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, February 19, 2009

New York City

Wherein our hero indulges himself in an actual blog entry...

So while it was all hitting the fan in Austin, I find myself in New York City where people don't know what picante sauce is supposed to taste like, mostly for the  purpose of talking about applying the web to the problem of elicitation of expert opinion and formal development of collective wisdom with Paul Baer. This is entirely off budget and outside my formal job description as far as U Texas is concerned, though perhaps it shouldn't be. (I took vacation time.) 

Today I had an interesting conversations with Paul and his colleague Sivan Kartha, who gave me quite a bit more respectful a hearing about the prospects and limitations of climate modeling than I normally get. 

It turns out that a well-known climatologist of our mutual acquaintance has said "We're done", meaning that physical climatology has essentially nothing more to contribute to the discussion of global change. I said that barring a substantial change of direction which I could imagine but could not envision getting funding for, that I agree. (I exclude geochemists, who have a lot more work to do on the carbon cycles. We also talked about glaciologists a bit. I think the glaciologists are doing fascinating science, but I don't think they'll answer the motivating policy questions effectively. Even though the field is suddenly data rich doesn't mean the question "when will what melt" is sufficiently constrained.) 

Climate change policy is not about climatology anymore, as I've said here on more than one occasion, and all efforts to make it so are efforts to divert from the much harder questions of what to do about it. All this conversation took place at an amazing and bustling pizza/pasta/pannini/patisserie near Penn Station. I'm grateful to my companions for letting me expound on my opinions, which I have had few chances to express in as much detail.

Then we all took the subway up to Columbia where Paul and Sivan had a meeting with the Nobel prize winning economist Joe Stiglitz who is a professor there (and who was giving a public lecture that evening). Remarkably, my Twitter feed had made me aware of a panel the very same day at Columbia, at which Coturnix/Bora Zivkovic was one of the panelists. So I had an excellent opportunity to keep myself occupied while Paul was busy.

While I was familiar with Bora's work on the PLoS journals, the other two panelists were new to me. I was very entertained by the idea of treating DNA sequences "like one of those old Texas Instruments TTL chip catalogs" I may have been the only person in the audience who knew exactly what Barry Canton meant by that. A certain shade of bright orange-yellow was activated in my memory. (It certainly revived my confidence that such if such a thing is possible for living beings, it certainly should be possible for a computational model of physics, contemporary boondoggles notwithstanding.) But that's not the direction I seem to be heading. 

I was much more interested in Jean-Claude Bradley's description of "open notebook" science, which ties in both with the work I am doing with Paul and with some of the ideas I have for extending Sergey Fomel et al's Madagascar package. In open notebook science, every little step you take is instantly public.  Bradley showed how this sort of collaboration can advance chemical science; both contemporary work and ancient out-of-copyright texts can be imported into a vast and multilayered structure. This seems to me the a substantial step in the real world into the sort of science envisioned back in the 1940s by Vannevar Bush. I also see applications of their approach in the world of collective intelligence, but here it might pay me to be a tiny bit closed for awhile.

I also had a brief conversation with the charming Miriam Gordon, who was very taken with my mention of Marshall McCluhan in the context of new media and my question about the future of peer review. (The panel really threw up their hands with a vague hope that peer review might somehow get better someday in the distant future.)

I would be remiss, though, if I didn't mention how much I enjoyed Bora's talk. He's obviously a fine writer, but I didn't expect what an extraordinarily engaging speaker he is. Though there was little in his presentation that was new to me, I enjoyed it thoroughly and highly recommend you give Bora a listen if you ever have the opportunity.

I can't begin to express how much more productive this day was than my day job in Texas. It's not just that I had more fun. I meanwhile added more value to the world. (I captured essentially none of the value, though Paul is covering my expenses through a small grant of his.) I have been miscast and I don't know quite what to do about it. I think I might convince Paul to coauthor some papers to help get me into policy space for real. Cause y'know, blogging "doesn't count".

I do know that the answer has something to do with the internet. I would never have met Bora, or Paul, or for that matter Damon Gambuto, Clifford Johnson, Sheril Kirshenbaum, Chris Mooney, John Mashey, Erik Conway, John Fleck,  Julia Hargreaves, James Annan and and even (indirectly through Paul) Michael Mastrandrea and Stephen Schneider were it not for this blog, meetings which easily constitute some of the most interesting and encouraging moments of the past two years. (I hope to meet more of y'all in future!)

So then I walked directly from this remarkable event to the next building, home of the business school at Columbia, to meet up again with Paul and to listen to Joe Stiglitz talk about the Great Unwinding. While, to be sure, he was in many ways far more sophisticated than I, it reamined, like most talk by most economists, profoundly unsatisfying to me, in that "growth" was treated as normal and non-growth as pathological. This whole way of thinking still seems divorced from reality. 

In any case, Stiglitz made what seemed to be a strong case for nationalizing banks (ironic that capitalism itself will be the first business to be socialized) a strong case for Keynesian expenditures, and a strong case against rescuing shareholders and executives of banks and other companies that are struggling. Those latter he considered "looting". He adamantly opposes any sort of government action where the government is left holding the bag in case of failure but gains nothing in case of success. That, he said, "is not capitalism". 

He also allowed that things would not go back to "normal" until "demand" reasserted itself, and he could not see how that was going to happen. He believes that the consequences of the recent financial failures will be very long-lasting. 

Isn't it bizarre? Nothing was burned or flooded. How do such things happen? Why is paper more important than physical reality? Anyway, it all seemed so weird to me. Not only did the word "climate" not pass his lips, but not even the word "energy" entered the picture for Stiglitz. Sometimes I wonder if most economists live on a different planet from the rest of us. 

In any case, there is little question that I have to be here in the land of bad salsa (try Ollie's Noodles for some easy on the wallet and the palate consolation. Yumm!) burning vacation time under the terms of Tim O'Reilly's rule of productivity: "Work on stuff that matters". It sure would be nice if the stuff that matters didn't actually compete with the stuff that pays, though. I hope to have more to report to you soon.

13 comments:

Mitchell said...

I would think that aerosols and the Arctic remain important topics. We don't know the magnitude or the timing of the Arctic threat, and it seems likely to me that managing anthropogenic aerosols with attention to their cooling influence is going to be part of future climate policy, with or without geoengineering.

Bunty said...

"Sometimes I wonder if most economists live on a different planet from the rest of us."

Legend has it that they used to.

The planet we now know as Venus. ;D

gravityloss said...

Financial crisis... Stupid name really.

People took too much debt they couldn't pay. They bought all kinds of fancy things with that debt.

I guess the loaners don't have use for that fancy stuff like huge houses etc.

So in a way the effort was wasted in building them.

On the other hand, if the price of houses dropped, who now has the money, who sold those houses at the high point?

It's just a transfer of money from the past home buyers (who got it from the banks) to the past home sellers.

So somebody must have gotten rich from the bubble...

Probably the building contractors as well.

I see nobody cheering them.

It's also interesting that they don't seem to be buying stuff ie that "lesser growth", lacking better words is occurring.

Maybe it's just that the previous growth was based on spending savings (money but ultimately labored products and raw materials) and that stash has now run dry.

So people can't spend at the previous level anymore.

Miriam Gordon said...

Michael,

It was a pleasure to meet you as well. Its so great to connect with people who can discuss things with intelligence and flair, such as you and Bora. I am currently enjoying reading all of your sites. I really hope we will keep in touch.

Aaron said...

Sometime in the next decade there will be a Lake Missoula like event on Greenland resulting in a “sea level rise event”. Nine billion people will have an “Oh! Shit!” moment. A very few of us will shrug and say ”MCP”. In the following hours, there will be a lot of funding for modeling ice, and better GCM that includes ice sheets. Then, and only then will there be a chance to build FASTER, SMARTER, CHEAPER models. Of course, then it will be a lot easier because we will have a real live example of how fast it goes and people will not be saying, “Oh! It could not possibly happen that fast!”

Physical climatology is not done, it need to rethink its roles. We need risk assessments and cost analysis for ALL of the impacts of climate change. Good RA and CA need numbers a thousand times better than what we have. Only then, can we estimate how much we should tax carbon. Only with good RA and CA can we plan a cap and trade system. To say that Physical climatology is done, is to say that policy makers should just pull those programs out of their asses without any decision support or justification. We only have time for one whack at this problem. We need to go after the problem with every tool that we have. We must get it correct the first time, and that will take real data, prepared without reticence. Until we have very good numbers on which to base policy, and assure ourselves that our policies are in fact working, physical climatology is not done.

David B. Benson said...

I immediately thought of aerosols, but I see Mitchell beat me in mentioning the matter.

Michael Tobis said...

It's not outside the realm of possibility that ice sheet modeling will make sufficient progress to constrain the behavior of ice sheets effectively. It is certainly worth a try.

Carefully targeted expenditures on science can be effective, but you cannot hire nine women to make a baby in a month. Intellectual progress can reach some maximum rate but then it reaches a point where more manpower and more funding is just redundant.

Some problems in earth science are undecidable. We may never understand the ocean circulation of the Eocene, much though we might want to.

My guess is that the most likely outcome is that there will be several viable scenarios for the West Antarctic Ice Sheet until one of them occurs. Maybe Greenland will turn out to be a tad less capricious.

As for aerosols, (and as for clouds, and so on) yes improvements can perhaps narrow the uncertainty of climate prediction a bit, and I'm happy to be helping out in that regard, but the chances of the first order picture changing very much are slim.

The real issues are environmental, agricultural and civil engineering problems, and the response issues are in the social, political, economic and geopolitical realms.

Climatology is a worthy pursuit in itself, and if geoengineering is necessary you will need to rely on huge advancement in the field. Possibly we can improve our abilities for local and regional predictions, which would add a lot of value. So by all means support climatology, but don't look to us for input into what needs doing now on the mitigation front. We have said our piece and it is unlikely to change, not because we are stubborn, but because there are some things we understand pretty well.

naught101 said...

Scientists going into policy = good.

Michael, on economics, you should check out "Debunking Economics" by Steve Keen. He's an Australian economist, and the book completely ignores the problems with the basic assumptions of neo-classical economics, and concentrates on explaining the internal logical flaws of the theories - which are many, and serious. It's pretty mind blowing how far separated from reality most economist are.

guthrie said...

Naught101, I found Keen's blog a while ago, whilst trawling for online information to backup what I have read in various books such as "The Grip of death" by Michael Rowbotham, and also had confirmed by a friend who studied history of economics. The url is:
http://www.debtdeflation.com/blogs/

Suffice to say, far too many economists don't look at history or the history of their topic, and take up positions based upon ideology rather than a rational exploration of what has happened and does happen in the economic realm.

Hank Roberts said...

> It sure would be nice if the stuff
> that matters didn't actually
> compete with the stuff that pays,
> though.

Those of us grinding through work lives at the clerical level and trying to be of use to the world in our copious spare time can but agree.

EliRabett said...

In NY you want to find a good pizza shop, not Tex-Mex

rcram said...

The climate models are not close to correct yet. The Petr Chylek paper published in 2007 shows the cooling impact estimated for aerosols is way too high. Climate models have not made the necessary adjustments yet.

Also, Spencer and his team wrote a paper in 2007 regarding the negative feedback over the tropics they identified as the Infrared Iris effect hypothesized by Lindzen. Climate models still are not calculating in this negative feedback.

Climate models still do not get clouds correct or understand what causes clouds to form. Is it galactic cosmic rays? (may or may not have a long term trend) Or is it the PDO which has a 30 year oscillating trendline? Or is it a combination of both? The climate models have no clue.

Orrin Pilkey at Duke University wrote a book about how climate models (and he respects the work of the IPCC) will never have predictive power. It's called "Useless Arithmetic." You should read it sometime.

Michael Tobis said...

rcram's doubts about climate science are excessive in my opinion. If he were sincere these doubts would push him to support vigorous emissions restraints.

His choice of what to believe and what not to believe seems not based on a grounding in the discipline but in a preference for certain results, though. Accordingly one wonders about his strategies for evaluating evidence.

Regardless, rcram's way of treating science that agrees with his inclinations as demonstrated and science which disagrees with his inclinations as repudiated discourages conversation.

The proposition that extant GCMs have no predictive value is demonstrably false.

GCMs of the late 1980s correctly predicted very unusual aspects of the large scale behavior atmosphere over the past including polar amplification and stratospheric cooling. Transient models since then appear to be on track regarding global temperature trajectories and expansion of the arid subtropical zones.

What does that tell us about the role of modeling, and of physical climatology in general, in future governance? That's what I'm trying to address.

The position I'm trying to sketch out isn't all that simple. I'm certainly in favor of more work in climatology.

I think climatology as a purely theoretical discipline could be of great importance in itself. It seems to me that Lorenz proved that. I'll try to sketch this out sometime; I've believed it for a long time but I don't think I've expressed myself clearly on the subject as yet.

As an applied discipline, though, it is reaching practical constraints that will at best take some considerable time and effort to overcome.

In the long run such work may yet serve to inform regional adaptation. Those advocating a geoengineering fallback had best support a very vigorous discipline of physical climatology.

Perhaps progress of that sort is impractical. It's a plausible outcome. This doesn't mean climate science itself should be shut down, but it would mean that climate science has reached its limits as an input into policy.

At least for the present, I believe that we have said our piece. It will not change much for a generation.

It's daunting enough if you believe us, and worse yet if you don't.