Undoubtedly everybody read the news that the hurricane researchers at Colorado State are still predicting an active hurricane season with a 69% probability of at least one hitting the U.S. shoreline. The media, which doesn't have time to read a 43-page report pretty much just quoted the abstract and said it had something to do with El Nino. Turns out El Ninos cause "vertical wind shear" in the Atlantic which makes it hard for hurricanes to form. And if the The Kid does dissipate you can combine it with a oddly warm North Atlantic this year for a serious hurricane season. For a further geek out and to link to the full report click here.

Every year for the last whole bunch, they predict an above average season - sooner or later, they'll get it right

OK, so El Nino makes it hard for tropical storms to form.
I am not sure why it would disrupt them instead of seeding them, but OK, I will accept that as given until I can find out more.

You have a 40 page report there which is a bit deep. Thanks for the read. I found their comments after page 30 interesting.

Live along the gulf coast, east coast or Florida long enough and your going to get hit by a hurricane. Events are independent. Every year the odds that that year is the good/bad resets. Nine years at 90% doesn't mean tenth year is a sure thing. Odds of 90% approaches an exact 9 out of 10 only with an infinite string of events.

Bottom line is that the predictions are not wrong. They are just statistics and subject to the limitations of statistical methods. We may be being protected for a time by El Nino. Keep your powder dry.

A hurricane is a gigantic heat pump. Lop off the top half - which is what wind shear does - and the pump breaks.

Everyone has had problems with seasonal estimates recently. Colorado State's contribution was that such analysis could be done at all - that conditions dis/favorable to hurricanes were the result of other _observable_ conditions months earlier - at a time when hurricane seasonal forecasting was as credible as watching goats to predict earthquakes. It's still a hard problem and I suspect the annual press releases have more to do with getting research funding than providing homeowners with actionable information.

It's interesting that Bill Gray's name is on the paper. At one time he was nearly forced out at Colorado State (politics).

The problem with hurricane predictions is whether it's bad or not doesn't depend on the hurricane but where it hits land - and then that matters to within a few miles.

It's like being able to accurately predict how many drunk drivers will be on the road tonight - but not being able to say how many will be on the same road at the same time as a child is crossing.

If you look at the history of people predicting hurricanes before the season starts, it is pretty abysmal.

They can't even accurately tell me if it is going to rain tomorrow and they want me to believe they can predict a whole season of hurricanes?

They don't predict. They state that the statistical chance of major hurricanes coming to life is increased. That also increases the chances of major hurricane making landfall. No attempt to nail down when or where.

Pretty far from prediction by my standards.

To a homeowner that's true, but not at all to other groups.

FEMA and the electric utility groups would benefit greatly from accurate 6+ months estimates in terms of resource & manpower needs planning. Dates & places would be nice of course, but merely being able to budget, hire in the spring, etc is a big win even if they don't know precisely where or when hurricanes will make landfall.

It's those groups that I bet are funding the hurricane season prediction research, because if it got good enough it would be really useful to them.

It's probably impossible to predict time & place 6+ months ahead since the processes are "chaotic" in a technical sense. A scientific paper with one of my favorite titles ever nails it: "Predictability: Does the Flap of a Butterfly's Wings in Brazil Set Off a Tornado in Texas?"

Actually I think 24 hours rain forecasts are pretty good now. Make sure you know what the forecast means: "20% chance of rain" usually means there's a 20% chance of rain somewhere in the forecast region. It doesn't mean there's a 20% chance of rain where you are (the NWS forecaster doesn't know where you are!)

There's a good analogy here. Two week rain forecasts are nearly useless at saying if it will rain. But seasonal total rainfall estimates are a lot better, saying that there will be more or less rainfall than "normal" for spring/summer/fall/winter. Farmers use seasonal forecasts as one factor in deciding what to plant etc. The seasonal rainfall forecasts aren't perfect and they don't bet the farm on them, but the forecasts are definitely useful in increasing the odds of a successful farming season.