Institute of Physics submission to CRU Inquiry

Published March 3, 2010 by Sean

I was meaning to write about this when the story first appeared on 28th Feb of the IoP making a very forceful submission to the Science and Technology committee’s inquiry into the data disclosure (climategate emails). The submission is very strong on the behavior of the scientists and their actions which appeared to make replication and verification of the temperature series and proxy reconstructions impossible. The submission was highlighted by ClimateAudit, WattsUpWithThat and BishopHill, where some of the contributors seemed to read it as saying that this was final acceptance that the science was a sham.

Whilst RealClimate don’t have a seperate thread, their refutation is pretty entertaining – they seem to struggle to work out who the Institure of Physics are to start with, then accuse them of being industry led (actually, they seem to be a predominantly academic organisation, and have printed many many articles in Physics World lamenting the fact that the planet is doomed). RealClimate site responses (from the ‘real climate scientists’) are all on message, supporting the ‘standard practice’ of not releasing code.

By 2nd March (the day after the oral hearings, which recieved a good amount of media coverage), the IoP issues a clarification stating that their submission does not address the question of the consensus, and the Institute does not doubt that. Quite what has caused this is not clear, but there is certainly a severe mis-understanding regarding the principle of repeatability.

The view from RealClimate is that it is OK to keep the data secret, if it gets into the wrong hands it could be mis-used in a dangerous way, and anyway the process has been replicated independantely and shown to be good. Since people like Gavin speak from a modeling point of view, I can see some logic in saying that making the models open wouldn’t be much of an advancment of the science. (The code is probably ugly, there are several pseudo-independant models, and the models are not particularly precise anyway)

Frank O’Dwyer keeps repeating that open-access should apply to all, sceptics included, yet fails to accept that several people have been attempting to verify the methods used in several studies in order to check the methods. The accusation of (unintentional) selection bias in datasets, inapropriate adjustments to data series, manipulation of the peer review process, etc. are not issues which the skeptical community need to answer. Much of the interest is in series which to date show a fairly small degree of warming, so small errors are significant. Even if most users of the data only use it to confirm correlation with (e.g. hurricane disaster losses) there is a need to enable the reconstruction itself to be challenged. The various tree ring cronology questions which have been asked in the past show that if the reasons for selecting a specific series are secret, it rapidly becomes impossible to determine how robust a method is to being fed bad data.

Some commentors have argued that the IoP’s submission justify doing nothing, and that the precautionary principle is not applicable to the climate change scenario. it may be this reasoning which prompted the clarification. The precautionary principle theory goes along the lines of ‘even if it’s not proven for sure, we should react just in case it is important to start now’. All that has been expressed by the IoP is that the scientific method has not been sufficiently open. Extrapolating from that to say that the data we have should be thrown away, and we need not worry about anything is pure flat-earth territory. We have a number of facts, which justify a level of concern. Some of those facts have not been tested as well as an independent scientist might have assumed, but that is still the best case analysis to date. Some policies are still sensible, regardless of the need/desire to restrict CO2.

The IoP has said it is good to question the consensus, and to encourage critical analysis of the reconstructions. It should be enough for now to agree that more open access to the data and methods is a priority – a point which is still contested by many of the alarmist proponents. questions like UHI adjustments and divergence must be brought back to the table at the scientific/statistical level. This is a straightforward view which requires no judgment of value or personal priority, it is basic scientific process.

From a policy point of view, there are many things changing. The datasets are slowly growing, agreement with model projections is decreasing, trends are changing, the independence of the consensus position is coming under question. The validity of some of the data has a very small influence on this today. It is likely that people will become less prepared to support expensive measures in the short term, but it is best to assume that nothing will emerge to prove that the science is completely wrong.

Filed under Science

Comments (0)

Comments RSS - Trackback - Write Comment

No comments yet

Write Comment

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.