It is time to stop quivering in our boots in pointless fear of the future and just roll up our sleeves and build it.
- Ray Pierrehumbert

Thursday, May 13, 2010

How Deep is this Result?

There are deep results in the climate sciences. By "deep result" I mean something that would take some years of study to understand or place into context. This is an intrinsically interesting branch of physical science. (Indeed, if there were no policy issue it would probably be better funded and attract a larger number of highly talented people than it does now. But let's avoid that topic for now.)

William Stein is an accomplished mathematician and a the leader of a large open source project (that scientific programmers in the audience would do well to look into, SAGE, a replacement for Maple, Mathematica, Matlab and "Magma", the last being more of a niche product that especially interests Bill.) I had the pleasure of having dinner in his company along with various other very smart people last Monday, several of us among the climate-obsessed. Bill, however, being a very focussed person, knew very little about it. Accordingly he asked a couple of remarkably insightful newbie questions. The one I remember clearly was "how deep of a result is it?"

The response we came up with first was a bunch of caveats. Of course we cannot "prove" anything in the sense a mathematician might. Once we got that across, I pondered the question.

It emerged that Bill's question could be quantified as follows. Suppose you have a non-ideological person with a bachelor's degree in a serious mathematical science. How long would it take to complete the argument?

I quickly came to the point where I asserted that, if "it" means "greenhouse gas accumulation presents us with serious consequences over the next century" it was difficult to think of any intelligent, open-minded graduate student in the climate sciences who didn't find the proposition obvious after they had attained to a master's degree level. "So, two years then?" Bill asked? I said that was excessive: these people are studying particular phenomenologies, not the "is global warming scary" question.

So I've reached an upper bound of a few months. And I think that's roughly right. I'd think a single course (full-time equivalent of a month, possibly represented by David Archer's book "Global Warming: Understanding the Forecast") should suffice. A Ph.D. level person ought to be able to read a book like that as if it were a novel, say a few evening's concentrated work. Would that be convincing? A couple of online lectures, say Alley's and Hansen's, to ice the cake, and yes, I'd say the picture could be conveyed to a scientifically educated person in a couple of weeks of serious concentration, perhaps just a few days if that person were familiar with some of the underlying physics.

Steve pointed out that the deniers could monkey-wrench any individual argument, since we are doing earth science and not mathematics. That said, their are a number of convergent arguments, and little evidence to the contrary. And of course, denying the science is insufficient to derail the policy. It is necessary for the zero-policy position to prove a low sensitivity, but not enough to question the consensus sensitivity. So a week isn't really enough to manage the minefields of debate. In fact, time constrained "debate" as construed by many people is never really a good way to examine a broad scientific context, though one could defend an individual result that way.

But "global warming is scary" isn't a scientific result in the ordinary sense.

From a strictly philosophical point of view, "scary" isn't an objective standard. Some real-world metric for "scariness" is needed, and therein we get to difficult territory. But it's hard to compare the scope of the changes we expect with the scope we are used to and be sanguine about it.

A more serious issue is that "scariness" isn't a result in that it isn't a straightforward conclusion from laws and data to a consequence. It's a "balance of evidence" argument that leads to an estimate of how big a deal this all is.

Usually the number we use is "sensitivity", i.e., the amount of warming associated with a doubling of CO2 (or equivalent radiative forcing). That number appears to be in the neighborhood of 3 C, and that in turn leads us to expect 1) detectable changes by 1990 2) noticeable changes by 2010 and under business-as-usual 3) disruptive changes by 2050 and 4) catastrophic changes by 2100. Also, failing a scheme for removing CO2, likely a more expensive prospect than not emitting it in the first place, these disruptions will be very long-lived.

So this article is my long answer to Bill's question. I also came up with a short answer, as Steve said, "the outlines of the argument" that could be understood in a few minutes and was intended to be maximally persuasive in a few minutes. That will be another article.

This article, I think, would give any fair-minded scientist enough material to at least understand the perspective of mainstream climatologists in the absence of deliberate obfuscation (*).

* - The obfuscation, alas, is there. That is a problem we have to deal with, and how we deal with it in addressing an energetic and intelligent person who is also fully occupied with their own projects and interests is another matter. It's an important one, though, and we can't rely on PR people or politicians to do it for us.


David B. Benson said...

Here is a slightly longer than short argument, Global Warming, Decade by Decade in which the first formula is missing a right parenthesis and ought to read
AE(d) = k(lnCO2(d-1) - lnCO2(1870s)) - GTA(1880s)
and for the doubters, blow them away by citing Tol, R.S.J. and A.F. de Vos (1998), ‘A Bayesian Statistical Analysis of the Enhanced Greenhouse Effect’, Climatic Change, 38, 87-112, (h/t to James Annan).

That won't do for the scary part. For that, read Mark Lynas's "Six Degrees".

S said...

You have more interesting dinners than I do.

Michael Tobis said...

S, well, not as often as I'd like, but that's sort of the point of doing science.

It isn't the money.

Aaron said...

The problem is getting people to discriminate between real science and the nonsense that is being tossed out, and called “science”. Cutting between the two requires a very sharp and well honed chisel. We get the science right, but often say it badly. They say things with emotional appeal but offer bad science.

Most people do not have a framework of good science to hang climate science on. Thus, they tend to depend on how well the lesson is presented, to decide whether they will accept that bit as real science. Then, one good “pitch man” with the right voice tones and eye contact can undo all of your teaching.

Or people go to the internet. Denialists on the internet learned their craft selling tobacco when most climate scientists were still just kids. Those guys know how to tell a story so people believe them. That is their job and they are good at it. When was the last time you met a climate scientists that had spent the last 30 years learning to be a better salesman?

Mal Adapted said...

Excellent post, Michael. The topic of just how much a person needs to know, to apprehend the reality of AGW, comes up all the time in the blogosphere. As you point out, it depends on just how the question is posed, and on the motivation of the person asking the question.

You say "the picture [of the serious consequences presented by GHG accumulation over the next century] could be conveyed to a scientifically educated person in a couple of weeks of serious concentration, perhaps just a few days if that person were familiar with some of the underlying physics."

I agree that a "reasonable" (i.e. non-ideological, non-Dunning-Kruger-afflicted) person familiar with the underlying physics will readily grasp this much:

- GHG accumulation means warming is happening, and will continue;
- A large fraction of the accumulating GHGs are anthropogenic;
- Positive feedbacks will amplify warming beyond that attributable to the primary forcing gases;
- AGW will progress for centuries, even if anthropogenic GHG emissions cease immediately.

I'm not so sure that an understanding of basic physics, however, will lead to recognition of the consequences of AGW, for both human society and for biodiversity. I think that requires a knowledge of both economics and biology, and especially ecology, that goes beyond "a bachelor's degree in a serious mathematical science."

The reason is that complex phenomena like climate and ecosytems are at least weakly emergent from simple physical processes. The character of much of the debate leads me to think that General Systems Theory isn't included in typical undergraduate science curricula. I didn't discover GST until the second year of my Master's program in Environmental Science, an expressly inter-disciplinary field. GST is indispensable for understanding how AGW leads to economic and ecological consequences. If it were routinely introduced in high-school or undergraduate science training, deniers would have less less to gain by monkey-wrenching any individual argument.

Michael Tobis said...

Um. Well, therein lies a tale.

While I've definitely been influenced by many of the people mentioned in the Wikipedia article, I'm very dubious of von Bertallanfy, who I think was just a shallow imitator of Norbert Wiener. I think Wiener's dalliance with this crowd did not serve him well and actually contributed to the trivialization and eventual disappearance of cybernetics as a serious discipline.

Wiener's legacy is everywhere in the hard sciences and engineering disciplines, but his name and achievements remain associated with wild handwaving and premature generalizations.

I think that broad analogies across disciplines can be extremely useful, but I don't think that there are universal principles that can be applied to all systems. There are universal principles of continuous linear systems, which is what Wiener studied, but he was too easy to convince that the analogies to other systems would be easy to find.

This is the tragedy of his forgotten legacy.

Bateson and Mead were marvelous and brilliant human beings but they fell into physics envy and contributed to the destruction of Wiener's reputation as a significant figure in modern science.

I wish I had time and skill to write a biography of Wiener from the point of view of someone who actually understands his substantive achievements. There are several biographies of him but they uniformly fall into the same trap WIener did.

Anyway, to Mal's point, I don't think you can teach generalizations until you understand some instances. It's a very common didactic mistake, and one which I am very prone to myself. You need examples. The deeper the generalization, the broader the set of instances the student needs to have experienced.

This all said, I would very much like to hear how Mal has been exposed to GST and how it affects his thinking. How mathematical is your exposure? DId the name Lyapunov come up?

Please contact me via email or reply here. I am more than a little interested in how this material was presented to you, and by whom.

Hank Roberts said...

How long did it take Bill to read Spencer Weart's history, if he did?

(If he hasn't, is there a better place to start than that?)

Michael Tobis said...

No, as far as I know he hasn't spent any time on this. The question was hypothetical: if he were to spend time, how much time would that be?

Mal Adapted said...

Whew! Michael, I probably should have linked to Systems Thinking rather than to formal Systems Theory. My exposure to systems thinking, in my MS-level Ecology coursework, was not very mathematical at all, nor was there much discussion of its history; rather, it was presented as a useful framework for thinking about complex natural systems, as a hierarchy of simpler ones interacting with each other to generate emergent phenomena. For me, it's a recognition that ontologically, the Universe isn't divided into departments, but needs to be understood as a dynamic whole. That's in contrast to the kind of stove-pipe thinking that can lead to Gerlich and Tscheuschner 2009.

From my experience, the general inability to see how basic physical processes can give rise not just to climate, but to ecosystems and economies, explains why some denialist arguments, especially of the "what's the optimal temperature?" and "warming won't be so bad" categories, have been as successful as they have.

Michael Tobis said...

Well, G&T is just incorrect, "stovepipe" or otherwise.

Sorry, you hit a half-buried obsession of mine, obviously.

The holist/reductionist dichotomy shouldn't be a debate, I think. Clear thinking requires an array of tools; each tool and approach has its place.

At root, though, sustainability requires holistic thinking that is also rigorous. We don't have much experience with that; macroeconomics seems deeply flawed and nothing else seems to make the attempt. The old "general systems theory" doesn;t do it for me either.

It's time for an Earth System Science, more or less as Francis Bretherton proposed, but it's not an easy question how to get there.

Mal Adapted said...

"The holist/reductionist dichotomy shouldn't be a debate, I think. Clear thinking requires an array of tools; each tool and approach has its place."

I absolutely agree, Michael. I also agree that everyone needs a broad knowledge of Earth System Science, to have the deep understanding required for a transition to sustainability. I have no more clue than anyone else how to get there.

Kooiti MASUDA said...

Unfortunately for us, from a viewpoint of obfusicators it is not necessary for the zero policy position to prove a low sensitivity, but it is enough making a public impression that all science about climate sensitivity is as junky as their soldiers'.

Kooiti MASUDA said...

Excuse me for previously responding to a peripheral part of the subject.

I think that it is a robust knowledge that the steady state response to CO2 doubling is likely to be between 1.5 and 6 degrees C. And it is enough to warrant policy action. I think that it is important for us to convince many people who have literacy in basic physics of this.

On the other hand, it is desirable to reduce the uncertainty because of both reasons intrinsic to science and policy relevance. It is likely to require "deep" understanding on one hand, and somewhat expensive scientific enterprises on the other hand.

The both example issues of ventilated thermocline and moist convection will considerably (though not definitely) be clarified by high resolution numerical modeling. And high performance compting is surely big (=expensive) science, but not so big among big science. Still it may be a science-techology-sociey problem that everyone cannot afford to reproduce the experiments. It is a larger problem how to supply observational data to check reality.

Hank Roberts said...

> The question was hypothetical:
> if he were to spend time, how
> much time would that be?

Okay, I guess I'd time him on reading Weart's online book (more comprehensive than the paper one); add the time to look up whatever puzzled him in that (remembering he can email the author); then add, oh, 50-100 percent more time to that.

Reading and comprehension are the critical things there; someone who stops at every statement and says Oh Noes Not the IPCC will run the meter way up and never learn. But anyone who's been known to read outside his field browsing say in Science and Nature for curiousity likely will do fine.

So my guess is the number will be in hours, somewhere between 8 and 24 hours, all told.

I've occasionally been called hopeful ....

Hank Roberts said...

So, ah, you opening a betting pool on this?

I've noticed more appearances of older academic types who have just started looking into climate blogs, starting very skeptically, but who have good credibility with their colleagues and, if not offended and insulted right off the bat, can get past the loaded terms that set some of the anklebiters off too quickly (I may resemble that remark myself at times).

So I'm gonna stick with my estimate. Love to see results.

William Stein said...


Thanks for pointing out Spencer Weart's book, which I'll read as I find time. I very much appreciate that Weart's book is made freely available online.

-- William (the person who is the subject of this blog post).