And thus another chink in the armor surrounding discussions of climate change. It is all very well to make predictions about how warm the world is going to get, and to try and frighten politicians into changing policies on the basis of alarming predictions of the effects of global warming , but the problem comes when those predictions fail to come to pass. It is all very well for folk to state, as George Monbiot does that this is not the case.
Even their premise – that the Met Office "confidently predicted a warmer than average winter for Britain" - is wrong. Here's what it actually said:I guess he and I don’t speak the same language, since I read the Met Office prediction as being that there is an 87% chance of a warm winter – which this surely isn’t. But when you’ve tied yourself as tightly to the global warming mantra as he has, it is going to be hard to face up to the reality that perhaps it isn’t so.
Early indications are that it's looking like temperatures will be near or above average. But there's still a one in seven chance of a cold winter – with temperatures below average.
No confidence there, no certainty, and no single prediction.
Instead we are at a point where businesses are being closed in Europe to conserve fuel.
The national grid was forced to reduce the supply to companies in the North West and East Midlands shortly after issuing its second warning in three days that the system was running out of gas. Demand in recent days has been 28 per cent above seasonal norms and is likely to increase today, when temperatures could fall to minus 22C (-8F) in some areas after what was expected to be the coldest night of the winter so far.and in a reminder to the EPA Administrator that cold can be a killer there is the possibility of up to a million sheep being lost because of the continued cold in the Scottish Highlands.
And it is not just farm animals that suffer, again in contrast with the EPA Administrator's views, data is now available on how many folk die due to the increased cold, and why.
Last year's low temperatures saw the highest number of "excess deaths" - the number of those who perished over and above what is normal for the time of year - for nearly a decade.This is not just a theoretical debate being carried out based on computer models and controlled publication of only papers favorable to one position. But this is a real ongoing situation where those who rely on the predictions that have been made must now face the actual reality of what is happening.
The 40,000 "excess deaths" in England and Wales represented a rise of nearly 50% from the previous year. In the South East, where people were perhaps least prepared for a cold snap, deaths nearly doubled.
Now it is true that one winter does not, alone, make the climate change story untenable. But given that there has to be a reason for this, and it is not evident in the climate models on which so much has been invested, then it is fair to ask why not? It is reasonable to ask what is causing this drop in temperature, and to question whether whatever is causing this cold has not, in the last decade, been working in the opposite sense to cause warming. And if my comment from a year ago relating the a change in attitude toward an unconditional acceptance of global warming theories is beginning to be validated, then I think that the questioning will intensify somewhat when the costs of a colder period are added up.
The problem goes beyond the immediate short-term, because the predictions for providing energy for power and heat, and the quantities of natural gas that would be required are proving inadequate in the face of the current situation. This is the “knock-on” consequence of putting too much faith in the models. There has been talk of the need to apply the Precautionary Principle, in taking steps to ensure that increases in carbon dioxide levels in the air do not lead to global warming and the consequences that are increasingly exaggerated in order to drive the debate. There has been too little talk of the precautions that should be taken if the theory is wrong and we will need more fuel rather than less. Or, in this case, more salt, since there are problems in the countries where supplies are proving inadequate.
Police in the Republic are set close a swathe of major roads over the weekend as supplies of salt to keep them open finally run out.In reality the winter will, after a while, be over, and the debate will continue. What the current weather has done, as the ABC comments illustrate, is to break the mind set that unquestioningly accepts the climate change mantra. Just how badly that break has been is going to become evident in the coming months. (Note how I become very nervous of making a prediction).
As the Irish government was last night accused of dismally failing to deal with the crisis, even more severe weather is forecast for the next 72 hours. Thousands of schools will not now reopen until well into next week at the earliest.
Local authorities are likely to run out of salt in just two days, raising fears that gardai may be forced to close roads.
So HO, since you don't like it when I'm mean to you, let me be socratic instead.
ReplyDeleteYou have had a career as a senior faculty member in an engineering discipline at a research university, yes?
So you have some training in the kinds of statistical inference used in engineering and science, yes?
How improbable must the event be in the case of the null hypothesis in order for the event to be considered statistically significant?
So in a single trial, if a prediction of condition X with 87% probability does not come true, is that statistically significant?
Oh, one last question. At what level in the curriculum at Missouri University of Science and Technology is the concept of statistical significance taught?
ReplyDeleteIn my education it was covered in upper division statistical methods when I was undergraduate.
Grin:
ReplyDeleteStuart, Stuart - you should watch the Head of the Met Office being interviewed on the BBC.
This is the third major prediction in a row that they have got wrong (last winter, last summer and this winter) and M. Hirst has just been given a 25% pay raise. He does a very poor job of defending the prediction.
And, as an experimental researcher (rather than a theoretical one) I have run numerous experiments over the decades and thus have more than a little familiarity with statistics. (I have even taught geostatistics for a couple of lectures in one of my classes, some years ago).
But the experiments that I do are generally expensive and it is rarely possible to do the full statistical runs that would be ideal, and so we have also learned how to extract data from smaller test runs, and to grasp the possible meaning of anomalous results. (Which have led on to two significant improvements in the technology).
As to where I learned the statistical tools that I use (which is I presume the point of the second comment) - while I got some statistical training as a student (bear in mind that finding where a mineral vein is going underground relies in part on a statistical analysis of grade levels from boreholes; and the failure of mine structures which can have varying conditions and loads, that are rarely fully known, is also increasingly analyzed statistically) I have been both self taught and have sat in statistics classes here that were taught by faculty in our statistics department, and did actually take a course in Statistics and Probabilistic Methods at MIT some years ago - information which I have used and built on in the years thereafter.
I'll mention that, at the same time that it's unseasonably cold in the UK, I've been unseasonably warm in the Pacific Northwest. I don't think of that as AGW either but it might, in my case, have something to do with el Nino.
ReplyDeleteH.O - I'm not taking any position on the Met office's performance, just pointing out that you are making elementary errors of statistical analysis in your writing.
ReplyDeleteI understand the point, blame it rather on my bad habit of over-condensing information to the point that I simplify too much, and leave out additional information (such as the previous record).
ReplyDelete