by Bryan Walsh
Chelyabinsk was a violent reminder for those in the NEO community that space can still throw things at us that we can’t see coming. The asteroid approached from the direction of the sun, so it couldn’t be spotted by any Earth-based telescope. (Point a powerful telescope in the direction of the sun and its expensive innards will melt.) Its size was well below even the 140-meter cutoff that NASA is tasked with finding. Although observatories like Catalina have found and cataloged about 500 asteroids of comparable size, astronomers estimate there are probably millions of Chelyabinsk-class NEOs out there, waiting to be discovered.
Chelyabinsk also showed that it doesn’t take a large asteroid, or even one that makes it all the way to the ground, to cause panic and destruction—especially if it strikes without warning. Chelyabinsk Oblast is home to one of the biggest nuclear processing facilities in Russia. I visited the region in 2007 while reporting a story for Time magazine on Russia’s environmental woes. (The Mayak processing facility in Chelyabinsk Oblast was the site in 1957 of one of the worst nuclear accidents on record, though the full extent of the damage was long kept secret by the Soviet Union.) Even in 2007, when relations between the United States and Russia were less strained, police pulled over my car to ask what a foreign reporter and photographer were doing nosing around a militarily sensitive area. The airburst from the 2013 meteor could have easily been mistaken for a nuclear strike by the United States, which was indeed the first reaction of many witnesses on the ground. Nuclear tensions between Washington and Moscow are even higher now than they were in 2013. It’s not difficult to imagine—but horrifying to picture—what a knee-jerk Russian reaction to a seeming nuclear attack could have led to.
The Chelyabinsk meteor proved there are still holes in NEO surveillance, as did news that a meteor exploded above Russia’s Kamchatka peninsula in December 2018 with the force of ten Hiroshima bombs. But there are new tools coming online soon, most notably the Large Synoptic Survey Telescope (LSST), a wide-field telescope under construction in the mountains of northern Chile. Set to begin full operations in 2022, the LSST will be able to photograph the entire available sky every few nights.64 That should help multiply the rate of NEO discovery as much as tenfold.65
Asteroid hunting will also benefit from advances in artificial intelligence, as machines get better at picking out actual NEOs from the background visual noise. As I found on my visit to the Mount Lemmon Observatory, machines have a tendency to produce false positives, which is why human observers are still needed to screen the initial results. But NASA’s Frontier Development Lab has brought together AI experts and astronomers to trawl the enormous data sets generated by observatories and develop algorithms that can rapidly pick out potentially hazardous asteroids.66 In 2017, researchers from Google and the University of Texas fed data from NASA’s Kepler space telescope into a machine-learning algorithm and found two new planets that had been previously missed.67 The algorithm was correct 96 percent of the time—a batting average any asteroid hunter would love to have. And AI doesn’t need coffee.
What could really turbocharge the NEO search would be a space-based, infrared telescope dedicated to planetary surveillance. Orbital telescopes can run day and night, and they aren’t subject to bad weather. Because NEOs are warmer than other space bodies, they’re easier to locate through an infrared telescope that hunts for heat rather than an optical telescope that uses visible light. “We’ve had several studies over the years, and they come back with the same answer,” Johnson told me. “We really need to have a space-based capability that is able to get above the Earth’s atmosphere so that we can hunt for them twenty-four seven.”
The downside is cost. Space is not cheap. It’s expensive to get objects successfully off the planet, and it’s expensive to keep them running when every repair mission requires an orbital rendezvous. NASA has a project on the books for a space-based infrared telescope called NEOCam, but as of early 2019 it remains in what the agency calls “extended phase A study,” hovering between budgetary life and death.68 Without the aid of space-based telescopes like NEOCam, NASA says it could take thirty years or more to meet its congressional mandate to find 90 percent of the NEOs above 140 meters69—a mandate, it should be noted, that came with no extra funding, just expectations. That means decades when the Earth could be surprised by the next Chelyabinsk—or something much worse.
When I began working on this book I knew I wouldn’t be satisfied with merely investigating the ways our world might end. I wanted to determine what we can do to protect ourselves, which policies and which priorities need to be put into place to give human beings the best chance of making it to the next century and beyond. These aren’t easy questions—the world has no shortage of needs, many of them far more immediate than the remote chance of a species-ending catastrophe. That’s why existential risks tend to be overlooked and underfunded. Which is understandable—until the day that delay becomes fatal.
So how much should we spend and how hard should we work to try to avert a disaster that would be as catastrophic as it is unlikely? This is where asteroids provide a useful starting point for breaking down the tricky math of existential risk. The true probability of most other existential threats is unknown and unknowable. Scientists have no way of accurately forecasting the chances of an AI uprising, epidemiologists can’t foresee a pandemic before the first cough, and no one knows if aliens even exist, let alone if they want to turn Earth into their latest colony. But if they can track all the asteroids out there, the very smart astronomers and physicists dedicated to planetary defense can predict with near-clockwork precision the probability of a too-close encounter with an NEO over the next century, or even longer. And that predictability means that we can make a case—not merely invent one—for spending much more to protect the planet.
Impacts from asteroids larger than about three miles across could plausibly lead to human extinction. A hit from an asteroid that big is predicted to occur once every 20 million years, which translates to a 0.000005 percent chance of it happening in any given year. This is lower than your chance of dying in a lightning strike, which amounts to 0.000009 percent.70 Yet people do die from lightning strikes—sixteen Americans were that unlucky in 201771—while there was not a single recorded human death from an asteroid impact last year, the year before, or, for that matter, all of recorded human history.72 This raises the question: if we’ve made it this long on Earth without a single asteroid-related fatality, why spend a single dollar on NEO defense?
This is where we need to understand what makes existential risks different from ordinary risks. Decades of experience have told us that on average lightning will kill a couple dozens of Americans a year at most—but we also know that there’s a zero percent chance that some kind of mega-lightning strike will zap everyone on Earth and bring an end to the human race. That’s how it works for conventional risks. The number of people who die each year from airplane crashes or hurricanes or from accidents while taking selfies (dozens annually in that last category, according to scientists73) can rise or fall based on external factors or the preventative measures we might take, but there won’t be one year when the entire population of the world dies in a plane crash. These risks—the risks we navigate every day—are finite.
Existential risks, by contrast, are highly improbable but carry the threat of extreme and even infinite consequences, which can skew the numbers in unexpected ways. The National Research Council has estimated that an average of 91 people a year will die in an asteroid strike, but of course those 91 people did not die last year and it’s unlikely that 91 people will be killed by an asteroid next year.74 Rather, it means that at some point—unless we develop an impregnable asteroid defense system—we can expect that millions or even billions of people will die in one major impact event. Average that global death toll over the very long period of time that’s likely to pass before that asteroid strike, and you get 91 deaths a year. That’s more than the number of people who died globally in airliner crashes i
n 201775—and that same year the Federal Aviation Administration alone spent nearly $2 billion on aviation safety.76
This is the mathematics of end times. It may seem impossible to put a dollar figure on human extinction, if for no other reason than that if the world ended, there’d be no one left to pay the bill. That plays havoc on the conventional cost-benefit analyses economists use to determine whether a given policy is worth its price tag, but we can at least attempt to approximate the value of avoiding extinction. The authors of a report by the Global Challenges Foundation, a Swedish nonprofit that raises awareness of existential risks, estimated that if human beings could protect themselves from global catastrophic threats like asteroids, the species could expect to survive for 50 million years. That would be enough time for 3 quadrillion future humans, or three thousand trillion people. If each of those lives, and the lives of the 7.6 billion people on Earth today, were valued at the bargain-basement price of $50,000—far stingier than the price the U.S. government puts on a single American life when judging the impact of regulations—the cost of a total extinction event would be $150 quintillion, nearly two million times larger than the current value of all the money in the world.77
As the jurist and economist Richard Posner points out in his book Catastrophe: Risk and Response, one way to try to determine how much should be spent to avert a risk is to multiply its expected cost by its probability.78 Remember that 0.000005 percent chance of an extinction-level asteroid strike happening in a given year? As minuscule as those odds are, the $150 quintillion cost of extinction is so high that the math suggests we should be spending $750 trillion a year to avert NEO strikes. That would be a thousand times more than the United States currently spends in total on defense. It’s more than nine times the value of the global economy. It is a lot of money.
It would be insane, not to mention impossible, to spend that much money on NEO defense. There are other risks that demand attention and dollars, including existential risks we’ll learn about later in the book. Even NASA, which spent nearly $150 billion in today’s dollars to put men on the moon, would run out of ideas well before it could spend anything close to three-quarters of a quadrillion dollars on shooting down asteroids. But what this thought experiment does tell us is that spending $50 million a year, as NASA has been doing, or even $150 million, is almost certainly too little to keep the Earth safe from a threat that has ended life on this planet before, and which will do so again.
What’s true of asteroids is true of existential risks more generally. The cost of losing the world to any cause, however unlikely, is so great that we should be spending far more of our money and effort to offset those risks however we can. But as I discovered again and again while working on this book, we’re held back less by what we can budget than by what we can imagine. Human beings are terrible at evaluating risk—especially existential risk. We rely on feeling rather than fact, and privilege emotional memories over hard numbers. “We treat something as impossible unless there is an experiential aspect to it, and no one has experienced an asteroid strike,” said Paul Slovic, a professor of psychology at the University of Oregon and an expert in risk perception. “The human tendency is to take that small probability and sweep it to zero.” That’s how we end up ignoring risks that could wipe us off the face of the planet. Not because we’re making a reasoned decision to spend money on one need over another, but because we’re not being reasonable at all.
That’s an understandable tendency. It’s also one that just may get us all killed, unless we’re brave enough to come face-to-face with the end of the world. The universe may be trying to kill us, but that doesn’t mean we have to let it.
VOLCANO
A Decade Without a Summer
Seventy-four thousand years ago, give or take a few millennia, Homo sapiens had a very bad day, perhaps the worst day that we’ve ever experienced. On what is now the Indonesian island of Sumatra, a mountain called Toba exploded—though the word doesn’t do justice to the act of sheer geophysical violence perpetrated that day. What happened to Toba was so destructive that scientists would eventually coin a term for volcanic disasters of its scale: a supereruption.1 But the eruption was only the beginning. Toba’s aftereffects dimmed the sun and draped a volcanic winter around the world. It might have brought our species closer to the brink of extinction than we have ever been before or since. At a moment when Homo sapiens was far from the world-dominating force we are today, Toba was our ultimate trial. It was also a warning—the most dangerous natural existential risk we face comes not from the skies above, but from the ground beneath our feet.
As a journalist with Time in Hong Kong, I helped report on the aftermath of the 2004 Indian Ocean tsunami. More than 230,000 people, in fourteen different countries, died in the disaster, which was triggered by a 9.1-magnitude earthquake that ripped through the bed of the Indian Ocean off the coast of that same Indonesian island of Sumatra. The blunt power of the quake created a rupture 600 miles long; the scale of human death and injury was so great it extended across an ocean. I didn’t think the Earth was capable of anything worse. But as I researched the Toba supereruption, I learned that catastrophe had categories beyond my imagination.
A volcano is a mechanism to transport what is under the ground into the air, and so eruptions are judged first on the sheer amount of rock they emit. As much as 9 million tons of sulfuric rock and dust per second erupted from the Toba volcano over what scientists estimate was about two weeks of devastation. By the time the eruption had finished, Toba had spewed the equivalent of as much as 700 cubic miles of volcanic ash and magma into the air.2 Collect it all in one place and there would be enough to load the Grand Canyon nearly three-quarters full.
If that doesn’t help you imagine the scale of destruction, picture Mount St. Helens in Washington State. Its eruption in 1980 was the largest and most destructive in U.S. history, killing fifty-seven people and causing a billion dollars in damages. It was captured on video and even made the cover of Time magazine. Yet on the basis of the amount of volcanic tephra—rock, ash, and tiny microscopic pieces of glass with a telltale hook shape—that was emitted in the eruption, Toba was the equivalent of 2,800 Mount St. Helens eruptions, or one a day for more than seven and a half years.3
Closer to Toba’s eruption site on Sumatra, a dense mass of molten rock and gas—some of it as hot as 1,200 degrees, with boulders stampeding as fast as 60 mph—flowed for hundreds of miles. Anything that could burn caught fire, including plants or animals.4 The eruption pulverized what had been a mountain and carved out a caldera—a volcanic crater—that today measures more than 60 miles by 18 miles, with a depth of a third of a mile.5 That caldera would eventually fill with water, forming the lake that now bears Toba’s name.
Eruptions are ranked through the Volcanic Explosivity Index (VEI), a scale that runs from 1 at the lowest to 8 at the highest. Toba was far and away an 8, a level technically described as “mega-colossal.” The eruption cloud reached as high as nineteen miles above the surface, well into the stratosphere.6 The noonday sun would have gone dark. “Toba was undoubtedly the biggest eruption the planet has seen over the last one hundred thousand years, if not far longer,” Steve Sparks, a volcanologist at University of Bristol in the United Kingdom, told me.
Toba made its immediate surroundings hell on Earth, but its significance for the story of existential risk came in the effects felt thousands of miles away. Eruptions produce ash with a unique chemical signature, a volcanic fingerprint that allows scientists to trace Toba’s legacy around the globe. Beginning in the 1990s, researchers found evidence of large Toba ash deposits scattered in marine sediments in the Indian Ocean. Further studies found ash in the South China Sea, the Arabian Sea, even in Lake Malawi in southeastern Africa, more than 4,000 miles away from the eruption site.7 In India and Pakistan ash drifts accumulated as deep as 20 feet,8 and today researchers estimate that nearly 1 percent of the planet’s surface would have been covered by Toba ash.9
Vol
canic ash is toxic in its own right—breathing ash can cause lung impairment and lasting respiratory damage, and it can poison pasture and cropland. But the climatic effects of the Toba ash cloud were far more devastating. The ash and dust blown into the stratosphere darkened the sky. While dust particles are usually washed out of the sky within weeks by rainfall, the eruption cloud also contained huge concentrations of sulfur dioxide—just as the debris cloud made by the Chicxulub asteroid impact did. The SO2 combined with water in the stratosphere to generate sulfuric acid droplets, producing a lingering haze that could have reduced the amount of incoming sunlight by as much as 90 percent.10
We know from direct experience that major volcanic eruptions can trigger temporary global cooling by blunting sunlight. After Mount Pinatubo in the Philippines blew in 1991, the resulting clouds of sulfur aerosols reduced global temperatures by nearly one degree Fahrenheit over the next couple of years.11 Pinatubo was one of the strongest eruptions in the twentieth century, but it only scored a 6 on the VEI scale, which, like the Richter scale for earthquakes, grows logarithmically—an increase of 1 on the index equates to 10 times as much ash and rock. That means Pinatubo—which had a measurable cooling effect on the planet’s climate, enough to temporarily offset global warming—was only about 1 percent as powerful as Toba likely was.
Just as paleontologists can reconstruct extinct animals by searching the fossil record, paleoclimatologists can reconstruct past climates by sampling bubbles of air trapped tens of thousands of years ago in Arctic ice. Ice cores and other geologic evidence recovered from Greenland indicate a sudden and lasting drop in global temperatures around the time of Toba, with the eruption as a prime suspect. It’s not conclusive—over the past few million years, the Earth has swung between ice ages and warmer periods, and around the time of Toba the planet was entering into a prolonged glaciation.12 But ice ages take thousands of years to develop, and what could have happened after Toba would have been immediate and drastic.