Predicting the future is sometimes possible on the short time scales of weather forecasting, but sensitive dependence on initial conditions makes long-term predictions iffy.  Spotting trends, and acting to head off disasters, is one of the ways in which human intelligence improves upon mere physics.  So our stories about possible futures are not predictions so much as scenarios.

     Our attitude toward climate change has long been like that of the driver who cannot stop when his headlights illuminate an obstacle, because of going too fast.  Science is providing us with somewhat better headlights.  Indeed, it has just told us that global warming has an evil twin, that there are going to be some missing bridges on that dark road ahead, not just some bumps but some voids where the comfortable road drops out from under us.

     And it isn’t just a matter of slowing down (though that would help).  We are near the end of a warm period in any event; ice ages return even without human influences on climate.  The last warm period abruptly terminated 13,000 years after the abrupt warming that initiated it 130,000 years ago, and we’ve already gone 15,000 years from a similar warm-up starting point.  But we may be able to do something to delay an abrupt cooling.

 

Do something?  This tends to stagger the imagination, immediately conjuring up visions of terraforming on a science-fiction scale – and so some shake their heads and say, “Better to fight global warming by consuming less,” and so forth.

     Surprisingly, it may prove possible to prevent flip-flops in the climate – even by means of low-tech schemes.  Keeping the present climate from falling back into the cool-and-dry mode will in any case be a lot easier than trying to reverse such a change after it has occurred.

     Were fjord floods causing flushing to fail, because the downwelling sites were fairly close to the fjords, it is obvious that we could solve the problem.  All we would need to do is,  using highway-construction amounts of explosives, open a channel through the dam – and do it before dangerous levels of fresh water build up. 

     This works only if floods of fresh water prove to be the problem, because the downwelling sites turn out to be close enough to the fjords.  The jury’s still out on that one.  Timing could be everything, given the delayed effects from inch-per-second circulation patterns, but that, too, potentially has a low-tech solution:  build dams across the major fjord systems and hold back the meltwater at critical times.  Or divert eastern-Greenland meltwater to the less sensitive north and west coasts.

     But relying on such simple fixes presumes that you know what you’re doing.  You get to be an expert in this field only by knowing the data and knowing the processes, backward and forward.  And I mean that literally:  you have to have computer models that successfully predict the past before you’ll even think of trusting them to predict the next fifty years.  Before we become as busy as beavers, we’ll want to try out the dam schemes on the computer models.  That’s the way you find out the common mistakes, not by experimenting directly on our one and only global climate.

     Fortunately, big parallel computers have proved useful for both global climate modeling and detailed modeling of ocean circulation.  They even show the flips.  Computer models might not yet be able to predict what will happen if we tamper with downwelling sites, but this problem doesn’t seem insoluble.  We need more well-trained people, bigger computers, more coring of the ocean floor and silted-up lakes, more ships to drag instrument packages through the depths, more instrumented buoys to study critical sites in detail, more satellites measuring regional variations in the sea surface, and similarly for studying the atmosphere.  Eventually you’d progress to some small-scale trial runs of interventions.

An expert is someone who knows some of the worst mistakes that can be made in his subject, and how to avoid them. - Werner Heisenberg, 1971

     It would be especially nice to see another dozen major groups of scientists doing climate simulations, discovering the intervention mistakes as quickly as possible and learning from them.  Medieval cathedral builders learned from their design mistakes over the centuries, and their undertakings were a far larger drain on the economic resources and people power of their day than anything yet discussed for stabilizing the climate in the twenty-first century.  We may not have centuries to spare, but any economy in which two percent of the population produces all the food, as is the case in the United States today, has lots of resources and many options for reordering priorities.

 

Futurists have learned to bracket the future with alternative scenarios, each of which captures important features that cluster together, each of which is compact enough to be seen as a narrative on a human scale.  Three scenarios for the next climatic phase might be called population crash, cheap fix, and muddling through.

     The population-crash scenario is surely the most appalling.  Plummeting crop yields will cause some powerful countries to try to take over their neighbors or distant lands – if only because their armies, unpaid and lacking food, will go marauding, both at home and across the borders.  The better-organized countries will attempt to use their armies, before they fall apart entirely, to take over countries with significant remaining resources, driving out or starving their inhabitants if not using modern weapons to accomplish the same end: eliminating competitors for the remaining food.

     This will be a worldwide problem – and could easily lead to a Third World War – but Europe's vulnerability is particularly easy to analyze.  The last abrupt cooling, the Younger Dryas, drastically altered Europe's climate as far east as Ukraine.  Present-day Europe has more than 650 million people.  It has excellent soils, and largely grows its own food.  It could no longer do so if it lost the extra warming from the North Atlantic.

     There is another part of the world with the same good soil, within the same latitudinal band, which we can use for a quick comparison.  Canada lacks Europe's winter warmth and rainfall; it has, for example, no equivalent of the North Atlantic Current to preheat its eastbound weather systems.  Canada's agriculture supports about 28 million people.  If Europe had weather like Canada's, it could feed only one out of twenty-three present-day Europeans.

     Any abrupt switch in climate would also disrupt food-supply routes.  The only reason that two percent of our population can feed the other 98 percent is that we have a well-developed system of transportation and middlemen – but it is not very robust.  The system allows for large urban populations in the best of times, but not in the case of widespread disruptions.

     Natural disasters such as hurricanes and earthquakes are less troubling than abrupt coolings for two reasons: they're brief (the recovery period starts the next day) and they're local or regional (unaffected citizens can come to the assistance of the overwhelmed).  There is, increasingly, international cooperation in response to catastrophe – but no country is going to be able to rely on a stored agricultural surplus for even a year, and any country will be reluctant to give away part of its surplus.

     In an abrupt cooling the problem would get worse for decades, and much of the earth would be affected.  A meteor strike that killed most of the population in a month would not be as serious as an abrupt cooling that eventually killed just as many.  With the population crash spread out over a decade, there would be ample opportunity for civilization's institutions to be torn apart and for hatreds to build, as armies tried to grab remaining resources simply to feed the people in their own countries.  The effects of an abrupt cold last for centuries.  This might not be the end of Homo sapiens – written knowledge and elementary education might well endure – but the world after such a population crash would certainly be full of despotic governments that hated their neighbors because of recent atrocities.  Recovery would be very slow.

 

A cheap-fix scenario, such as building or bombing a dam, presumes that we know enough to prevent trouble, or to nip a developing problem in the bud.  But just as vaccines and antibiotics presume much knowledge about diseases, their climatic equivalents presume much knowledge about oceans, atmospheres, and past climates.  Suppose we had reports that winter salt flushing was confined to certain areas, that abrupt shifts in the past were associated with localized flushing failures, and that one computer model after another suggested a solution that was likely to work even under a wide range of weather extremes.  A quick fix, such as bombing ice dams, might then be possible.

     Although I don't consider this scenario to be the most likely one, it is possible that solutions could turn out to be cheap and easy, and that another abrupt cooling isn't inevitable.  Fatalism, in other words, might well be foolish.

 

A muddle-through scenario assumes, again, that we would mobilize our scientific and technological resources well in advance of any abrupt cooling problem, but that the solution wouldn't be simple.  Instead we would try one thing after another, creating a patchwork of solutions that might hold for another few decades, allowing the search for a better stabilizing mechanism to continue.

     We might, for example, anchor bargeloads of evaporation-enhancing surfactants upwind from critical downwelling sites, letting winds spread them over the ocean surface all winter, just to ensure later flushing.  We might try to create a rain shadow, seeding clouds so that they drop their unsalted water well upwind of a given year's critical flushing sites.     Perhaps computer simulations will tell us that the only robust solutions are those that re-create the ocean currents of three million years ago, before the Isthmus of Panama closed off the express route for excess-salt disposal.  Thus we might someday dig a wide sea-level Panama Canal in stages, carefully managing the changeover.

 

 Scenarios ought to capture the stage setting that precedes the main events, not merely describe major outcomes.  So let me spin a slightly more exaggerated version of our present know-something-do-nothing state of affairs:  Know Nothing, Do Nothing.

     My scenario doesn’t require the short-sighted to be in charge, only for them to have enough influence to create starvation budgets for the relevant science agencies, to send recommendations back for yet another commission report five years hence, and so forth.  In the USA, all it takes is either a know-nothing President or a do-nothing Congress.

     The short-sighted have, after all, dominated at many times during history.  Book burning and censorship are perhaps the most obvious signs, but countries have also withdrawn into shells, disdaining foreign inventions and knowledge in favor of fundamentalist visions of the good life (recall that Chinese retreat in 1433 from the route to Europe if you can’t think of modern examples).