Sunday, February 12, 2012
About models and math, semiotics and stories, global warming and sequestration, and my next book (Part I of a series)
I've been doing one kind or another of modeling and simulation, and a lot of different things with statistics, for money, for more than thirty years. (Never enough of it, by the way, so if you want to talk about what modeling and stats can do for a business, you can look at my blogs in TheCMOSite and AllAnalytics, and there's an e-mail link here.) Much of my future-building and world-building comes out of that.
The truth is, many times, I write stories set in a world that I think is interesting because it retro-finances all the time I spent jacking around with the models and stats. I take comfort in the fact that Tolkien often made up the languages first and only then got around to using them in stories. (Unfortunately I can't quite get Bob Dylan's voice singing I said "You know they refused Jesus too?" and he said, "You ain't him." out of my head here).
Back in 1991, the nice guys at NOAA released their hurricane models (astonishingly crude by present standards), and I played around with them on a Microsoft Excel 3.0 spreadsheet on a Mac SE, figuring out ways around the limitations of that nifty little gadget, and leading eventually to my writing Mother of Storms and having a whole lot of interesting adventures along the way.
These last few years I've had an acute interest in carbon sequestration, which is an interesting side-branch for those of you with an interest in the whole atmospheric-carbon/climate change issue.
If the term is unfamiliar—I'm surprised at how many people I know who seem not to have heard it—the idea is that if atmospheric carbon is an issue in keeping the planet fit to live on, then in addition to reducing how much carbon we release into the atmosphere, we (as in humanity) could also take some carbon out of the air, in a large variety of ways I'll talk about in more detail later, and then stuff the carbon someplace where it won't come back for a while, i.e. sequester it. There's already some controversy in just calling it "carbon sequestration." I say carbon instead of carbon dioxide-and-methane because in both cases it's the carbon that causes all the trouble, and there's no reason I can see to waste resources sequestering perfectly good oxygen or hydrogen. And "for a while" is a highly contentious issue because "a long enough while" is what defines sequestration. Almost everyone would say holding your breath is too short a return-to-atmosphere time to count as sequestration and would agree that injecting scavenged carbon beneath advancing tectonic plates (about 100 million years to return to atmosphere) is long enough to call sequestration. Just about everything between is contested.
I've written a couple of short stories about strange things that happen because the world tries carbon sequestration in one form or another – "Rod Rapid and His Electric Chair" (available: Amazon, B&N, direct from me) and "The Birds and the Bees and the Gasoline Trees" which still has a little while to run exclusively in Engineering Infinity but will be joining the short story collection pretty soon. Those first little games with the idea have been essentially Frankenstein stories (even though one* point of "Gasoline Trees" is that monstrousness is how you make it and what you make of it); those are a natural kind of story to tell about any new technology, and might even sometimes be truthful. We do often fail to foresee, and sometimes we think we're acting for high rational purposes but we're actually calling up Forbidden Planet's "monsters from the id."**
And although I have weeks to go on the current novel, since long experience has taught me that having a "next" to anticipate is essential to finishing, I'm doing some shopping for a next, and have some ideas for a novel that involves carbon sequestration, so a few days ago I started playing around with it, and this led to some thoughts I found interesting enough to blog about.
Getting the nuisance part out of the way right now, since people seem to require declarations of position before they decide to read on, I suppose I am now supposed to either say that I "believe in" human-caused global warming, or conversely that I am a "skeptic" (or "denier" if you play for the other team.***). My actual view will probably give no satisfaction to anyone, which is why I think it's probably a good starting basis for a longer work of fiction.
As a modeler who has looked at several of the models, I have to say that some of the most prominent global warming models are very well done, and that the less adroit ones have taken a pounding in peer review and appear to be losing influence. There are people who object to the whole idea that we can learn anything from computer modeling—they should stay out of airplanes; how do you suppose they are designed nowadays? There are people who think that only a perfectly specified model with every number and function in it nailed to nine nines can teach us anything, and they should stop listening to weather forecasts and give up a large number of medical innovations since about 1995. There are people who get worked up about tautology – in a reasonable sense every computer model is a repeated tautology, and I refer you to Chaim Perelman's The Realm of Rhetoric for a review on why the classical view of tautology does not hold up any better (or worse) than classical physics has. But as a modeler, I accept the essential value of working with a good model, even where data and understanding are incomplete, and the published models look like more than good-enough models to me; I've billed people for worse work and collected without shame.
As a statistician, I am compelled to say that the statistical basis for global warming is shaky because the data are necessarily shaky. There will never be enough non-shaky data within the time we have to make any decisions, because:
1. the good stuff doesn't go back very far and the older stuff has an unknown but large number of problems, but
2. it's the data we've got; many, many times I've been called in to extract information from a bungled study or poll, and that is a statistician's life; if you ski moguls, don't complain that the ride is bumpy,
3. it is by no means clear that better data was ever possible, in two senses:
a. the limits of the instruments and their positioning; if Arrhenius had convinced people back in 1896 to monitor global climate systematically, it is by no means clear that having many more recording thermometers, barometers, and air samplers world wide would have given us better data, only more of it. (As poll analysts frequently mutter, asking 1200 people the wrong question does not improve the result from asking 100 people the wrong question).
b. There are mathematical limits imposed by the way in which the poorly-sampled huge domain of overall global temperatures, gas concentrations, and insolation has to be projected onto the range of an immense array of individual measurements. (If I just lost you in the math words: I recently blogged over at All Analytics about what domain and range are, and a friend told me the piece was "almost painless." I want that on my next book jacket). Just about any effect smaller than whole continents going up in flames tomorrow is going to produce very fuzzy, messy, ambiguous data, and statisticians don't like that. (By contrast, environmental factors that affect human mortality, conditions that affect traffic accidents, and news stories that affect voting, offer easily sampled tight domains that project very nicely into compact, explicable ranges – which is why stats people like working with those, and sometimes acquire the view that all data should be like that or not analyzed at all, rather like a child who will only eat mac and cheese because s/he knows exactly how it tastes and how to eat it with a spoon).
So as a statistics person, I look at global warming and say, "Wow, now there's a mess to sort out. Glad I'm not the one that has to." But the world has never been arranged for statisticians' convenience*****, and very often important issues force you to work with crummy data, at which point you suck it up and do your best. (And expect that there will be some quibbling with whatever you did, so don't do anything you'd quibble with yourself about; that way at least one person will always approve of your work).
A slightly heretical admission for one who deals so much in quants: Sometimes anecdotal evidence is better than the statistical. As a church member, I've seen pencil and paper surveys elicit an overall response of, "It's all so great! Don't change a thing!" simultaneously with one usually-super-positive key member muttering angrily in the corner, and the lone angry mutterer was in fact a much better reflection of where we were. In the business world, I've seen sweeping decisions based on one encounter with one customer, ignoring all the carefully analyzed data, and sometimes that works. (Knowing when that will work, I guess, is what makes that particular businessman the success he is). Sometimes an anecdote is so awful or so wonderful that it weighs very heavily, and this is not necessarily a bad thing.
Anecdotes, in my experience, tend to be at their best and most valuable when they come form people who have long, intimate experience with the subject matter (e.g. knowing that the one angry person is critical to how the church functions, or that one customer has articulated an opportunity that was never even considered in the questionnaire). And much of the anecdotal evidence that the world is getting warmer is of this kind; a guy who has gardened the same back yard for 40 years finds he can plant three weeks earlier and harvest two weeks later. A serious fisherman for twenty years, who keeps a fishing diary, sees his water temps going up, slowly and haltingly, but up. People who lived for years with sea ice right up to the shore barely seeing any on the horizon. Photographs of glaciers then and now, fruit growers who spend much less on smudge pots, hunters who see warm-climate birds further north and south or later in the year, and the number who seem to be seeing it – this impresses me quite a bit. (There would actually be a way to aggregate all that via statistical semiotics, for a few million dollars, if anyone wanted to hire me). Of course exceptions crop up because statistically you expect them to. In a realigning election, there's always somebody's Uncle Jack who moved in the opposite direction; some three-pack-a-day smokers make it to 95 before a bus hits them; that's the nature of stochastic behavior. "The race is not always to the swift, nor the battle to the strong, but that's the way to bet." But anecdotal evidence in quantity is data (just messy, hard to process data) despite the claims of some people's sigs.
As a statistical semiotician, specifically, and to some extent as a fiction writer, I see something more tangled to unravel: a collection of stories into which the (admittedly cruddy) data is fit, and in the long run that is probably more significant. Although we hope our guiding stories are based on facts, it is the story, and not the facts, to which people respond at the deepest level. (Some of you are now thinking of the last minutes of The Man Who Shot Liberty Valance and those of you who are not should see it sometime soon, as it will make you a better person). There's an internationalist narrative about how finally there was a global menace so big that everyone had to give up nationhood, and learn to prefer a world run by well-educated suits filling out complex forms. There's a narrative of hubris where human beings are finally learning that nature always kicks back harder so we've got to learn to stop kicking it (and hubris is great story material). There's a narrative of the Jimmy Carter/puritanical sort in which it's just better and more moral and gosh-darn nicer for people to wear natural-fiber sweaters, eat whole grains, shower once a week, and so forth, and global warming is just the latest reason for everyone to be mildly uncomfortable in a virtuous sort of way. Over on the side that doesn't believe the world is getting warmer, there's a narrative in which the internationalists are going to confiscate all private business and order every business person to hire a Muslim transvestite who will spend all her time haranguing him, or the same vast conspiracy of evil scientists that is always out there taking over the world has just dreamed up global warming to get power.
I haven't actually seen a story that attracts me on any side of the question. Maybe I just don't like sweaters or I have a hard time picturing the climate scientists I know as Leninist agents, but there's no narrative that I want to sign on for, so far. That's a plus, for me; I want something new for the novel, and it's going to mean going back to the facts, including some catalog of what we don't know yet and what we can never know, and building from there. So I have a pleasantly difficult task in front of me, and I'll be wandering through it in this and the next few blog posts.
So suppose we just ask, is it getting too warm, how warm is it getting, and what's to be done about it?
Well, summarizing the above, the models are well thought out, and the balance of evidence is that it's getting warmer, and that a very large number of things having to do with the wild world***** are getting worse because of it. I don't believe we can yet rule out a possibility that processes connected to solar variation play a larger role and human-generated gases a smaller one than is being estimated, but while that might be highly relevant if we were holding a trial ("Your honor, I demand that we release the human race immediately and prosecute the sun!"), it is much less so if we are contemplating action or even purposive inaction.
This is because I'm more than willing to bet my planet that atmospheric carbon is at least a temperature increase amplifier. (If the music is too loud, I know it's likely a combination of the singer and the amp, but if I can turn down the amp, that's what I do).
We don't have to have created/emitted any of the extra heat trapping gases (likely though it is that we did)—they're still undesirable to have around. They mean more rapid heating no matter where the change in heat starts or comes from. If by any chance we arein a long term low solar cycle, this only changes the amount of time we have to work with, not what needs doing (though it may give us a wider or narrower range of options for exactly how to do what needs doing). Minima are followed by maxima, and whether the next solar maximum is in a few years or a hundred, it will be a bad idea to have the gases around, the equivalent of having a big leaky gasoline tank next to your house in a dry forest; the tank may not start the fire but if one starts you'll wish you didn't have it. Even if the solar minimum is low enough to significantly cool the Earth, we don't want the eventual recovery to be a massive backlash; a faster temperature rise is likely to result in more chaos and species loss, and a very fast one could trigger massive weather events. (Did I happen to mention Mother of Storms is coming out again with a new cover this year?)
So, as I said before, carbon sequestration is on my mind, and I don't find the stock narratives about global warming very interesting. I'm not going to find any new ones that I do like immediately, knowing how my mind actually works, but one way that works for me to stir the pot, prepare the road, water the soil, and mix the metaphor is to try to get some sense of scale and proportion and what must matter, the unignorables of the fictional world that is trying to be born.
For example, the model that led to A Million Open Doors and the other Giraut books suggested a universe of unimaginable affluence, where literally the poorest people had more of the good things in life than Bill Gates could have today, but also one where most people lived in concrete boxes.
The Century Next Door model, a very different kind, pointed me toward one of the central artistic interests in that series: how close the extraordinarily wonderful lies to the equally extraordinary horrible. St. Francis and Genghis Khan were almost exact contemporaries; so were Einstein and Stalin, and Henry the Navigator and Gilles de Rais. Caligula and Christ could have met as adults and neither was very long-lived.
And of course Mother of Storms was sort of a long meditation about how things move and transition from what state to another, and what it means for the system to have more to process than it can,with greater and greater volumes of information, money, and matter moving faster and faster. (I suppose I could have titled it A Hymn to Bandwidth or maybe The Year of Turbulence.)
So where I'm going, tomorrow I hope because I must get to other things today, is to walk you through one of my modeling exercises, which I built to see if the commonly given numbers for human-produced carbon make sense (they do), how things look for the next hundred years in light of that model (bad but not impossibly bad), and the what and why that might go into carbon sequestration (because, fundamentally, I'm not a particularly virtuous planetary citizen; gigantic bizarre engineering projects with huge unforeseen consequences are fun to read about, and that's what I'm after. Those who wish a sermon on our evil wasteful culture, or a denunciation of the manipulating one-worlders, can doubtless find it elsewhere).
So besides whatever intrinsic interest there may be in the model, or in the creative process, at least tomorrow we'll be talking about screwing around on a planetary scale, and making bigger messes than history has provided so far. (Always assuming I get that part done tomorrow).
*any decent story not only has more than one point, it has more points than any one reader or the writer is aware of, and contains some number of contradictory points as well. If you can say the point of the story, meaningfully, the story is not worth the bother. You may quote me to your English teacher, or to either Rationality Guy or Theme Girl at your writer's workshop, but I will not beat them up for you. Especially not that English teacher, as they are vicious, cunning, and treacherous on the mat.
**Now that we apparently don't have ids anymore, where are we supposed to go when we need a good monster?
*** There's a useful distinction between denier and skeptic that is made by perhaps 1% of everyone involved, so we'll skip it here, but you can find it discussed in thoughtful and not-so-thoughtful places all over the web.
**** Nor for the convenience of waiters, house painters, undertakers, mothers, traveling salespeople, or shepherds. The few rich people I know well assure me that it wasn't arranged for their convenience either, but I think they may be lying.
*****"oh, for god's sake, Barnes, why the wild world and not just wilderness?" I hear an impatient reader say. Because wildness happens all over the place all the time, and intertangles with our lives, and we are blessed by it; wilderness is grand stuff but most of us can have only very limited encounters with it. The wild world exists where anything is wild – the fox you see crossing the road in suburban dawn, the frogs and turtles in the irrigation canal, the complex web of animals and plants in a farmer's hedgerow or a borrow pit, the unbuildable ravine where a coyote lays up for the day. The wild world is in vacant lots, in the city treetops, possibly behind the wainscoting if you've got mice. And it blesses you when you see it, whether it's as quiet as a prairie dog watching from his hole in a vacant lot, or as spectacular as a peregrine picking off a pigeon in a public square. And it seems to me that the wild world is threatened at least as much by global temperature rise as true wilderness; when the rising temperature drives the dissolved oxygen too low for walleye in a big recreational reservoir, we're not losing wilderness, but we're losing a source of wildness in our lives. If there's less snow cover in city park, my chances of seeing an owl hunting there by moonlight are reduced. Admittedly, these are selfish pleasures, but what natural pleasures aren't? Who else experiences the pleasure besides the self?