Nic Lewis (an independent climate scientists) and the UK Met Office seem to be having a public debate about climate sensitivities and about the Met Office’s climate models. I find it quite interesting that the Met Office are engaging in this discussion. It’s commendable that they’re willing to engage, but it’s not obvious that they should be obliged to do so. Nic Lewis has a publication record in climate science, but it’s not particularly impressive. If one of my PhD students decided to publicly criticise a major research organisation, I’d probably tell them to wind their necks in and focus on doing science. If you think they’re doing something wrong, prove it in the scientific literature.
The basic issue seems to be that Nic Lewis has been involved in a couple of studies that use recent observations to determine the Transient Climate Response (TCR) and the Equilibrium Climate Sensitivity (ECS). The TCR is the change in surface temperature at the instant when atmospheric CO2 has doubled (assuming it rises at 1% per year) and the ECS is the ultimate equilibrium surface temperature after a doubling of CO2. Nic Lewis’s work suggests that the TCR and ECS are lower than previously thought. The Met Office’s model (HadGEM2-ES) produces some of the highest values for the TCR and ECS. Nic Lewis appears to think that there is an error in their models and somehow wants them to give more recognition to his studies that produce lower climate sensitivities.
I think Nic Lewis’s work (and it’s not him alone, to be clear) is interesting. One problem, though, is that it seems to be based largely on using the recent surface temperatures (last 30 years or so). Climate sensitivities are associated with changes in atmospheric CO2 that occurs over 100 years or more. Using short time intervals (30 years or less) means that the results are quite strongly dependent on natural variability and other effects (short-lived aerosols, solar variability) that might influence surface temperatures on short timescales, but average out on the longer timescales. Therefore, that Nic Lewis’s values differ slightly from other values is not that surprising. I suspect that many would see the method used by Nic Lewis and collaborators as a sensible sanity check, rather than as a method that is somehow more robust than other methods. What would have been surprising would have been if Nic Lewis’s method had produced results very different to those produced by other methods. That it didn’t (the ranges overlap) gives some confidence that the various other methods are at least consistent. If you want to know more about this, there’s a good discussion on Skeptical Science.
The UK Met Office has already responded to Nic Lewis’s criticisms in a blog post a few days ago. You can find Nic Lewis’s original criticism here. Julia Slingo, Chief Scientist at the Met Office, has written personally to Nic Lewis and the letter is posted here. I think Julia Slingo’s letter is reasoned and balanced and makes some very valid points. Nic Lewis has responded to this letter in a post on Climate etc. called Nic Lewis vs the UK Met Office.
Given how convoluted this whole episode seems to be I thought I would just make some comments with regards to Nic Lewis’s response to Julia Slingo’s letter. He seems to be critical that they focused mainly on the estimates using the period 1970-2009 and didn’t include the results from the period 2000-2009. I don’t know if they did or didn’t, but it would seem to me that focusing on the longer time period is more sensible than focusing on the shorter time period. I’m not even sure why they bothered considering the period 2000-2009 as it seems obvious that the longer the time interval considered, the more robust the result (although maybe that’s just me).
It seems as though one of Nic Lewis’s criticisms is that the Met Office model cannot explore low values of the ECS. Julia Slingo’s response is
But the key point is that the relationship between aerosol forcing and ECS is an emergent property of the detailed physical processes sampled in the PPE simulations.
I take this to mean that the the ECS is a model result that is a consequence of the various initial conditions and the physical processes included in the model. If the model does not produce a low ECS, then that implies that a low value is not possible given those input parameters and the physical processes included. You can’t force the model to give you a low ECS. As far as I’m aware, Nic Lewis’s results are based on a short time interval compared to the total time over which CO2 will double. The models consider the entire doubling time and continue on until the ECS is reached. It’s not unreasonable that there will be differences. In some sense, all these different results are useful and interesting and should, in some sense, remain independent. You can’t insist that the Met Office model produces lower ECS values simply because you think your work has more validity than others. You might be wrong. Of course, if one can find an error in a model, it should be fixed but getting a different result using a different method does not immediately imply an error in the model.
Maybe the most confusing thing about Nic Lewis’s response was the comment below
That strength cannot overcome the basic problem that HadCM3 cannot sample low aerosol forcing, low climate sensitivity combinations and is therefore an unsuitable model for this PPE study.
I understand how aerosol forcing is an input to the models, but he seems to be implying that somehow the models should be able to probe low climate sensitivities. These, I thought, were outputs so either the model produces a low climate sensitivity or it doesn’t. You shouldn’t tune the model to probe low climate sensitivities. The models should be physically motivated. How is it even remotely acceptable to argue that the model is wrong because it doesn’t produce the result you want? If we knew what the TCR and ECS actually were, then maybe one could conclude that the model was wrong, but we don’t.
Julia Slingo finishes her response by saying
As I said we appreciate your contributions to the literature on these topics; but the implications of climate change are so profound that it is essential that scientific debate takes place in the appropriate forum. With this in mind I think it is appropriate that further discussion be subject to proper peer review, through the scientific journals.
Nic Lewis, sadly, gives the standard response to such a suggestion. This appears to represent an attempt to stifle reasoned scientific debate. No it’s not an attempt to stifle reasoned debate. It’s a recognition that public debates about science aren’t particularly valuable. Exchanging ideas at conferences, or exchanging a few emails or letters is part of the process. At the end of the day, however, you need to publish your work and have it evaluated by others in the field. It’s almost as if Nic Lewis thinks that this public debate could lead to some kind of resolution. No it can’t. It could lead to some changes in the modelling and in the studies that might ultimately resolve the issues, but until the works been done, tested, published, evaluated and checked we really won’t know.
That’s why Nic Lewis should get on with his work and should really leave the Met Office scientists to get on with theirs. I’m not suggesting that all correspondence should stop. Simply that expecting the correspondence to lead to some kind of scientific resolution is unscientific. Ask questions. Make suggestions. Engage scientifically. All fine. Don’t, however, assume that because you’ve published a paper suggesting that results obtained by others might be wrong means somehow that they all have to bend over backwards to incorporate your work. That’s not really how it works. I think what Nic Lewis is doing is interesting and valuable, it’s just not – in my opinion – more interesting and valuable than the work being done by others.