Cloud Feedbacks

I think the problem here is just that Spencer isn't being very precise with his wording.  My guess is that the point he's making is that the models show the net feedback from clouds is positive.  Clouds have both negative (reflecting) and positive (heat trapping) effects (depending on the type of cloud), so the question is how large each of these effects is, and thus what the net feedback is.  The models used in the IPCC certainly account for both, and my understanding is that Spencer is correct (if this is what he's saying) that the calculated net effect is always positive.  He's trying to argue that the models are all wrong and it's actually negative, which I strongly doubt.

The only peer reviewed article that Spencer has published on the subject of cloud feedbacks was not criticizing the IPCC, but rather a single paper from 2005. Basically, Spencer's entire article in Journal of Climate is one long criticism of one of the methods used in this paper, a paper which doesn't seem to be particularly influential in the first place (it was based on a mere 11-year run of a computer model, for instance.) 

Plus, even with my shallow knowledge of climate modeling, it seems there is a lot of room for error in Spencer's "admittedly simple" climate model. He assumes anthropogenic radiative forcing is 0, for instance, and so he only has 2 variables that he is changing: heat transfer in the oceans, and some "unknown" forcing, which could stand for anything but which he seems to assume is cloud cover. I don't quite understand the details, but it seems he just tinkers with these two variables until they match the temperature data (now where have I heard that technique criticized before) and ultimately spits out a small negative number which he concludes must be the true feedback of the entire climate system.

It's funny how skeptics criticize the methods of other scientists who make climate models, yet have no reserves about accepting Spencer's "very simple" models as pure fact.

Oh and on a side note, I noticed that the author of this paper he was criticizing, Forster, was mentioned 4 separate times in article you referenced to, 3 times in this one, and probably many other places on Spencer's blog. Makes me wonder who that guy is and why Spencer is so obsessed with calling him out.

Quote:

Originally Posted by dawei 

 

He assumes anthropogenic radiative forcing is 0, for instance, and so he only has 2 variables that he is changing: heat transfer in the oceans, and some "unknown" forcing, which could stand for anything but which he seems to assume is cloud cover. 

Interesting - that's the same tactic he used in another entry on his website, where he was trying to determine how much of the recent warming could be attributed to the PDO.  Basically he set up a few crude models, one of which had PDO as the only forcing.  The end result was that he concluded in theory, the PDO could account for a bit less than half the recent warming - if no other radiative forcings existed and the PDO was hugely amplified, or something along those lines. 

It seemed to be a pretty pointless exercise to me.  I don't understand why you would make completely unrealistic assumptions like setting the anthropogenic radiative forcing to zero.  It seems like Spencer is just playing games, which is probably why RealClimate hasn't devoted any time to discussing his recent feedbacks paper.

That does seem to be his strategy though.  From the paper linked by dawei:

Why you would do that, I don't know.  It doesn't make sense to me.

There was another paper written by a Japanese physicist I've seen referenced a couple of times.  Basically what he did was assume the global temperature increase in the first half of the 20th century was purely natural (in reality, there was a significant anthropogenic component as high as 50%) and then concluded that based on his calculations, warming in the latter part of the century could have been natural.  In short, by assuming the warming was natural, he concluded the warming was natural.  It seems to me like Spencer is doing something similar.

Quote:

Originally Posted by dana1981 

That does seem to be his strategy though.  From the paper linked by dawei:

 

 

Why you would do that, I don't know.  It doesn't make sense to me.

 

There was another paper written by a Japanese physicist I've seen referenced a couple of times.  Basically what he did was assume the global temperature increase in the first half of the 20th century was purely natural (in reality, there was a significant anthropogenic component as high as 50%) and then concluded that based on his calculations, warming in the latter part of the century could have been natural.  In short, by assuming the warming was natural, he concluded the warming was natural.  It seems to me like Spencer is doing something similar.

My take on Spencer is that he is clued in enough on the physics to understand that he can't argue anthropogenic CO2 will have no effect on climate.  Therefore, he argues it has a small effect.  This is what Christy argues as well.  For CO2 to have a small effect, the climate sensitivity to changes in the radiative forcing have to be small.  The problem is that some of the feedbacks aren't well known, and to calculate the total climate sensitivity you have to know all the feedbacks.  This means that there is some wiggle room for guys like Christy and Spencer to exploit. 

By fiddling with parameters in models to exploit peculiarities in satellite datasets, Spencer is able to show that the need feedback from clouds *has* to be negative, implying that the overall sensitivity of climate is less than the IPCC estimates.  However, the problem for Spencer, which nobody really bothers to call him on because it's too obvious, is that there are independent estimates of the climate sensitivity available from the paleo record (where they can correlate increases in global temperature with increases in GHGs and then, using a model, derive an overall sensitivity).  If you put Spencer's negative feedbacks from clouds into those models, you can't simulate paleo climate. 

This means one of two things.  The first is that there is some unknown process operating in the paleo climate that gives a large positive feedback to compensate for the large negative feedback from clouds so that you can get the observed sensitivity.  Spencer ignores this problem because he hasn't a clue what that process would be, and he has the additional problem that he not only has to propose a mechanism for paleo climate but that mechanism has to be inoperative in the modern situation (or he is right back where he started with a larger climate sensitivity (am I making sense so far?))  Alternatively, you could conclude that the paleo climate sensitivities are correct, the processes are fairly well known, the total feedback from clouds is positive (or not very negative), and Spencer is fudging things again. 

Yeah that's how I understood it.  Hansen for example is convinced the climate sensitivity will result in a 3 +/- 0.5 deg C warming for a doubling of CO2 based on paleoclimate data.  Other climate scientists aren't convinced the climate will respond exactly as it did in the past, but you have to invoke some very different effect for the sensitivity now to be much lower.  That's why I don't buy Spencer's arguments.  He doesn't explain what's different, just kind of assumes that something is because he wants it to be.

Nothing to do with deniers - most people claim or deny a correlation depending upon their preconceptions or needs at the time.

Nothing against beating on a denier but this error is far more common than just with the deniers. 

Well they're not exactly a consistent group of people.  Don't forget that Spencer's work is based on models, and extremely simple ones at that.  Deniers are always saying that climate models are worthless, but once Spencer uses one, it's golden.  And the temperature record is unreliable, unless it appears to show cooling.  And....well, I could go on forever listing denier contradictions, but you get the point.  It's what we've come to expect.