The Intersection


Today is officially the last day of the 2007 Atlantic hurricane season. Not that we can’t have a storm after November 30…but it becomes less and less likely.

And how do we summarize the year? Well, less active than expected, especially when it comes to the frequency of the most intense storms. That’s two years in a row the forecasts have overshot in this respect.

Anyways, over at the Daily Green I now have a more in-depth post-mortem on the season. Some highlights:

The relative quietude of the 2007 Atlantic hurricane season seems particularly mystifying in light of the fact that we saw La Nina conditions develop in the tropical Pacific — which usually correlate with above average Atlantic storm activity. So lets turn to Klotzbach and Gray to hear their reasons for why we had a quiet year. Notably, sea surface temperatures were 0.2 to 0.5 degrees (Celsius) cooler than the average of the past twelve years — a development that Klotzbach and Gray ascribe to high levels of African dust blown across the ocean, which deflected sunlight. Meanwhile, although La Nina tends to reduce storm-killing vertical wind shear over the Atlantic, this time around it really didn’t.

Meanwhile, the month of September was just plain quirky: No less than eight storms formed, but because some were close to land and others just happened to encounter other unfavorable conditions, none developed past short-lived Category 1 hurricane status.

But even on top of all this, October remains the deepest mystery. Or as Klotzbach and Gray put it: “We are still struggling to understand why October was not more active.” All the conditions were favorable. But very little happened. It’s still not really clear why.

The upshot from all of this: We still have a great deal to learn when it comes to forecasting seasonal hurricane activity in advance. Nevertheless, all forecasters agree that whatever may happen in a given year, we remain in an active hurricane era in general. And even though the United States has now enjoyed a welcome respite for two straight years, there’s no reason to expect easy times to continue.

You can read the full item here.


  1. #1 Karl
    November 30, 2007

    First – I do accept the theory of AGW.
    AND, I understand the difference between weather and climate.
    BUT, you (we) must be aware of the politico-scientific problems these failed predictions cause. All of the anti-AGW people are saying – obviously our knowledge about climate is not as good as was advertised, so why should we believe the predictions of GW?
    Maybe NOAA (or whoever) should get out of the prediction business.

  2. #2 Lance
    November 30, 2007


    You say that you accept AGW theory, yet you warn against making predictions based on it because,

    “All of the anti-AGW people are saying – obviously our knowledge about climate is not as good as was advertised, so why should we believe the predictions of GW?”

    Theories are only as good as their predictions. That is what science is all about. The response that you fear is exactly the correct one. When a prediction made by a theory is falsified then so is the theory.

    Karl, your last name obviously isn’t Popper.

  3. #3 Fred Bortz
    November 30, 2007


    You are wrong to say that NOAA should get out of the prediction business essentially because the general public doesn’t understand error bars and uncertainty. NOAA exists, in part, to enable people to make decisions based on the best predictions. The people who have to make those decisions do understand the limitations on the predictions.


    You have read Chris’ books, but you seemed to have missed one of the central points of Storm World (click my name for my review).

    The book notes that the science is simply not good enough to make firm predictions of the effect of increased sea surface temperatures on the number or intensity of hurricanes. Both the empiricists like Gray and those who trust scientific storm or climate models like Emmanuel recognize that there are too many factors that can derail a prediction for one basin in one season.

    It’s also important to distinguish between the IPCC’s consensus on AGW and its assessment of the relationship between GW and hurricanes.

    The IPCC consensus on AGW is based on vast data sets and models that, though necessarily imperfect, have been tested against a variety of historical data and are proving themselves robust by their predictions.

    There is no such consensus on the relationship of GW, whatever its cause, and hurricanes, only some trends. The historical data is much more limited, and the models are not fully tested. It’s too soon to declare either support of the theories or their falsification, especially by looking at one year’s data in one basin.

    I presume the models will be refined as a result of the totality of this year’s data around the world. Gray et. al. will refine their empirical interpretations, and Emmanuel et. al. will refine their models. Both refinements will be useful for future seasons.

    The success or failure of hurricane modeling will require many more years of data collection.

    An excerpt from my review that amplifies both of these comments:

    But on the question of how global warming would change hurricanes, the IPCC conclusion is much less certain. Warmer seas might produce more frequent or more severe hurricanes, but many other atmospheric and climate factors also contribute to storm development.

    To climate scientists, this is an important and exciting open research question. Policy-makers and businesses, on the other hand, prefer certainty. How strong should governments require levees and storm barriers to be, not only around New Orleans but also in low-lying coastal cities around the world? How much money should insurers set aside for future disasters?

  4. #4 Neuro-conservative
    December 3, 2007

    This forecasting debacle demonstrates, in microcosm, the problems inherent in prognostic “science” based on simplistic, discrete-variable models of complex systems.

    Even the most basic of data, such as the counting of storms (or the historical temperature data), are far less certain than the AGW religionists would have you believe.

    Of course, all the changes in the methods of counting storms lead in the direction of increased storm activity in recent years.

    Adding in all the hand-waving in these post-hoc models (African dust? Fairy dust? whatever), it would be an improvement to merely be GIGO.

New comments have been disabled.