Really Sciency

Visit my other blog 'Really Sciency' looking at Climate Science and its portrayal, misrepresentation and denial in the media.

Sunday, 5 December 2010

Web publishing is a poor substitute for peer review.

Every so often, too often to my mind, a piece of 'research' appears on the web that has science deniers pushing as evidence that supports all their misconceptions.This web publishing circumvents the very corner stone of science publication, peer review. This doesn't usually seem to be a problem for the deniers, in fact many claim that peer review is overrated or flawed, but rather than than try to offer something better they just dismiss it in favour of just about any sciency looking stuff that confirms their anti-science bias.

One such 'research paper' was about the intensity and frequency of hurricanes by the well know climatologist Les Hatton;

OK you got me, Mr Hatton isn't a climatologist but the lack of appropriate qualification is no barrier to those that need something to call research that supports their beliefs, but this paper was being pushed at me on a forum as good science.

I shouldn't have been surprise by some people taking this ‘paper’ seriously that isn’t by a climatologist and isn’t peer reviewed.

I always give such things a wide birth, because if it was good science it would not be self published. When I do get involved in looking at such ‘evidence’ I always regret it as it quickly becomes apparent that my time has been wasted and the flaws are obvious to anyone who truly thinks critically.

Yet against my better judgement I decided to see what all the fuss was about. So here is my appraisal of the paper and why, as Dr Hatton agrees, it is un-publishable. Don’t take my word for anything I say, I admit to being completely unqualified in climate science – however that may make me as qualified as Hatton himself.

First strange thing is that Hatton seems to be a true academic but does not appear to have used this to any advantage. I’m certain he must have published in his own field at least so he must me well aware of the quality of material needed to get published in a peer reviewed journal but he does not make any attempt to do so here. Also if he contacted scientists working in climatology there is a good chance he would get replies and answers to any queries he had about the IPCC conclusions – he does not appeared to have done this.

The paper looks at 6 statements made by the IPCC and compares them to his own analysis of some of the raw data.

Here is the first problem, the IPCC statements are made based on many papers all of which are available and referenced in WG1. I have only looked at Chapter 3;

Among the hundreds of references I can find over 60 that relate to hurricanes, cyclones or storms. All of which use much more detailed analysis than Hatton uses. Basically if he wants to compare his analysis with that expressed by the IPCC he needs to use the same data sets, like for like, – or have a very good reason for not doing so. I suspect any journal considering this for publication would expect that at least.

There is no evidence that he has read any of these papers or understands how the IPCC conclusions were arrived at. Why should his simple graphing and analysis be more credible than any of the papers published by researching climatologists?

His conclusions are also lacking.

Conclusion 1, he agrees with the IPCC but adds “this is only a return to the level of activity it had in the 1950s and 1960s after a 25 year period of relatively low activity from around 1970.”
But his analysis does not look at much before then. On what basis can it be said to be a return? There is data going back to 1851 for this area, and scientists say that the data is fairly good for the last 100 years. So why hasn’t he used as a check before making such a conclusion?

NASA and PEW appear to have and do not conclude as he does;

About twice as many Atlantic hurricanes form each year on average than a century ago, according to a new statistical analysis of hurricanes and tropical storms in the north Atlantic. The study concludes that warmer sea surface temperatures (SSTs) and altered wind patterns associated with global climate change are fueling much of the increase."

Is the frequency of hurricanes increasing?

Globally (not just in the North Atlantic), there is an average of about 90 tropical storms every year. According to the 2007 Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC-AR4), globally "[t]here is no clear trend in the annual numbers [i.e. frequency] of tropical cyclones."
However, in the North Atlantic there has been a clear increase in the frequency of tropical storms and major hurricanes. From 1850-1990, the long-term average number of tropical storms was about 10, including about 5 hurricanes. For the period of 1998-2007, the average is about 15 tropical storms per year, including about 8 hurricanes. This increase in frequency correlates strongly with the rise in North Atlantic sea surface temperature, and recent peer-reviewed scientific studies link this temperature increase to global warming.

Conclusion 2, he dismisses on the basis that climate models were used – “Climate models are corrupted by unquantifiable errors. This could be due to inadequate physics, software error or both.”

He is a software specialist; shouldn’t he be more specific about what the problems are? Shouldn’t he give some evidence that what he claims is actually the case?

And yet the IPCC conclusion was that “The observed increase in hurricane intensity is *LARGER* than climate models Predict”. So it seems that if there is anything wrong with the models it is to err on the side of conservative estimates.

Conclusion 3 – Agrees with the IPCC.

Conclusion 4, he claims the IPCC has made “a worryingly incorrect statement”. But that is only based on the data sets he used and with no input form any one qualified in climatology. Here is where his lack of qualifications lets him down. As a respected academic he could try to contact the authors of the relevant chapter of WG1 AR4 or the authors of the papers it was based on and ask on what basis such a conclusion was drawn. This again is the least that a journal would expect before publishing a paper stating that other research could not actually be confirmed.

Conclusion 5 & 6 suffer from exactly the same problems as conclusion 4. They were made based on his limited data and climate knowledge. He has made no attempt to qualify the IPCC statements and on which basis they were reached.

So for a paper that is aimed at confirming or refuting statements made based on many published science papers it apparently fails to look at any of the evidence presented in them, and his conclusions are based on an analysis of  very narrow data, both in quantity and time frame, without seeking any input from those who are experienced in the subject. He also dismisses, without presenting any evidence, climate models that may have been used in some of the many papers that the IPCC conclusions are a summary off.

Yet his paper in un-publishable, simply because no effort has been made to write to anything near the required standard for a credible journal to consider.

I do of course welcome any thoughts and opinions on the issues I have raised.

1 comment: