When the detective Dirty Harry points his huge gun at a criminal, he states that there is some doubt about whether the chamber that is currently cocked is actually loaded. The policeman then utters the movie’s famous lines: “You’ve got to ask yourself a question: ‘Do I feel lucky?’ Well, do you punk?”
In a sense, this question captures the state of debate about climate change on the cusp of the COP26 global summit in Glasgow. Quantifying uncertainty is always a crucial part of science. In the Dirty Harry example, a thoughtful criminal will first infer a one in six chance of dying (assuming one loaded chamber in a six-chamber pistol) before deciding whether fleeing is worth the risk.
Similarly, to decide whether recalcitrant governments at the forthcoming summit are unwisely pushing humanity’s luck, it’s best to focus on the well-established science whose uncertainties can be quantified. It’s reasonable to suspect the wilder speculations of doom, and it’s obviously reasonable, too, to doubt whether a particular detailed and specific prediction will be accurate. But it’s not reasonable to deny that we’re running some risks the existence of which has been well established. Gauging what can and cannot reasonably be doubted is indispensable to settling just how “lucky” we can feel. But it can be tough to glean what is sure and what is dubious in a “debate” that is mired in emotion, politics, money and sensationalism as much as it is in science.
The lurch away from science is not hard to understand. Some socialists might welcome a wholesale reworking of the business model of modern industrial civilisation which they judge, rightly or wrongly, to be a happy requirement of dealing with the climate crisis. Hair-shirted puritans, meanwhile, may enthusiastically embrace the darkest scenarios almost because of the implied need for immediate dramatic restrictions on energy production, consumption and travel. For others, including libertarians in particular, this restriction of “personal freedom” is anathema: they will thus have an ideological reason to doubt apocalyptic predictions. In such ways, and even before we get to the potential of vested personal interests that distort one’s reading of science, a person’s immediate reaction to the issue of climate change tends to become deeply entangled with their political DNA.
The debate is now reaching a new level, with the release this summer of a dire new report from the Intergovernmental Panel on Climate Change (IPCC), its first major assessment of global climate change in eight years, warning of potential disastrous consequences if carbon emissions are not curtailed in the coming decade. The UN secretary general António Guterres labelled the report “a code red for humanity,” as the new administration in the United States looks set to use the coming summit as an attempt to return to a position of leadership on the climate issue.
But just how much can Washington, or even the IPCC, be reasonably sure of? Developing sound public policy proposals in the current intellectual climate—never mind winning wide political approval for painful choices—will be challenging until we can agree on what it does, and does not, make sense to doubt.
Irreducible uncertainties
At one level, it might seem peculiar for any open-minded scientist to be talking about the limits of doubt. Because, in some respects, the scientific method is doubt: never take anything on trust, always interrogate whether you’re really sure, and stand ready to falsify and jettison anything if experiment disagrees with your predictions. And indeed, there is properly plenty of doubt, and all manner of uncertainties at the frontiers of climate science. It is one of the great hallmarks of good science that we not only acknowledge the existence of uncertainties, but also attempt to accurately quantify them. One way of thinking about climate policy is as an organised societal response to doubt. In theory, at least, governments and NGOs should be weighing various alternatives, attempting to predict what policies might produce the best outcomes, recognising that their current understanding may be incomplete, and that outcomes may differ from expectations.
During the lead-up to decisions and summits like COP26, debate and discussion can usefully explore such issues as the economic and environmental utility (or lack thereof) of carbon taxes, carbon caps, subsidies for renewable energy and zero-carbon technologies, tariffs on coal, funding for carbon capture and so on. Acknowledged uncertainty should be integral to this discussion, with costs being weighed against a range of harmful possible outcomes, together with the likelihood of different potential impacts of various
proposed interventions.
But we are unlikely to devote adequate time and resources to such potentially disruptive moves if fundamental doubt lingers about the main predicted consequences of climate change so that governments and citizens can rationalise a “wait and see” attitude. Campaigns that plant the seeds of this sort of sweeping doubt have been remarkably effective at forestalling an organised response in places like the United States. The all-important distinction to draw is between a candid admission of the existence of the sort of uncertainty that has to be grappled with in all areas of science including climate, and a cynically confected ultra-scepticism, which invites paralysis by falsely cowing us into believing we don’t know anything at all.
Now it is true that no science is ever completely settled: even Newton’s classical law of gravity, which accurately predicts most of the physical phenomenon humans can directly observe, was supplanted by the theory of general relativity. General relativity may in turn be supplanted by something like string theory. Nevertheless, we need to counter our natural tendency to jump from the reality that we don’t know everything about the universe to the conclusion that we know almost nothing. The fact that we don’t have the slightest idea what is the fundamental gravitational cause of the accelerating expansion of our universe doesn’t thwart our ability to navigate the gravitational forces in our solar system accurately enough to drop a rover on Mars precisely where it was intended to land.
In the climate context, natural scepticism is only emboldened by just how much remains uncertain at the forefront. Detailed regional predictions require complex (and sometimes interactive) modelling of the atmosphere, oceans and landmasses, as well as amassing a large number of measurements around the globe. Moreover, data about past climate events is sometimes sparse, and the conclusions we draw based on paleoclimate data invariably get refined as new numbers come in.
Take a couple of recent examples. Just as the new IPCC report was about to be unveiled, Science magazine reported that most climate models used in it projected implausibly fast warming rates. When projecting backwards to compare with paleoclimate data during the coldest point in the most recent ice age, about 20,000 years ago, the data suggested the Earth cooled by less than 6C while the models suggested almost twice as much cooling. Projecting into the future, estimates that use current data and trends suggest warming between 2.6C and 3.9C if CO2 in the atmosphere doubles compared to pre-industrial times, whereas many of the models considered by the IPCC predict warming in excess of 5C.
Similarly, a recent report from Columbia University’s Lamont-Doherty Earth Observatory and Utrecht University suggests that sea level rise in the past was less significant than previously estimated, and another study out of Utrecht this year suggested that sea level rise in the coming century is likely to be 25 per cent less than previously predicted, due to slower global ocean warming than earlier estimates suggested.
Sure things
These are real, scientific concerns that cannot be wished away. If you hear a campaigner claim that all such uncertainties are over and it is now a settled fact that there are only 12 years left to save the world, ignore them. They, and you, would do better to switch attention to the real question: whether, in the face of the evolving and sometimes imprecise predictions, there is sufficient uncertainty regarding the fundamental nature, extent and impact of climate change to justify stepping back from concerted, and potentially costly, efforts to curb CO2 production?
The answer is no. The inevitable uncertainty at the cutting edge of atmospheric science does not undermine well-understood basic predictions. To return to another gravitational analogy, letting uncertainties at the forefront of science cause us to ignore fundamental well-understood science is like using the fact that uncertainties exist about quantum gravity to justify jumping out of a seventh-storey window, just in case you might fall up rather than down.
The fundamental science that underlies climate change is now over 150 years old, and it works, having been tested by easily replicable experiments—as basic as showing how much glass bulbs filled with different gasses will heat—for over a century. Moreover, the general climate predictions based on this fundamental science agree with observations that have been made over the past 60 years.
When one has an underlying, well-founded theory whose predictions accurately correlate with known causes and observed effects, it is a good bet that one is on the right track. One can never completely rule out some weird cosmic conspiracy that reproduces, by coincidence, the exact same effects as well-established theories would predict. But the right attitude to such freak possibilities is best expressed colloquially: if it walks like a duck, and quacks like a duck, it is probably a duck.
One way of thinking about climate policy is as an organised societal response to doubt
Here are the basic physical principles and observations that underlie the reality of human-induced climate change:
(a) If energy in equals energy out, the Earth’s overall temperature will remain constant: we can measure the total “energy in” when the Sun is directly overhead, and it is 1,361 Watts per square metre. Averaged over the Earth, accounting for nighttime and regions where the Sun is at a lower angle, this total average energy input is cut by three-quarters, to a value of 340 Watts per square metre. But whatever the number, the temperature will not rise if the same amount of energy bounces back out to space as is incident on Earth.
(b) If there were no atmosphere and the Earth radiated back into space 340 Watts per square metre, then our 175-year-old understanding of thermodynamics (which specifies that all bodies radiate an amount of energy that is proportional to the fourth power of their temperature) allows us to calculate that the average surface temperature of the Earth would be about 18 degrees below 0C.
(c) The actual average temperature of the Earth is roughly a positive 15C. This is because the Earth is blanketed by an atmosphere that traps much heat energy. About 30 per cent of solar energy impinging on Earth is directly reflected back into space from the top of the atmosphere. The rest is absorbed by the atmosphere and the Earth. The Earth’s surface then re-radiates its energy back out, as infrared radiation, but before this can make its way out into space it has to pass through the atmosphere. It is the proportion of energy absorbed here which determines how much warmer our world ends up than if there were no atmosphere—and that, in turn, depends on the mix of gasses there.
(d) Increasing carbon dioxide in the atmosphere makes it less “transparent” to outgoing infrared radiation. Carbon dioxide absorbs heat energy, as does water. This was first measured in the 1860s by Irish scientist John Tyndall, who measured the amount of heat radiation that successfully transmitted through vials of different gases. En route to space, all the re-radiated energy has to pass through the upper atmosphere, where there is little water vapour, so it is substantially the CO2 there which determines how transparent it is to the outbound heat energy: more of it will mean less of the energy makes it out to space. And if less radiation is emitted outwards into space, while the amount of radiation coming in from the Sun remains the same, the equilibrium that would otherwise exist is disturbed (a departure from equilibrium called radiative forcing). There is extra heat energy which makes its way to the Earth’s surface, heating it up.
(e) The predicted “radiative forcing” due to the additional CO2 in the atmosphere is between 1.5-3 Watts per square metre over the Earth’s entire surface—and this is consistent with measurements. Eventually, even with more CO2 than before, so long as the new mix of gasses is stable, equilibrium would be restored: a hotter planet and hotter atmosphere will emit enough heat energy into space to balance the incoming radiation from the Sun. But, if CO2 continues to be pumped into the atmosphere, then additional “radiative forcing” will occur and equilibrium will not be achieved. Instead, the temperature of the Earth-atmosphere system will continue to rise.
These are the well-tested bits of physics and chemistry underlying the current science of climate change. They are confirmed by a host of observations, including satellite- and ground-based measurements of upward and downward radiation, and CO2 and temperature measurements across the globe. And the basic prediction—that measured human production of greenhouse gases between 1900 and 2020 would increase the Earth’s temperature by about 1.2C—has been validated.
The Earth’s surface is dominated by water, and the oceans govern the balance of heat across it. Here measurements are particularly telling. The world’s oceans had increased in average temperature in 2019 by 0.075C compared to the 30-year average temperature between 1980 and 2010. This may not seem like a lot, but it corresponds to the additional heat that would have been produced by setting off 3.6bn Hiroshima-level atom-bomb explosions in the ocean, about four every second, day and night, 365 days a year for the past 25 years.
It takes some time for the additional energy impinging on the Earth due to radiative forcing to be fully and uniformly distributed through the ocean’s depths. Even if greenhouse gas production stopped today, the radiative forcing that has already occurred up to now will result in an additional ocean temperature rise of almost 0.5C over the coming centuries.
Slipping below the waterline
As we consider the likely impacts of climate change, amid many uncertainties, this additional heat in the oceans is telling. It has been known for millennia that water expands when heated and, when considering sea level rise, the basic change in ocean temperature this century will produce a sea level change of at least a quarter of a metre.
And that is before we come on to sobering uncertainties regarding the melting of glaciers and icesheets. As I am writing this, it has been reported that it rained on Greenland’s snowy summit for the first time in recorded history. Over the past decade, the melting of Greenland’s ice sheet has persisted at a mostly constant, and sometimes accelerating, rate. Even conservative estimates of this melting, along with the melting of glaciers in western Antarctica, when combined with sea level rise due to increasing ocean temperatures imply a minimum total sea level rise of about 0.5 metres during this century, even if the world reduces carbon production in line with the promises made at the last important COP summit before Glasgow, at Paris in 2015.
Doubt about the likelihood and timing is completely legitimate; doubt about the seriousness of what is at stake is not
A half-metre may not strike you as much, but a lot depends on the eyes of the beholder. If someone is precariously holding a bowling ball 50 centimetres above your foot, it won’t seem trivial.
A broken toe heals, however. Over 100m people live in coastal regions which will be below high-tide levels by the end of this century. Many of these people will face loss of property and livelihood without hope of recovery.
Last year, I spent several weeks on the Mekong river, visiting a region where over 14m people depend directly on the health of that ecosystem, and where over 60m less directly depend on the rice produced in the flooded paddies, making it the most fertile rice-producing region in the world. If the South China Sea turns the flooded plains brackish, then rice paddies will make way for salty mangrove swamps, and a river system that today produces more fresh fish than the fish presently harvested in the US will be no longer.
For the people of southern Vietnam—almost all of which is less than one metre above sea level—and many other places immediately threatened by rising seas, the current uncertainty about whether temperatures will rise by 2 or 4C, or whether sea levels will rise by more than one metre, is rather academic. It cannot be reasonably doubted that these communities will have to respond soon to the immediate ongoing effects of climate change, not due to future CO2 emissions, but the change already locked in by past emissions.
In sum, it is now known with close to certainty that mediating the known effects of climate change in the near term will be required if we are to stave off some of the worst consequences for the world’s poorest populations living in equatorial regions. Ensuring access to modern agricultural methods in the face of increasing droughts, and to clean water and reliable power, will be important, as will building coastal infrastructure to deal with rising seas.
Indubitable arithmetic
Because the impacts of climate change elsewhere in the world may take many decades or centuries to manifest, some may still find it hard to muster much urgency to address current carbon production. But here too another basic bit of atmospheric chemistry, independent of current uncertainties in high-resolution climate modelling, is important—the firmly established reality that the additional CO2 concentration we add to the atmosphere will not dissipate for over 1,000 years.
Humanity pumps about 10bn tons of carbon into the atmosphere each year, roughly five billion of which will remain in the atmosphere for a millennium. The cumulative effect of this is measurably material: over the past 60 years, for example, we have already increased the abundance of this gas in the atmosphere by over 30 per cent. And if the current production rate continues, by 2050 the atmospheric abundance will have doubled compared to the pre-industrial concentration of 600bn tons. And almost all of this extra carbon will remain in the atmosphere for the foreseeable future—even if fossil fuel burning now stops or slows.
If we set an ambitious goal of limiting the rise in the total CO2 concentration in the atmosphere to peak at less than twice the pre-industrial level—basically the goal of the Paris agreement—then every decade in which we continue to produce CO2 at current levels makes the required future reductions more severe, and so more disruptive for our society and economy. Had we begun to reduce carbon production in 2010, the required annual reductions would have been 3.7 per cent. Waiting until 2020 to begin already means that future reductions would need to have been 9.0 per cent per year. Waiting for another decade to take action would mean that hitting the same goal would require annual reductions that are likely impossible.
The basic change in ocean temperature this century will produce a sea level change of at least a quarter of a metre
Beyond this, there is the question of climate tipping points that might produce irreversible dramatic changes, due to feedback between various different effects—the possible complete disappearance of that Greenland ice sheet is a particularly disturbing example here. Should it vanish, as it has at times in the distant past, worldwide sea levels will increase by about seven metres, enough to flood almost all major coastal cities on Earth—an outcome that nobody could doubt would be ruinous.
Models suggest that a mean temperature increase of 5 to 7C is guaranteed to irreversibly destabilise the ice sheet, which would then disappear over the course of several centuries to a millennium. Some models, though, suggest a temperature increase of as little as 2C could irreversibly, albeit more slowly, threaten its stability. While the timescale over which it would disappear is different in the two scenarios, the end result is the same.
We cannot say with certainty when such a tipping point might occur. Feedbacks are inherently fiendish—the interaction of multiple processes. So even if we can see where they could occur, it is very difficult to predict exactly how far things can slip before they are triggered. Doubt about the likelihood and timing is completely legitimate; doubt about the seriousness of what is at stake is not.
In the face of admitted uncertainty about the imminence of globally devastating effects, we could gamble on those models that suggest the sheet will survive until the warming reaches five-plus degrees. Or we could instead try to ensure temperatures do not rise as high as the lower value at which other projections suggest the trouble might start. Again it comes down to Dirty Harry’s question: are we feeling lucky?
With time, better models and more data, the uncertainties associated with climate predictions will decrease. But even today, when important residual uncertainties remain, they cannot erase the fundamental realities of climate change now. Nor can they be used as an excuse for inaction. As the former publisher of the New York Times once stressed, keeping an open mind is a good thing, but not so open that your brains fall out.
