Proposing A Non-Cynical Red Team Exercise

(Guest posting by Michael Tobis. Opinions expressed here are mt’s only.)

Some people who dismiss the climate change issue like to call it “the CAGW hypothesis” for “Catastrophic Anthropogenic Global Warming”.

The name is quadruply irritating, inasmuch as 1) a noun is not a hypothesis, 2) the lack of definition of the hypothesis allows one using the name to rule specific claims in or out for purely rhetorical utility  3) it conflates several ideas and 4) it injects jargon into the debate that the climate-related professions do not use, thereby obfuscating rather than clarifying.

Nevertheless, I think it is fair to say that, um, those people mean SOMETHING by CAGW. They know a CAGW proponent when they see one.

It is true that the consensus is making a claim that on first hearing seems utterly radical and implausible: that very large energetically viable fossil fuel resources must remain untapped.

But that’s not a hypothesis!

The proposition that “catastrophic climate change due to fossil fuel emissions is sufficiently likely that very large energetically viable fossil fuel resources must remain untapped” is in a sense a hypothesis. To be more precise it needs a precise definition of catastrophe. I’ll propose “mortality-driven decline in global population” as a definition of catastrophe.

This leaves us with proposition P: “Mortality-driven decline in human population attributable to large anthropogenic climate change is sufficiently likely to justify urgent and stringent efforts to eliminate fossil fuel energy long before reserves are depleted.” 

That’s a mouthful, and while I’d prefer to call this by some other name, I’m willing to concede the rhetorical point at least for present purposes. I’ll refer to P as the Precise Catastrophic Anthropogenic Global Warming Hypothesis, or PCAGWH.

(It’s pronounced exactly as it’s spelled. I am a genius with acronyms.)

PCAGWH is in a Popperian sense a testable hypothesis. One can fail to restrain fossil fuel consumption and see if catastrophe (as defined above) ensues. The trouble with this test is obvious – the consequences of the test are catastrophic. Avoiding that test is exactly what we are arguing for.

Because the policy we are arguing for is so extreme and unusual, though, the desire for a thorough and responsible review of PCAGWH is a reasonable goal. This brings us to the “red team” proposal.

While many criticisms can be directed at the current administration of the US, which is proposing the exercise, most are not relevant to the idea that PCAGWH should be tested in some other way than simply ignoring it and waiting to see if catastrophe ensues. Regardless of what else you may think about its proponents, this claim is on its face more than reasonable.

Some criticisms of the administration, however, are very relevant.

The administration has expressed vocal hostility and suspicion of advocates of the hypothesis, and its congressional allies have habitually called upon a few credentialed scientists of dubious skill and questionable objectivity to justify such hostility and suspicion. In this context it’s hard for the very community that has been the subject of attacks based on  systematic misattribution and outright fabrication to participate with enthusiasm.

Indeed, more of the same is quite likely to be the intent – fringe characters will hurl accusations as though they were prosecuting attorneys, and legitimate scientists will be forced into a defensive stance rather than allowed to construct a coherent case.

Nevertheless, refusal to participate encourages the view that we are arguing from baseless authority. People inclined to disbelieve the argument for our surprising and, if wrong, very expensive proposition, will have their misapprehension reinforced. This we do not have an ethical right to allow, even if the motivation of those proposing the exercise is entirely disingenuous.

===

Again, testing PCAGWH directly amounts to taking no action until global catastrophe emerges, which is exactly what the proposed policy intends to prevent. So since we would rather not test PCAGWH directly let’s look at the reasoning that supports it.

Here’s my initial attempt to simplify it to five key points.

1) Global sensitivity to greenhouse gases is unlikely to be much below 2 degrees C per doubling with a best estimate around 3 C

2) Economically valuable (in an unchanged regulatory environment) fossil fuel reserves provide enough fuel for at least two doublings of CO2 background

3) Consequent changes in excess of 2 C over less than a few centuries are very rapid compared to earth’s past, except possibly in a couple of extremely catastrophic extinction events

4) Economic, social and environmental losses climb rapidly and nonlinearly with temperature change , and may already overwhelm the short-term benefits of fossil fuels, and will very likely do so in the near future. UPDATE for clarification: Costs of a unit of emission aggregated over time may already overwhelm the benefits, which appear immediately. This balance shifts further against the value of emissions as geologically rapid climate change proceeds.

5) Environmental concentration of CO2 emissions is to a good approximation cumulative. CO2 does not naturally vanish from the system on time scales of policy interest.

The conclusion is that, consequently CO2 emissions must cease as quickly as practicable.

Since I’m just a guy on a laptop, I may have missed some points or expressed some poorly. I welcome  suggestions for improvement.

I’d say that each of the points is very robust, each extremely unlikely to be false, and together constitute a strong basis for PCAGWH.

===

It’s not a “house of cards” argument, either. Each of these claims is robust, with small but perhaps measurable uncertainty., We are in a risk management scenario. As long as each of these points is as likely as not, we are left in a situation of unacceptable risk.

I’d argue they’re all well established, but I also would like to emphasize point is they don’t have to be to justify a policy response.

Now, when the administration proposes a “red team” exercise, they are not doing so because they care about the state of science itself. They are not comparably interested in herpetology or particle physics. Clearly, their interest is in the policy implications of the science.

So, arguments ad hominem, arguments about specific statistical nits or personalities, arguments about specific excessive claims by overzealous individuals, arguments that this or that person has been ill-treated by this or that community or institution, all the nitpicking about this or that graph are beside the point. The first thing we must expect of a fair red team exercise is that it will eschew all the marginal noise that exists mostly to reflect on the morality of the community of relevant scientists (justifiably or not!).

A legitimate red team exercise would look at the component arguments and establish whether any of them is so grossly unlikely to be true as to overturn the claim that fossil fuels must remain untapped.

===

Insofar as the red team is examining the role of climate science in the PCAGW hypothesis, this limits questions of interest specifically to some narrow topics. In my own formulation, these would be #1 and #3:

1) Global sensitivity to greenhouse gases is unlikely to be much below 2 degrees C per doubling with a best estimate around 3 C

3) Consequent changes in excess of 2 C over less than a few centuries are very rapid compared to earth’s past, except possibly in a couple of extremely catastrophic extinction events.

===

These are consensus claims in climate science. Each of these is quite amenable to a fair-minded red-team exercise.

My interest would be in a fair assessment of point 1 which addresses physical climatology, but point 3 (paleoclimatology) is also important and I’ll address it briefly.

I think point relatively 3 is easy to address, insofar as it is a question to which expertise in geological science can be applied. One could easily assemble a red team from branches of the earth science that are not closely involved in climate who would be competent to address the evidence. The proposition that the realistically anticipated changes would be very large compared to events in the natural history of the planet can be fairly evaluated by people without a dog in the hunt, beyond the desire to get the answer right.

===

Let me get back to the physical climatology question, point 1.

What all this is leading up to is a challenge I’ve meant to address to those claiming that climate science is weak and that climate modeling doesn’t provide any evidence worthy of note.

I strongly believe, to the contrary, that general circulation modeling, which is the common core of climate models and weather models, is an extraordinary triumph of science.

This is because GCMs produce model weather which is sufficiently and so sustainably analogous to the actual weather of the world that a realistic climate emerges. Climate is not an input to theses systems. (They are unlike economists’ Integrated Assessment Models in this way!) What the models are told about the world is the constitutive small-scale phenomenology, and realistic approximations to boundary conditions and forcing. Storms and dry zones form in the models in the right places. The models explicitly represent the radiative transfer of the planet among the small-scale phenomenology. So the greenhouse effect is itself not explicit – it’s an emergent property of the radiative transfer.

There’s a lot to argue about in this effort, and I have been close enough to the field that I can attest that healthy and skeptical arguments do occur constantly and vigorously.

It’s important to understand that climatology exists as an independent discipline, studying many phenomena outside the domain of policy interest, and these models are crucial tools in that effort. But there’s no denying that projection of future climate change is a key objective, and that this application presents a number of interesting challenges.

But for the purposes of testing PCAGWH, the issue at hand boils down simply to this: if the sensitivity is low enough not to worry about, why is there not a single working CGCM that produces a respectable climate simulation with such a low sensitivity?

===

It’s not as if the petroleum industry lacks the skill or resources to produce such a model if it were possible. Nor, obviously, do they lack the motivation. Indeed, I am acquainted with a PhD climate modeler who worked for many years for a major petroleum corporation in their oil exploration department, modeling the carboniferous climate to identify places where oil deposits might have formed! That is, at least one oil company had climate modeling experience in house!

Those determined to claim that we don’t need to stop emitting CO2 are being utterly illogical if they don’t assert a sensitivity to CO2 doubling well lower than the IPCC low end with confidence.

They also claim the evidence of models is negligible, in that models in their view can be tuned to produce a “desired” result. (That we don’t actually desire the result we are getting escapes their imagination altogether!)

Well, the Hungarians have a lovely word, tessek (pronounced tesh-shake), which means, by all means, please, help yourself. Sometimes it is meant politely, and sometimes sardonically. But by all means, go ahead. I’m sure somebody would provide you with CGCM code. Go fix it so it has a low sensitivity without breaking the model climate.

===

There are several reasons to think of this as a fair exercise, in a way that the usual bluster from the usual suspects would not be.

For one thing, any member of the ordinary climate modeling community could easily participate. It’s a legitimate form of research – find the lower bound for plausible CGCM Charney sensitivity. Call me up, I’d probably pitch in.

Also, the otherwise inevitable vexing question of whether Dr X or Dr Y who is now clamouring to be on the “Red Team” is or is not a legitimate source of climate expertise becomes moot. Either they can help to build a working climate model or they can’t.

This also moots the question of whether the models are in any sense “good” or “bad”. The objective of the Red Team would be only to produce a version of a CGCM that is *as good* on various metrics as extant models. If that isn’t very good, so be it, it just leaves them a lower bar to hurdle.

Finally, until and unless they meet this challenge, I think we have an objective basis to claim that the critics do not have a grip on the state of the science as embodied in the models.

===

Please note that a single model with low sensitivity would be a very long way from establishing the proposition that low sensitivity is certain enough to moot the problem. But I think the absence of such a model to date constitutes very strong evidence that a low sensitivity is not possible.

Again, the motivation, skill and resources are all available. Yet nobody, and specifically none of the oil majors, has done so.

Why not?

If they want to red team the models, I say we try very hard to help them. Anyone willing to take this challenge seriously would learn something, and they might teach us something as well.

As long as they don’t have a red team GCM, all they have is bluster and hostility. If they want to be taken seriously as a team, let them play the game.