Recipe for a Storm — Forecasting a Hurricane Season

University of Wisconsin-Madison researchers investigating the ingredients that go into a hurricane think they have found a reliable basis for predicting the overall strength of a hurricane season. Jim Kossin and Dan Vimont have found a basin-wide circulation pattern that offer one possible explanation in the previously unexplained differences in long-term hurricane trends. "Kossin and Vimont, a professor in the Department of Atmospheric and Oceanic Sciences, noticed that warmer water is just one part of a larger pattern indicating that the conditions are right for more frequent, stronger hurricanes in the Atlantic. The atmosphere reacts to ocean conditions and the ocean reacts to the atmospheric situation, creating a distinct circulation pattern known as the Atlantic Meridional Mode (AMM). The AMM unifies the connections among the factors that influence hurricanes such as ocean temperature, characteristics of the wind, and moisture in the atmosphere."

Comments

Re:Wisconsin

[nido]

Good thing these fine young scholars are boldly venturing forth into the areas of meteorology most crucially important to the Midwestern region of the United States. Oceans drive climate systems across the entire planet. Surely you've heard of El Nino and La Nina? One is a 'warm ocean', and the other is a 'cool ocean'.

Furthermore, the positions of warm and cool spots in the ocean control where the jet streams flow, and the jet streams determine who gets rain and who gets drought. I understand that the warm anomalies are probably caused by underwater volcanic activity, but this is one aspect of the earth's geology that we have precious little data about - those underwater volcanoes are notoriously difficult to study...

Wisconsin has lots of farming which is dependent on rainfall, so it's entirely appropriate that they're trying to improve their forecasting models.

Every June

[midmopub]

Every June in Florida the local news is full of reports by experts that this year would be the worst hurricane season on record. After 7 years of hearing the same stuff I started to tune it out.

Re:Every June

[Titoxd]

Unfortunately, that is a very real risk associated with long-term hurricane forecasts: Assuming that because there may be a lot of storms, they may devastate an area in particular (something the mass media is particularly good at). A hurricane season can have dozens of storms, and having none affect land. On the other hand, a season may have very few storms but be extremely damaging, like 1992 was. It really takes only one bad storm, like Andrew in 1992 or Mitch in 1998, to turn lives around.

In reality, people have to realize that predicting weather is an inherently unstable mathematical problem, so longer-term forecasts are usually not that accurate. On the other hand, short-term forecasts keep getting better as the understanding of the physical phenomena increases, along with more computational power to throw at the good old models. A bit of preparation before hurricane season never hurts, though.

Re:Every June

[erc]

In reality, people have to realize that predicting weather is an inherently unstable mathematical problem, so longer-term forecasts are usually not that accurate.

And probably never will be. When I studied meterology in college years ago, I remember the complex math in Methods In Climatology, one of our textbooks, and it was every bit as bad as the math I had to take for physics. And that was 30+ years ago - it's only gotten even more complicated since...

I'll give you a storm

[cerberusss]

You want a storm? Forget your wife's birthday, that'll bring a storm.

Re:I'll give you a storm

[flyingfsck]

No, it will cause the start of an ice age.

"I'm so hot and she's so cold... cold as a tombstone."

I wonder how many times Mick forgot Jerry's birthday.

Well, yeah

[FroBugg]

These guys aren't the first to "discover" this connection. The article does a piss-poor job of explaining it, but basically the global thermohaline circulation varies in speed. Sometimes it runs fast and sometimes slow. The fast periods tend to last about 15-20 years, with the slow ones a little shorter, and it's a self-correcting cycle. Our observed records of this pattern correspond very well with the last hundred years of Atlantic hurricanes.

Global warming is a major threat, and it's going to be responsible for a lot of weather problems, but Atlantic hurricanes aren't one of them. Once you increase Atlantic surface temperatures to a certain point, you actually tend to increase upper-level shear, which is extremely disruptive to hurricanes.

The 2005 season was so terrible because four of the storms that made landfall passed over the extremely warm loop current in the Gulf of Mexico shortly before making landfall. It was a busy season and we just had some really bad luck on top of it. Even considering this, they were all weakening when they actually hit, and the destruction of New Orleans is entirely due to shoddy construction of the levees. Katrina may have been a cat 5 at sea, but the levees failed in category 1 conditions.

Re:Well, yeah

[spvo]

Yes, nature will most likely always be too complex to predict, but that is because its a massive chaotic system. I would guess there will always be too many unforseen initial conditions that would blow up and lead to very different, and unpredictable, results. In any case, it has nothing to do with the heisenberg principle. It doesn't apply to macroscopic systems, and I think a hurricane definitely qualify as large.

Re:A little late for this past season

[ThreeGigs]

What's that old joke?

Dear Weatherperson,

I'm writing to let you know I just finished shoveling 20 inches of 'partly cloudy' off of my back porch.

Yours truly...

Re:Canadian forecasters: Very cold winter ahead

[aevans]

Canada isn't a certain localized area. It's a huge chunk of the earth's surface.