Back at the beginning of our warm period, there was mile-high ice here at the U.S.-Canadian boundary, between Victoria and Vancouver.  That’s because the Puget Lobe was rapidly advancing only 14,000 years ago, triggered by the great episode of rapid global warming that preceded the Younger Dryas.

     When the ice sheet over the Canadian Rockies started to collapse and spread sideways, the ice came all the way down the Strait of Georgia.  It went past the Strait of Juan de Fuca and on into the southern part of Puget Sound, covering Seattle’s site up to the height of a 300-story building.  The local mountains had their own glaciers then, as they still do, but they didn’t have the major ice sheets capable of filling in between mountain ranges.  It was the surge from the north that filled up the space between the Cascade Range to the east and the Olympic Mountains to the west, lasting for over a thousand years.

     A few hundred miles to the east, another such ice advance blocked a river, creating a great meltwater lake (Lake Missoula) over northern Idaho and the western part of Montana.  The ice dam broke, of course, with a great outburst flood surging across the middle of Washington State (we call them the “scablands” because of the deep channels and plunge pools cut in the old lava).  It must have been rather loud, too.  Temporarily, the Columbia River had a flow ten times greater than all the rest of the world’s rivers combined.

     The ice dam reformed, and broke again – at least 59 times.  Just imagine the biblical flood on fast forward and stuck in a repeat loop.

     We’re descending now, and it’s a little bumpy as we cross the San Juan Islands.  Landing is always the hard part, and it’s only twenty minutes from now, as I see the Strait of Juan de Fuca off to the west, stretching out into the Pacific Ocean.  The end of our warm period isn’t as predictable as the end of an airline flight, but it, too, surely isn’t far off, if we let nature take its course.  And flip again.

     We humans have consciousness in a big way, compared to the great apes, enabling us to better evaluate risk, danger, and pain.  We pay a big price for it, too:  we can know when we can’t do anything about imminent suffering.

     But maybe we can do something about it.

 Stabilizing our flip-flopping climate is not a simple matter. We need heat in the right places, such as the Greenland Sea, and not in others right next door, such as Greenland itself. Man-made global warming is likely to achieve exactly the opposite – warming Greenland and cooling the Greenland Sea.

     A remarkable amount of specious reasoning is often encountered whenever we contemplate reducing carbon-dioxide emissions.  It’s not a simple matter of global temperature – I hope it’s now clear that no simple-minded argument in favor of more global warming (to “avoid cooling,” or “increase agricultural productivity”) should be taken seriously.  These clangers are sometimes beginners’ mistakes, but can also be confuse-the-issue propaganda designed to “buy time” for polluters.

     That increased quantities of greenhouse gases will lead to global warming is as solid a scientific prediction as can be found, but other things influence climate too, and some people try to escape confronting the consequences of our pumping more and more greenhouse gases into the atmosphere by supposing that something will come along miraculously to counteract them.  Volcanos spew sulfates, as do our own smokestacks, and these reflect some sunlight back into space, particularly over the North Atlantic and Europe.  But we can’t assume that anything like this will neatly counteract our longer-term flurry of carbon-dioxide emissions.  Only the most naïve gamblers bet against physics, and only the most irresponsible bet with their grandchildren’s resources.

     To the long list of predicted consequences of global warming – stronger storms, methane release, habitat changes, ice-sheet melting, rising seas, stronger El Niños, killer heat waves – we must now add abrupt, catastrophic coolings.  Whereas the familiar consequences of global warming will simply force expensive but gradual adjustments, the abrupt cooling and drying promoted by human-enhanced warming looks like a particularly efficient means of committing mass suicide.

     We cannot avoid trouble by merely cutting down on our present warming trend, though that’s an excellent place to start.  Paleoclimatic records reveal that any notion we may once have had that the climate will remain the same unless pollution changes it is, well, wishful thinking.  Judging from the duration of the last warm period, we are in the declining centuries of our current one.  Our goal must be to stabilize the climate in its favorable mode and ensure that enough equatorial heat continues to flow into the waters around Greenland and Norway.  The stabilized climate must have a wide “comfort zone,” and be able to survive the El Niños of the short term.  We can design for that in computer models of climate, just as architects design earthquake-resistant sky scrapers.  Implementing it might cost no more, in relative terms, than building a medieval cathedral.

     Yet we may not have centuries for acquiring wisdom, and it would be wise to compress our learning into the years immediately ahead.  We have to discover what has made the climate of the past 8,000 years relatively stable, and then figure out how to prop it up.

 Just because great climate flips can happen in response to global warming doesn’t mean that they are the most probable outcome of our current situation, what one might “forecast” (that’s one of the reasons why I’ve been careful not to “predict” a cooling in the next century).  Climate scientists have a maddening tendency to focus on “the most likely” outcome in the next century, just as economists and politicians tend to do, and I believe that is a serious mistake.

     The issue here is managing high-risk situations, not the usual stuff:  you can make a lot of mistakes when you try to extrapolate ‘business as usual.”  How much effort should be expended on a minority possibility, particularly one with a history of having occurred many times in the past?  Other than military and disaster planners, physicians, and people in the reinsurance business, not too many people are knowledgeable about high-risk management.  (I learned some only because my father was an insurance executive and I later listened to many neurosurgical conferences discuss the failure of the referring physician to think beyond “the most likely cause.”)

     Guessing the most likely diagnosis and outcome is something that any second-year medical student can often do.  Even I can do it in some areas like neurology, just from hanging around real doctors for decades:  given some symptoms and lab findings, I can sometimes say that it’s most likely disease X.  I might even be right half the time.  But that’s not good enough.  The reason you consult real experts is because they know all the common mistakes and how to avoid them. And what they do, that the second-year medical student can’t do yet, is to rule out the less likely scenarios, particularly the ones that could kill you quickly if left untreated.  Even though such possibilities are “less likely,” they’re exactly what you have to focus on.

     A homey example would be when you are awakened during the night to hear some strange gurgling noises.  The most likely source, you realize as you lie there in bed, is simply a downspout clogged with leaves.  Not a serious matter, something that can wait for a sunny weekend.  But you also know that the sounds could be coming from a ruptured hose to the washing machine, and you know what a mess a flood can make, in short order.  Even though there’s a 80 percent chance of the noise being innocuous, you crawl out of your warm bed and go check things out.

     And you see the same focus on the “less likely,” both in medical diagnosis and therapy, where the physician must often act on incomplete information because of the serious consequences of delay.  Suppose that with your symptoms and lab findings, the chances are 80 percent that you’ve got disease X, a nuisance in the long run but not catastrophic.  The trouble is, the symptoms are also consistent with another more serious disease, lymphoma, which can quickly kill and needs early treatment.  Even though the chances are only 20 percent that you’ve got lymphoma, your physician may tell you that you are going to need chemotherapy “for insurance against cancer.”  You can’t just wait to see what develops.  The possible consequences of delay are simply too great.  The physician who waits until “dead certain” of a diagnosis before starting treatment may wind up with a dead patient.

     That’s our situation with gradual warming and abrupt cooling.  It isn’t that abrupt cooling is the most probable outcome in the next century but that an atmospheric warming from any cause looks capable of triggering a loss of the warm water loop through the Labrador and Greenland Seas (the front-runner candidate for what has caused the observed global abrupt coolings of the past).

     One should not get distracted by which-came-first issues (Is the warming “our fault”?) but focus on consequences – and particularly the possible consequences of postponing action, of simply waiting to see what develops.  The failure to flush the cooled-down water from the ocean surface isn’t even the 20 percent possibility at the moment; it’s the best-understood candidate for what can trigger global abrupt cooling.  The alternative candidates should not be used to discourage preventative action on the rapidly-fatal scenario.  Promptly studying how to stabilize the North Atlantic Current ought to be high on the agenda.

     Focusing on “the most likely” outcome is a beginner’s mistake when the stakes are so high.  Climate scientists have not, heretofore, had to cope with managing high-risk situations because they’ve had few interventions to offer.  As that changes, thanks to the magnificent science now being done on climate, they’ll need some appreciation for how to manage situations described 2,500 years ago by the Hippocratic aphorism, “Life is short, the art long, opportunity fleeting, experience treacherous, judgment difficult.  The physician must be ready . . . .”

 The boom-and-bust climate cycle shows the challenges that our ancestors had to contend with.  At the beginning of the ice ages 2.5 million years ago, our ancestors were walking upright but had ape-sized brains; they were just starting to make stone tools.  By the time of our most recent ice age, we had a brain three times larger, capable of making good guesses about what’s likely to happen next, and often capable of heading off trouble.

     But I doubt that another episode of the bust-then-boom climate will jack us up another notch in brain power.  Ever since agriculture was invented, and transportation minimized isolation, natural selection has become decoupled from human evolution, at least the ways it used to operate.  We’re now smart, however, in ways that owe little to our present brain power, but rather to the accumulated experience of the people that have lived since the last ice age ended.  Education.  Writing.  Technology.  Science.  We accumulate knowledge and refine wisdom from it.

     There’s an element of “use it or lose it” here.  Use our civilization’s achievements to prevent the next whiplash, or lose much of civilization’s gains as our warm period suddenly ends with a population crash.