I was hesitating about writing this post because I’m going to find it hard to avoid some snark. Apologies in advance. A number of people seem to think that Nic Lewis’s submission to the UK government’s IPCC inquiry is very good. Judith Curry seems to think it’s a tour de force. Personally, it seems a little like someone trying to say: I wrote a paper. I’m proud of my paper. I think my paper is better than everyone else’s papers. Why didn’t the IPCC take more notice of my paper? Also, I’m only going to mention uncertainties when it suits me to do so.
I don’t have any particular issues with Nic Lewis’s work. I’ve worked through Otto et al. (2013) in quite a bit of detail. I haven’t quite managed to work out what he’s getting at in Lewis et al. (2013), but have no reason to think that it’s not a reasonable piece of work. It’s just that, in my view at least, if you’re interested in doing science, you immerse yourself in the subject, you build up a portfolio of work, and you gain credibility with your peers and with others. Getting your name on 2 or 3 papers doesn’t suddenly make you an expert and doesn’t, typically, give you the credentials to review the validity of other aspects of the same field. There’s nothing fundamentally wrong with this, it just seems an odd way to build a scientific career. Might make it seem that your desire to do science is driven more by the policy implications of what other work suggests, than by some deep desire to really understand something.
Anyway, Nic Lewis’s submission is quite long, so I thought I might highlight a few things and make a few comments. He discusses the Transient Climate Response (TCR) and the Equilibrium Climate Sensitivity (ECS), but – as far as I can tell – makes no mention of the Earth System Sensitivity (ESS). The ECS is the equilibrium temperature after a doubling of CO2, but is based only on fast feedbacks. It’s now accepted that there are likely slower feedbacks that on longer timescales will lead to a higher equilibrium temperature than ECS estimates would suggest. He goes on to say
The IPCC considers all observational ECS estimates in AR5 WG1. It concludes that estimates based on
- paleoclimate data reflecting different past climate states
- climate response to volcanic eruptions or solar changes
- satellite measurements of short-term changes in heat radiation
are unreliable and/or unable to provide usefully well-constrained estimates. I agree with this conclusion. That leaves in essence only estimates based on observations of warming over multi-decadal periods. Useful surface temperature records extend back approximately 150 years (the ‘instrumental period’).
I’d love to know if this is correct. I can’t find anything in the IPCC document that seems consistent with this claim. Essentially he seems to be arguing that the only reliable estimates are his own.
He goes on to say,
However, as noted above, between AR4 and AR5 there has been a major reduction in the IPCC’s estimate of the cooling strength of aerosol pollution, which necessarily implies that estimates of climate sensitivity should be substantially lower than previous estimates. However, the current CMIP5 generation of GCMs still very largely reflects the earlier understanding of a stronger aerosol effect.
I’m not convinced this is true either. I believe the influence of anthropogenic aerosols is quite uncertain. The AR5 document suggests a likely value of around -0.5Wm-2, with a range from about -1.3Wm-2 to around -0.1Wm-2. However, there are some who think – I believe – that the actual aerosol forcing is quite likely more negative than the IPCC most likely estimate. If so, that would imply a higher ECS, not lower. Again, maybe I’m wrong and the GCMs are using aerosol forcings that are too large, but that was not my impression.
He continues with,
A particularly robust way of empirically estimating climate sensitivity is the so-called ‘energy-budget’ method, which is based on a fundamental physical law – the conservation of energy. Energy-budget best estimates of ECS fall in a range between 1.5°C and 2.0°C (1.25–1.4°C for TCR), depending on the exact periods chosen for analysis. Using the longest available periods that were free of major volcanism gives a ECS best estimate of approximately 1.7°C (1.3°C for TCR).
Particularly robust. Really? Where does that come from? Is that some kind of objective judgement? I notice that he doesn’t mention the ECS range from Otto et al. (2013). Most of the Otto et al. (2013) estimates were around 2oC for the ECS, with a range from around 1.3oC – 3.9oC. I appreciate that there are estimates with the ECS below 2o, but if one considers the recent Trenberth & Fasullo (2013) and Cowtan & Way (2013) papers, then even ‘energy-budget’ estimates give ECS likely values around 2.5oC. Again, there is a large range, but it’s still not clear that ‘energy-budget’ estimates are particularly inconsistent with other estimates.
He also adds
Note that the CMIP5 GCMs give an estimate for the warming over the next two decades as 0.48–1.15°C.21 In the AR5-WG1 final draft, however, that estimate was reduced by 40% to 0.3–0.7°C, apparently recognising that overall the models were warming unrealistically quickly. Inconsistently, no change was made to the longer term GCM projections.
Well, I believe that this is because we are aware that there is a mismatch between models and observations. There are various explanations for this. It could be internal variability, volcanoes, anthropogenic aerosols, or – possibly – something else altogether (maybe there are fundamental issues with the models). However, this mistmatch clearly exists and so will likely influence the warming over the next two decades. It, however, doesn’t immediately imply that the long-term trend is wrong. Hence, no change to the longer term projection is not necessarily inconsistent.
He finished his submission with
If TCR really is 1.35°C then under RCP8.5 – the worst-case, business-as-usual scenario – the end of the 21st century will be approximately 2°C warmer than today.
The meta-analysis in Tol (2009)22, of fourteen estimates from economists, suggests that a temperature of 2°C warmer than today is likely to have a negligible impact on welfare.
Indeed, if the TCR is 1.35oC, the end of the 21st century would be about 2oC warmer than today. However, that will make it higher than at any time in human history and means that we will have locked in at least a further 1oC of warming. I discussed Tol’s meta-analysis in an earlier post (although – as usual – the comments may be more informative than the post itself). I would argue that Nic Lewis’s statement is, even based on Tol (2009), incorrect. Tol (2009) suggest that 2oC warmer than today is when there will likely be no net benefit (roughtly 50% chance of benefit and 50% chance of damage). I wouldn’t say that a 50% chance of the impact being negative is consistent with likely negligible impact on welfare.
There is, however, also a typo in Tol (2009). The paper indicates that Hope (2005) was +0.9, rather than -0.9. When one corrects for this typo it not only reduces, slightly, the warming for which we will see benefits, but also makes it very clear that the possibility of a net benefit is based very strongly on a single piece of work (Tol 2002).
Additionally, many argue that acting against global warming will harm the poor who will lose access to cheap energy. What those who promote Tol’s (2009) meta-analysis fail to point out is that even this analysis acknowledges that- in a warmer world – most of the benefits will go to the developed world (who contributed most to global warming) while the poorer parts of the world will likely suffer. This seems like a moral issue that is at least worth considering, not something we should be ignoring altogether.
Anyway, that’s all I was going to say. There probably is more that could be said, but I’ll leave it at that. Apologies if the beginning seemed a bit snarky. Hard not to be when someone with the publication record of an average PhD student thinks that there are issues with the IPCC because they didn’t take his work more seriously (or, at least, that’s how it comes across). As usual, comments welcome, especially if they can clarify the bits about which I’m uncertain (most of it really).