POWERHOUSE HURRICANE REVIEW 2013

When Sir Isaac Newton was asked why he was able to achieve so much in science, he simply replied it was because he stood on the shoulders of the giants before him! Indeed, the history of the growth of science is a history of standing on the shoulders of giants. That is why we can say that Einstein’s relativity theory is a restructuring of Newton’s theory of universal gravitation in a way which brought about new knowledge or growth of knowledge.

The intention of this article is not to seek to restructure or re-order the current meteorological paradigm which make predictions about hurricanes. I would be naïve to entertain that type of wishful thinking. But I would be happy if I could stand on the shoulders of hurricane forecasting giants and see a little further ahead as my contribution to that area. The bottom line is hurricanes are a threat to me personally and to my little island – a massive threat of which we  have booth been victims - and sometimes I say to myself that “I wish I could do something to help in the attenuation of hurricanes”. And I know that there are so many people who think like me in this regard.

It is against this background that I have penned this article which I hope will in a small way at least contribute to the store of knowledge and perhaps remotely to the growth of knowledge in that area.

Three major forecasters have released their predictions for this year’s hurricane season and there appears to be much consensus among them. The table summarises those predictions:

Forecaster	No of storms	No of hurricanes	Major storms
CSU	18	9	4
Weather Bell	16	12	5
Weather Services	16	9	5
Average	16	13	4

If those predictions turn out to be right, then the Caribbean may be in serious trouble!

But can we or should we trust those predictions? They may have high levels of statistical confidence, but the major questions is: how much of the “statistical variance” have they explained? A model can have a 99% “confidence level” but may only explain 25% of “the variance”, leaving a massive 75% of unexplained variance. Is that the case of the predictive models for Atlantic hurricanes?

In the past, I have argued that  meteorological predictions are derived from “scientific measurements”, but because of the dynamic and unpredictable nature of weather, they are largely imprecise and in many cases seemingly “pseudo-scientific”. Weather behaviour tends to be like human behaviour. It is multi-plastic, multidimensional and multivariate and these introduce a wide margin of error in measurements and predictions.

Let’s look at the variables that generally constitute a prediction model. They are: El Nino phenomenon, wind shear and changes in heat. Meteorologists postulate that the presence of those three factors can significantly inhibit or undermine the formation of a tropical cyclone. Indeed, they do so and do so very well; but the big question is: are they the only variables in the equation? Obviously, they are not! That is why the models are never fool-proof! So then, what are the missing elements?

Perhaps, if the meteorologists lived in the Caribbean on a day to day basis, they might have considered changing their research paradigm from a largely armchair “positivistic” one to a more realistic “post-positivistic” one. What do I mean?

Positivism is at the heart of scientific research. Inductivists tell us that scientific practitioners use the tested “scientific method” – which they refer to as the calculus of discovery - to make observations and predictions about natural phenomena. However, many post-positivist philosophers of science (like Karl Popper, Imre Lakatos and Peter Medawar) have challenged that notion, claiming that inductivism is an intellectually dishonest “blank mind” philosophy of science.

Perhaps, it may be that “blank mind” philosophy that may have guided the meteorological research on Atlantic storms. Let me explain!

I believe that if the meteorologists and climatologists “lived” in the Caribbean and made “real” observations about the weather, then their predictions would perchance have much greater reliability and relevance. Overtime, Caribbean people have amassed a universe of historical and anecdotal data about their weather patterns and factoring that input into the existing prediction blueprints would only to enhance the validity of “scientific” predictions.

Inferences from that existing “universe of data” suggest there may be three broad variables that may have a definite moderating impact on storm formation, at least in the Caribbean. Crudely, they are “off-hurricane season” winds which moderate temperatures in the “dry season”, off-season rain and local thermodynamic variables. The combined effect of those factors significantly moderate heat levels in the tropical Atlantic and therefore significantly reduce the direction, frequency and intensity of hurricanes in the Caribbean. Most importantly, these factors are all rooted in thermodynamic theory which provide a sound theoretical platform for explanations of hurricane phenomena.

The discussion is not to discount the existing models; but rather to advise that the local thermodynamic variables as well as the naturalistic element of research should have a greater profile in the prediction models. Indeed, the existing models are very relevant in the sense that they raise our alert levels of preparedness; and for that alone, we must respect them.

So based on my theoretically crude considerations, what is my hypothesis based on what I believe to be "operationally true" for 2013 hurricane season? If the early April sprinklings persist, the salubrious trade winds continue their moderating influence and the atmospheric conditions continue to remain generally as “cool” as they are, then the number and intensity of hurricane affecting the Caribbean will also reduce!

Mind you! My hypothesis does not translate to the belief that there will be no hurricanes in the Caribbean in 2013.

Meanwhile, start getting your act together for the hurricane season which begins on June 1. We will talk about my hypothesis again in November!