthday present his arrived. No, it’s not a camera — that comes tomorrow. It’s a new set of projections for future hurricane Dean, keeping it out to sea, well NE of Florida. The East Coast trough is digging again. Models have consistently over-estimated the Bermuda high this summer. It has been weak or absent to the east at all times. Now the digging trough seems likely to recurve Dean enough to make for landfall somewhere from the Carolinas to Nova Scotia. Or there may be no landfall at all. …

The current trough alone should turn Dean clear of the islands, but it will probably not persist long enough to recurve Dean entirely. It will take a second one to turn the storm decisively away from eastern North America. We still have at least a week of uncertainty. But the threat to the Caribbean and Florida is diminishing.

Dr. Jeff Masters has more on Dean, as well as information on the potential Tropical Depression Five in the Gulf of Mexico (proto-Erin??) and Hurricane Flossie in the Pacific. (He’s got a radar image of Flossie nearing the Big Island.)

I’d post more, but I have baby shopping to do. :)

Mattel recalls 9 million toys because of hazards to children

Visit CNN for the latest.

My apologies if anyone is having trouble accessing the blog. It’s been very spotty tonight, possibly as a result of the two simultaneous Instalanches I’ve got going, and/or possibly because of some unrelated broader WestHost problems (I’m having trouble accessing my manager page, and the support pages, too). I’m not sure what’s happening exactly, and although I’ve submitted a help ticket, I may not be able to seriously tackle it for a while if the problem persists, since I’m going to bed now, and then Becky and I are going to Babies R Us tomorrow morning to create our registry. In other words, life is intervening and de-prioritizing the website for the moment. Not for the last time. :)

UPDATE: A-ha! It was a WestHost issue! They e-mailed me at 12:58 AM:

We are currently under a heavily DDOS attack on our Network which is currently affecting access. Our network administrator is currently working on thwarting the attack such that Network Stability returns to normal. A DDOS (Distributed Denial of Service) attack causes millions of network requests sent to us every few seconds from various locations. This is performed by the attacker to exhaust network resources which is causing the issue.

Apparently the attack has ended now.