I popped over to Bishop Hill today – something I haven’t done for a while – and noticed that Andrew (to whom I haven’t always been particularly complementary) was promoting a new paper about clouds and global warming. The paper is by John McLean,
apparently a Professor of Physics – apparently not, and is called Late Twentieth-Century Warming and Variations in Cloud Cover.
Despite being written by a physicist, there are many reasons why this particular paper should be treated with a great deal of skepticism. For example, the conclusions argue that
the reduction in total cloud cover accounts for the increase in temperature since 1987, leaving little, if any, of the temperature change to be attributed to other forcings.
So, no role for anthropogenic forcings? Really? Despite what would appear to be a fairly obvious red flag, there appears to be – so far – not a single skeptical comment on the Bishop Hill post. I know that people object to the term climate science denier, but I’m starting to think that this is simply because it’s impolite, not because it isn’t true.
Let’s take this a little further, though. The basic premise of the paper seems to be that one can use the Trenberth-like energy flow diagram to determine how changes in cloud cover influence the incoming energy flux. Clouds have a total forcing of about 79Wm-2 . The paper indicates that, between about 1980 and 2000, cloud cover reduced by 4.8%, which is then interpreted as producing a change in forcing of 5.4 Wm-2 – significantly greater than the change in any other forcing. This would seem quite remarkable if true. [Update : Victor points out in this comment that this is only the influence of clouds on albedo. They also influence the outgoing longwavelength radiation and so the net cloud forcing is actually small, and a 5% change in cloud will not produce a significant change in forcing.]
Well, clouds are actually extremely complicated. Changes in low-level clouds influence albedo, while changes in high-level cloud influence the outgoing flux. This is explained really well here and indicates that cloud feedbacks are around 0.68 Wm-2K-1, but could be as low as 0 Wm-2K-1, or as high as 1 Wm-2K-1. Given that we’ve only warmed by around 0.85oC, this is hard to reconcile with clouds producing a radiative forcing of more than 5 Wm-2 over a period of a couple of decades.
So, that seems a little odd. Let’s look at the main figure though. This shows (below) the change in surface temperature and the change in cloud cover. Looks pretty good, doesn’t it? Maybe changes in cloud cover are indeed driving surface warming. Could we test this a little more? Sure. If the primary forcing is clouds, and all others are insignificant, then we should also see a similar signature in other components of the climate system – the oceans, for example. Do the oceans show rapid warming during the period from 1980 – 2000, followed by slower warming in the 2000s? No, there’s been no slowdown in ocean warming during the 2000s. How is this possible if clouds are – by far – the dominant forcing? Answer : it isn’t. Plus, if changes in cloud cover can produce a change of forcing of greater than 5 Wm-2, we’d certainly have noticed that in the ocean heat content.
Now, could there be another explanation? Well, assuming the data isn’t garbage, what this could be illustrating is actually how cloud cover responds to changes in temperature – i.e., it’s illustrating cloud feedback (H/T to Karsten who pointed this out on Twitter). There’s probably much more that could be said about this paper, but I’ve wasted enough of my time, so I won’t bother. I’ll just point out that a paper that argues that there is little influence from greenhouse gases, and confuses clouds responding to changes in temperature, with temperature responding to changes in cloud cover, appears to have generated little skepticism on Bishop Hill. A regular theme amongst “skeptics” is that they are somehow excluded from discussions, and are not taken seriously. Well, if they accept dross like this without question, it’s maybe not that surprising. Maybe they should try to be actually skeptical, rather than only claiming to be skeptical.