Dr. Roy Spencer’s Ill Considered Comments on Citizen Science

Guest Post by Willis Eschenbach

Over at Roy Spencer’s usually excellent blog, Roy has published what could be called a hatchet job on “citizen climate scientists” in general and me in particular.  Now, Dr. Roy has long been a hero of mine, because of all his excellent scientific work … which is why his attack mystifies me.  Maybe he simply had a bad day and I was the focus of frustration, we all have days like that. Anthony tells me he can’t answer half of the email he gets some days, Dr. Roy apparently gets quite a lot of mail too, asking for comment.

Dr. Roy posted a number of uncited and unreferenced claims in his essay. So, I thought I’d give him the chance to provide data and citations to back up those claims. He opens with this graphic:

Dr. Roy, the citizen climate scientists are the ones who have made the overwhelming majority of the gains in the struggle against rampant climate alarmism. It is people like Steve McIntyre and Anthony Watts and Donna LaFramboise and myself and Joanne Nova and Warwick Hughes and the late John Daly, citizen climate scientists all, who did the work that your fellow mainstream climate scientists either neglected or refused to do. You should be showering us with thanks for doing the work your peers didn’t get done, not speciously claiming that we are likeable idiots like Homer Simpson.

Dr. Roy begins his text by saying:

I’ve been asked to comment on Willis Eschenbach’s recent analysis of CERES radiative budget data (e.g., here). Willis likes to analyze data, which I applaud. But sometimes Willis gives the impression that his analysis of the data (or his climate regulation theory) is original, which is far from the case.

Hundreds of researchers have devoted their careers to understanding the climate system, including analyzing data from the ERBE and CERES satellite missions that measure the Earth’s radiative energy budget. Those data have been sliced and diced every which way, including being compared to surface temperatures (as Willis recently did).

So, Roy’s claim seems to be that my work couldn’t possibly be original, because all conceivable analyses of the data have already been done. Now that’s a curious claim in any case … but in this case, somehow, he seems to have omitted the links to the work he says antedates mine.

When someone starts making unreferenced, uncited, unsupported accusations about me like that, there’s only one thing to say … Where’s the beef? Where’s the study? Where’s the data?

In fact, I know of no one who has done a number of the things that I’ve done with the CERES data. If Dr. Roy thinks so, then he needs to provide evidence of that. He needs to show, for example, that someone has analyzed the data in this fashion:

Now, I’ve never seen any such graphic. I freely admit, as I have before, that maybe the analysis has been done some time in the past, and my research hasn’t turned it up. I did find two studies that were kind of similar, but nothing like that graph above. Dr. Roy certainly  seems to think such an analysis leading to such a graphic exists … if so, I suggest that before he starts slamming me with accusations, he needs to cite the previous graphic that he claims that my graphic is merely repeating.

I say this for two reasons. In addition to it being regular scientific practice to cite your sources, it is common courtesy not to accuse a man of doing something without providing data to back it up.

And finally, if someone has done any of my analyses before, I want to know so I can save myself some time … if the work’s been done, I’m not interested in repeating it. So I ask Dr. Roy: which study have I missed out on that has shown what my graphic above shows?

Dr. Roy then goes on to claim that my ideas about thunderstorms regulating the global climate are not new because of the famous Ramanathan and Collins 1991 paper called “Thermodynamic regulation of ocean warming by cirrus clouds deduced from observations of the 1987 El Niño”. Dr. Roy says:

I’ve previously commented on Willis’ thermostat hypothesis of climate system regulation, which Willis never mentioned was originally put forth by Ramanathan and Collins in a 1991 Nature article.

Well … no, it wasn’t “put forth” in R&C 1991, not even close. Since Dr. Roy didn’t provide a link to the article he accuses me of “never mentioning”, I’ll remedy that, it’s here.

Unfortunately, either Dr. Roy doesn’t fully understand what R&C 1991 said, or he doesn’t fully understand what I’ve said. This is the Ramanathan and Collins hypothesis as expressed in their abstract:

Observations made during the 1987 El Niño show that in the upper range of sea surface temperatures, the greenhouse effect increases with surface temperature at a rate which exceeds the rate at which radiation is being emitted from the surface. In response to this ‘super greenhouse effect’, highly reflective cirrus clouds are produced which act like a thermostat, shielding the ocean from solar radiation. The regulatory effect of these cirrus clouds may limit sea surface temperatures to less than 305K.

Why didn’t I mention R&C 1991 with respect to my hypothesis? Well … because it’s very different from my hypothesis, root and branch.

•  Their hypothesis was that cirrus clouds act as a thermostat to regulate maximum temperatures in the “Pacific Warm Pool” via a highly localized “super greenhouse effect”.

•  My hypothesis is that thunderstorms act all over the planet as natural emergent air conditioning units, which form over local surface hot spots and (along with other emergent phenomena) cool the surface and regulate the global temperature.

In addition, I fear that Dr. Roy hasn’t done his own research on this particular matter. A quick look on Google shows that I have commented on R&C 1991 before. Back in 2012, in response to Dr. Roy’s same claim (but made by someone else), I wrote:

I disagree that the analysis of thunderstorms as a governing mechanism has been “extensively examined in the literature”. It has scarcely been discussed in the literature at all. The thermostatic mechanism discussed by Ramanathan is quite different from the one I have proposed. In 1991, Ramanathan and Collins said that the albedos of deep convective clouds in the tropics limited the SST … but as far as I know, they didn’t discuss the idea of thunderstorms as a governing mechanism at all.

And regarding the Pacific Warm Pool, I also quoted the Abstract of R&C1991 in this my post on Argo and the Ocean Temperature Maximum. So somebody’s not searching here before making claims …

In any case, I leave it to the reader to decide whether my hypothesis, that emergent phenomena like thunderstorms regulate the climate, was “originally put forth” in the R&C 1991 Nature paper about cirrus clouds, or not …

Finally, Dr. Roy closes with this plea:

Anyway, I applaud Willis, who is a sharp guy, for trying. But now I am asking him (and others): read up on what has been done first, then add to it. Or, show why what was done previously came to the wrong conclusion, or analyzed the data wrong.

That’s what I work at doing.

But don’t assume you have anything new unless you first do some searching of the literature on the subject. True, some of the literature is paywalled. Sorry, I didn’t make the rules. And I agree, if research was public-funded, it should also be made publicly available.

First, let me say that I agree with all parts of that plea. I do my best to find out what’s been done before, among other reasons in order to save me time repeating past work.

However, many of my ideas are indeed novel, as are my methods of analysis. I’m the only person I know of, for example, to do graphic cluster analysis on temperature proxies (see “Kill It With Fire“). Now, has someone actually done that kind of analysis before? Not that I’ve seen, but if there is, I’m happy to find that out—it ups the odds that I’m on the right track when that happens. I have no problem with acknowledging past work—as I noted above,  I have previously cited the very R&C 1991 study that Dr. Roy accuses me of ignoring.

Dr. Roy has not given me any examples of other people doing the kind of analysis of the CERES data that I’m doing. All he’s given are claims that someone somewhere did some unspecified thing that he claims I said I thought I’d done first. Oh, plus he’s pointed at, but not linked to, Ramanathan & Collins 1991, which doesn’t have anything to do with my hypothesis.

So all we have are his unsupported claims that my work is not novel.

And you know what? Dr. Roy may well be right. My work may not be novel. Wouldn’t be the first time I’ve been wrong … but without specific examples, he is just handwaving. All I ask is that he shows this with proper citations.

Dr. Roy goes on to say:

But cloud feedback is a hard enough subject without muddying the waters further. Yes, clouds cool the climate system on average (they raise the planetary albedo, so they reduce solar input into the climate system). But how clouds will change due to warming (cloud feedback) could be another matter entirely. Don’t conflate the two.

I ask Dr. Roy to please note the title of my graphic above. It shows how the the clouds actually change due to warming. I have not conflated the two in the slightest, and your accusation that I have done so is just like your other accusations—it lacks specifics. Exactly what did I say that makes you think I’m conflating the two? Dr. Roy, I ask of you the simple thing I ask of everyone—if you object to something that I say, please QUOTE MY WORDS, so we can all see what you are talking about.

Dr. Roy continues:

For instance, let’s say “global warming” occurs, which should then increase surface evaporation, leading to more convective overturning of the atmosphere and precipitation. But if you increase clouds in one area with more upward motion and precipitation, you tend to decrease clouds elsewhere with sinking motion. It’s called mass continuity…you can’t have rising air in one region without sinking air elsewhere to complete the circulation. “Nature abhors a vacuum”.

Not true. For example, if thunderstorms alone are not sufficient to stop an area-wide temperature rise, a new emergent phenomenon arises. The thunderstorms will self-assemble into “squall lines”. These are long lines of massed thunderstorms, with long canyons of rising air between them. In part this happens because it allows for a more dense packing of thunderstorms, due to increased circulation efficiency. So your claim above, that an increase of clouds in one area means a decrease in another area, is strongly falsified by the emergence of squall lines.

In addition, you’ve failed to consider the timing of onset of the phenomena. A change of ten minutes in the average formation time of tropical cumulus makes a very large difference in net downwelling radiation … so yes, contrary to your claim, I’ve just listed two ways the clouds can indeed increase in one area without a decrease in another area.

So, examining how clouds and temperatures vary together locally (as Willis has done) really doesn’t tell you anything about feedbacks. Feedbacks only make sense over entire atmospheric circulation systems, which are ill-defined (except in the global average).

Mmm … well, to start with, these are not simple “feedbacks”. I say that clouds are among the emergent thermoregulatory phenomena that keep the earth’s temperature within bounds. The system acts, not as a simple feedback, but as a governor. What’s the difference?

• A simple feedback moves the result in a certain direction (positive or negative) with a fixed feedback factor. It is the value of this feedback factor that people argue about, the cloud feedback factor … I say that is meaningless, because what we’re looking at is not a feedback like that at all.
• A governor, on the other hand, uses feedback to move the result towards some set-point, by utilizing a variable feedback factor.

In short, feedback acts in one direction by a fixed amount. A governor, on the other hand, acts to restore the result to the set-point by varying the feedback. The system of emergent phenomena on the planet is a governor. It does not resemble simple feedback in the slightest.

And the size of those emergent phenomena varies from very small to very large on both spatial and temporal scales. Dust devils arise when a small area of the land gets too hot, for example. They are not a feedback, but a special emergent form which acts as an independent entity with freedom of motion. Dust devils move preferentially to the warmest nearby location, and because they are so good at cooling the earth, like all such mechanisms they have to move and evolve in order to persist. Typically they live for some seconds to minutes and then disappear. That’s an emergent phenomenon cooling the surface at the small end of the time and distance scales.

From there, the scales increase from local (cumulus clouds and thunderstorms) to area-wide (cyclones, grouping of thunderstorms into “squall lines”) to regional and multiannual (El Nino/La Nina Equator-To-Poles warm water pump) to half the planet and tens of years (Pacific Decadal Oscillation).

So I strongly dispute Dr. Roy’s idea that “feedbacks only make sense over entire atmospheric circulation systems”. To start with, they’re not feedbacks, they are emergent phenomena … and they have a huge effect on the regulation of the climate on all temporal and spatial scales.

And I also strongly dispute his claim that my hypothesis is not novel, the idea that thunderstorms and other emergent climate phenomena work in concert planet-wide to maintain the temperature of the earth within narrow bounds.

Like I said, Dr. Roy is one of my heroes, and I’m mystified by his attack on citizen scientists in general, and on me in particular. Yes, I’ve said that I thought that some of my research has been novel and original. Much of it is certainly original, in that I don’t know of anyone else who has done the work in that way, so the ideas are my own.

However, it just as certainly may not be novel. There’s nothing new under the sun. My point is that I don’t know of anyone advancing this hypothesis, the claim that emergent phenomena regulate the temperature and that forcing has little to do with it.

If Dr. Roy thinks my ideas are not new, I’m more than willing to look at any citations he brings to the table. As far as I’m concerned they would be support for my hypothesis, so I invite him to either back it up or back it off.

Best regards to all.

w.