How do we make choices in the face of uncertainty? In this episode of TILclimate (Today I Learned: Climate), MIT professor Kerry Emanuel joins host Laur Hesse Fisher to talk about climate risk. Together, they break down why the climate system is so hard to predict, what exactly scientists mean when they talk about “uncertainty”, and how scientists quantify and assess the risks associated with climate change.
In essence, risk is about probabilities and about costs, measured in human and monetary terms. For example, in deciding to ascend a stepladder to replace a lightbulb, we may estimate that the probability of falling off the ladder is small but of potentially great consequence, and weigh that against the large probability of successfully changing the bulb, with the attendant benefit of having light. This may be an easy one, but then there are the tough ones. A surgeon tells me that I have a 90% chance of surviving open-heart surgery. But if I do, I might have only a few years left to live. Given that the procedure will cost my family dearly whether it succeeds or not, should I go forward with it?
The assessment of risk therefore requires that we multiply the cost of the outcome by the probability of that outcome. We are then in a position to decide how much, if anything at all, we would be willing to spend to avoid that outcome. Quite often, the very worst outcomes have very low probability, and it is often quite difficult to assess the true probability of very low probability events. Economists call this the problem of “tail risk”, because it relates to the risks associated with the far ends—“tails”—of probability curves. The probabilities of tail risks might be very small, but we cannot ignore them because the costs can be very high.
For example, if you were told by a reliable source that there is a 1% probability that your child would be run over if you let them cross a busy highway, you would almost certainly not take that risk even though the odds are vastly in your favor. The costs are just way too high, particularly when weighed against the relatively low cost of walking to a pedestrian crossing.
When we confront the risks associated with climate change, we need to know something about the probabilities of different climate outcomes, the costs those outcomes might impose on society, and the costs and benefits of mitigating climate change. We also need to confront the tail risks associated with low probability but potentially catastrophic outcomes, such as large and rapid sea level rise due to a collapsing ice sheet.
But there are strong cultural biases running against any discussion of this kind of tail risk, at least in the realm of climate science. The legitimate fear that the public will interpret any discussion whatsoever of tail risk as a deliberate attempt to scare people into action, or to achieve some other ulterior or nefarious goal, is enough to make most climate scientists shy away from any talk of tail risk and stick to the safe high ground of the middle of the probability distribution. The accusation of “alarmism” is often quite effective in making scientists skittish in conveying tail risk, and talking about the tail of the distribution is a sure recipe to be so labeled.
After all, by their very definition, such risks are unlikely to be the outcome. If we want to be admired by our descendants, the best strategy is to stick with the most probable outcomes and with high probability we can then ridicule those “alarmists” who warned of the tail risks, just as the adult who advises the child to cross the street will, in all likelihood, be able after the fact to chastise the one who counseled against it.
Catastrophic climate change is unlikely, but it is possible.
As we explain in the next chapter, in the case of climate change, the most probable outcomes over the next century, barring any action to curtail the emission of greenhouse gases, incur serious costs to society. But if climate change is worse than what we currently think is the most likely outcome, we face the possibility of catastrophic outcomes, so catastrophic that it might be difficult to really attach any definite number to the likely costs. It becomes almost a philosophical question how much we might be willing to spend to avoid the unlikely, but not so comfortably improbable possibility of truly catastrophic outcomes.
To illustrate this a bit more concretely, take a look at Figure 14, which shows an estimate of the probability distribution of global mean temperature resulting from a doubling of CO2 relative to its pre-industrial value, made from 100,000 simulations with a particular climate model. We use this here as an illustration; it should not be regarded as the most up-to-date estimate of the probabilities of global temperature increases.
More or less in agreement with the most recent report of the Intergovernmental Panel on Climate Change (IPCC), the most probable “middle” of the distribution runs from about 1.5°C to about 4.5°C, while there is a roughly 5% probability of temperature increases being less than about 1.8°C and more than about 4.6°C. But, given the corresponding distributions of rainfall, storms, sea level rise, etc., the 5% high-end may be so consequential, in terms of outcome, as to be justifiably called catastrophic. It is vitally important that we account for this tail risk as well as the most probable outcomes.
So far it has been difficult to quantify tail risk beyond that implied by figures such as Figure 14. We have also tried to use paleoclimate data and the observed response of climate to large volcanic eruptions to narrow down the probability distribution. A wild card in climate risk assessment is the problem of abrupt, irreversible climate change, which evidence in ice cores and deep sea sediments suggests has occurred in the past. We also have to be mindful that the graph in Figure 14 and many risk assessment studies use doubling of CO2 as a benchmark, whereas we are currently on track to triple CO2 content by the end of this century. Unless we find a way to extract carbon from the atmosphere (which we discuss in the chapter on Solutions) the climate risks would become alarmingly high (and not just in the tails) in the 22nd century, even if we stopped emissions by the end of this century. Let’s explore those risks now.
08a Sea level rise
We begin by making a simple observation about past sea level rise and human civilization. Remember that as ice volume on Earth goes down, sea level goes up and vice versa. All that water locked in the ice came from the ocean, and so when there are extensive ice sheets there is less water in the ocean. Sea level must have been lower. How much lower? The answer is, roughly 130 meters (400 feet). We know this because we know the volume of land ice and also have direct geologic evidence of ancient shorelines.
Figure 15 illustrates sea level rise to modern values from its low point of about 130 meters (roughly 400 feet) below today’s level, about 22,000 years ago. Notice that sea level has been remarkably stable for the last 7,000–8,000 years—coincident with the time that human civilization developed.
Much damage would be done by a change in sea level of a few feet.
Civilization developed during a time of unusual climatic stability and is exquisitely tuned to the climate of the past 7,000-8,000 years.
And that is just the point. Because our prehistoric ancestors were nomadic, they did not build permanent cities. They probably didn’t even notice the 400 foot rise in sea level over 10,000 years (about 0.5 inch per year). Civilization developed during a time of unusual climatic stability and is exquisitely tuned to the climate of the past 7,000-8,000 years. But in our time, much damage would be done by a change in sea level of a few feet, let alone 400 feet. A modest climate shift in either direction will be highly problematic.
Runoff from melting ice in Greenland and West Antarctica is expected to further increase the rate of sea level rise over coming decades.
Projections range upward to an increase of around 1 meter (3 feet) by 2100, with a few estimates ranging as high as 2 meters (6 feet).
Sea level rose through the 20th century and has continued to rise in the present one; its rate has increased to a little more than 0.1 inch per year, mostly owing to Thermal expansion is the tendency of matter to change its shape, area, and volume in response to a change in temperature.thermal expansion as ocean waters warm. Runoff from melting ice in Greenland and West Antarctica is expected to further increase the rate of sea level rise over coming decades, and projections range upward to an increase of around 1 meter (3 feet) by 2100, with a few estimates ranging as high as 2 meters (6 feet). Most of the thermal expansion effect and at least some of the glacial melting has been directly attributed to anthropogenic warming.
Elevated sea levels make coastal regions more susceptible to storm-induced flooding, as evidenced by the aftermath of Hurricane Sandy in 2012, for example. Rising seas also infiltrate aquifers, putting freshwater supplies at risk. Many cities, such as New York, are weighing the costs and benefits of adaptation strategies such as building massive storm barriers versus hardening individual buildings.
But owing to the slow heating of the oceans, sea level will not stop rising in 2100 even if by then we manage to eliminate emissions. The last time Earth’s atmosphere had a concentration of over 400 ppm of CO2 was during the Pliocene period, about 3 million years ago, during which time sea level was about 25 meters (80 feet) higher than it is today. It may take thousands of years, but that is where sea level is headed, and scientists are not confident about forecasting how fast land ice will melt. There is no way that coastal cities can adapt to that level of change; they would simply have to relocate.
08b Heat and humidity
Warming is also of direct concern. Human comfort is measured by a quantity called the wet-bulb temperature, which is the lowest temperature a damp surface can have in air of a given temperature and humidity. When the wet-bulb temperature exceeds about 35°C (95°F) the human body cannot transmit heat to the surrounding air fast enough to compensate for its internal production of heat, and body temperature rises to lethal values. This limiting wet-bulb temperature is very rarely exceeded in today’s climate, but such values are projected to become common in certain regions, such as the shores of the Persian Gulf, by late in this century. Mortality from heat waves is already of concern; for example, the 2003 heat wave in Europe is estimated to have killed at least 50,000 people. As mean temperatures climb, such heat waves become more common. However, deaths from hypothermia decline with increasing temperature, and as of this writing the data are ambiguous as to the net effect on mortality.
Figure 16 presents an estimate of the number of days each year, by the end of this century, in which the combination of heat and humidity will be extremely dangerous, under emissions scenario RCP 8.5 is a pessimistic projection that assumes no serious effort to curtail greenhouse gas emissions, and robust economic growth.RCP 8.5. (By comparison, such conditions today occur no more than once every 10 years, mostly in a small region of the Midwest.)
08c Destructive storms
Violent storms are another risk to reckon with. Tropical cyclones cause on average more than 10,000 deaths and $700 billion (U.S.) in damages globally each year. There is now a strong consensus that the incidence of the strongest storms, which although small in number dominate mortality and damage statistics, will increase over time, even though there may be a decline of the far more numerous weaker events. The jury is still out on what might happen to the incidence and intensity of destructive winter storms and violent local storms such as tornadoes and hailstorms. Figure 17 shows projections of annual U.S. property losses as a result of the combination of higher sea levels and greater incidence of intense hurricanes.
08d Ocean acidification
Increased atmospheric concentrations of CO2 lead to increases in the concentration of CO2 dissolved in ocean waters. This makes the oceans more acidic. Laboratory experiments show that as ocean acidity increases, organisms that build shells, including certain mollusks, corals, and plankton, begin to suffer declining ability to build and maintain their shells. Thus ocean acidification poses significant risks to marine ecosystems; but these risks are only now beginning to be quantified.
08e Food and water
These changes will become apparent first and be most severe in regions, such as the Middle East, that today have only marginal food and/or water supplies.
Figure 19 shows a projection of the effect of climate change on U.S. agricultural losses, relative to today’s 1-in-20 event. By the end of this century, today’s once in 20 years agricultural loss events could occur every other year.
Political and social destabilization is perhaps the greatest and least predictable risk incurred by rapid climate change.
Historically, the disappearance of certain civilizations, such as that of the Anasazi in what is today the southwestern U.S., has been attributed to food and water shortages brought on by prolonged drought. Such shortages are also thought to cause or exacerbate mass migrations and armed conflict. The link between climate change and human conflict is well recognized in the defense community. For example, in its 2010 Quadrennial Defense Review, the U.S. Department of Defense states that: “climate change could have significant geopolitical impacts around the world, contributing to poverty, environmental degradation, and the further weakening of fragile governments. Climate change will contribute to food and water scarcity, will increase the spread of disease, and may spur or exacerbate mass migration.”
Political and social destabilization of a crowded, nuclear-armed world finely adapted to the highly stable climate of the last 7,000-8,000 years is perhaps the greatest and least predictable risk incurred by rapid climate change. Such existential risks are difficult to attach numbers to and represent extreme outcomes whose probability is not small under high-emissions scenarios.
How long can we wait to act?
Carbon dioxide is a greenhouse gas of special concern because of its long residence time in the atmosphere. Figure 20 shows estimates of the decline of CO2 levels assuming that emissions abruptly stop when concentrations reach various values. Over the first 100 years or so, concentrations fall fairly rapidly, but then the rate of decay drops off and it will take many thousands of years for concentrations to return to preindustrial values.
Temperature hardly drops over the first thousand years after emissions cease, mostly due to heat storage in the oceans.
Figure 21 shows projections of global mean temperature that correspond to the CO2 concentrations in Figure 20. Curiously, the temperature hardly drops at all over the first thousand or so years after emissions cease, reflecting mostly the effects of heat storage in the oceans. This is a crucial aspect of the challenge we face: absent technology for removing CO2 from the atmosphere, we will have to live with altered climate for many thousands of years. Thus we have a narrow time window within which to act.