Hurricane forecasts were way off again last year so if you're still wondering if a trained chimp 'can predict hurricanes better than NOAA'(1) a Nature Geoscience article has good news for you; forecasts can still be wrong 75% of the time but now can be wrong for years in advance too.

A group of researchers hope to make forecasting a lot more predictable.  Currently, when hurricane predictions are wrong, scientists say 'climate modeling is complex' which really does not earn climate science a lot of credibility with climate change deniers, but hurricanes are a lot trickier than the simple physics that says if you add more people and machines and animals, closed spaces warm up.  And climate skeptics tend not to understand what the word 'probability' means.

Heck, hurricane legend Dr. Bill Gray got his own 'perfect storm' from hyper-militant environmental kooks when he stated in 2006 that Al Gore was wrong for implying that global warming caused Hurricane Katrina, so hurricane forecasters basically can't win with either side.  Regardless, hurricane forecasting recently is on a par with tribal shamen tossing bones, to an extent this video has an actual chimp doing the prognostication...




...so help is welcome that doesn't involve a simian.   What researchers have going for them now is better understanding of patterns due to more thorough modern data, and their decadal climate prediction models can also incorporate events like volcano eruptions and more recent estimates of greenhouse gases.

Obviously when some groups see 'greenhouses gases' in a sentence they go to their web browser bookmarks, look up some pithy phrases and try to think of clever new ways to use the word "liberal" in a sentence, but large-scale climate factors, including volcanic eruptions and, yes, pollution, have to be factored in.   Tropical multi-decadal signals, La Niña cycles and temperatures in the tropical Atlantic Ocean and Caribbean Sea mean they can at least get closer to the perfect algorithm for determining how many hurricanes there will be.

But it's primarily an academic waypoint toward the true goal everyone wants to achieve; predicting not just how many hurricanes will occur, but how many will make landfall.   That sort of prediction is farther off but if they first learn what makes hurricanes more frequent (even if part of it turns out to be climate change) it would be a big step.

The NOAA has gotten smarter about predictions also, being more clear about what their predictions are and saying at first that 2010 would be  "above-average" (6-8 hurricanes) and then revising it to be "extremely active", (8-14 hurricanes) which is a huge range in science - and then only with a 70% probability.   Since there have been 12 so far, they can't really be wrong.

Though the chimp may disagree.

Citation: Doug M. Smith, Rosie Eade, Nick J. Dunstone, David Fereday, James M. Murphy, Holger Pohlmann, Adam A. Scaife, 'Skilful multi-year predictions of Atlantic hurricane frequency', Nature Geoscience (7 November 2010) doi:10.1038/ngeo1004

NOTE: 

(1) A second video will be released on December 1 by The National Center for Public Policy Research, at the conclusion of the hurricane season, with Dr. Hansimian's reaction to the performance of his forecast against the NOAA forecast.   The National Center for Public Policy Research says it is a 'free market' group, which is codespeak for 'we don't believe in global warming'.