The Man Who Predicted Climate Change
In the nineteen-sixties, Syukuro Manabe drew a graph that foretold our world today—and what’s to come.
By Stephen Witt December 10, 2021
Late in 1966, in the sprawling computer lab of the Washington, D.C., office building that housed the United States Weather Bureau, Syukuro Manabe was waiting for a print job to finish. At stake was the fate of the planet. Manabe, who was thirty-five, had come to the U.S. from Japan almost a decade earlier. He managed a team of computer programmers, tasked with building a mathematical simulation of the planet’s atmosphere. It had taken years to perfect, and cost millions of dollars. Now the simulation was complete.
With an alarming clatter, the printer came to life, and a single continuous sheet, striped in light-green and white, unspooled to the floor. The I.B.M. 1403 could print six hundred lines per minute, but Manabe couldn’t stand the noise it made, and usually avoided it by going out for lunch. This job couldn’t wait. If successful, Manabe’s simulation would quantify, for the first time, the relationship between carbon dioxide and the temperature of Earth’s atmosphere.
That the Earth’s atmosphere retained heat from sunlight had been understood since the early nineteenth century. Water vapor was the primary driver, trapping heat energy at lower altitudes and warming the planet’s surface by about sixty degrees Fahrenheit. (If Earth had no atmosphere, its surface temperature would average zero degrees Fahrenheit.) The open question was whether other atmospheric gases contributed to this greenhouse effect. Carbon dioxide was thought to have an effect, but it made up just three parts per ten thousand of Earth’s atmosphere by volume. Researchers wondered whether its impact was detectable.
Manabe speculated that it was. Three parts per ten thousand wasn’t much, but even a trace gas, with the right properties, could have an outsized impact. Without carbon dioxide, there would be no photosynthesis, and almost everything on the planet would die. Perhaps moving carbon-dioxide levels in the other direction—as the combustion of fossil fuels was doing—would have a similarly catastrophic effect.
There was no direct way to test this hypothesis; Manabe, who lived on Earth, did not have access to another, disposable planet on which to run experiments. Instead, he had to simulate the effects of atmospheric change from equations in basic thermodynamics. For the planet’s surface, these equations could be done by hand, but once additional atmospheric layers were added, the calculations grew more complex.
Fortunately, Manabe had access to a machine called Stretch, one of the most powerful computers ever built. Officially, it was an I.B.M. 7030, designed, at the request of the Pentagon, to simulate the effects of the hydrogen bomb. Nine such computers had been produced; others had been sent to the Los Alamos National Laboratory and the National Security Agency. This one, after much lobbying by Manabe’s boss, had been assigned to weather forecasting, to demonstrate to the public that computers could be useful. Stretch was larger than a single-family home, and had sixty freestanding components. The complete apparatus weighed about thirty-five tons, and was cooled by an air conditioner the size of a studio apartment.
Manabe had arrived in the U.S. in 1958; he had never before left Japan, and he spoke little English. But he shared his colleagues’ fascination with computer technology, and, outfitted with the default sportcoat and skinny tie, he fit in quickly. He was attracted to America’s informal social norms, which he preferred to Japan’s more hierarchical approach. “The hardest part was the Western toilet,” Manabe told me last year. “I’d never seen one before.”
Manabe eventually secured a post at Princeton, where he lives today. Last week, at the age of ninety, he was awarded the Nobel Prize in Physics. The prize committee cited Manabe’s 1966 simulation as the first reliable prediction of climate change. The simulation included a plot of points representing the sensitivity of Earth’s temperature to carbon dioxide at different altitudes. The printer didn’t have the capability to fit a curve to the data, so, for the final step, Manabe had to draw it in himself. “I used a pencil,” he said. “It took a long time.”
Manabe’s pencil-line graph revealed three unexpected results. First, according to the simulation, boosting carbon dioxide from three parts per ten thousand to six could cause Earth’s average surface temperature to rise by more than four degrees Fahrenheit. A comparable temperature increase at the end of the last Ice Age had caused ocean levels to rise a hundred feet.
Second, Manabe’s simulation predicted that carbon dioxide would trap heat energy in the lower atmosphere. The Earth’s surface and its oceans would therefore get hotter, while the upper atmosphere would cool. This combination—cooler above, hotter below—is now regarded by climatologists as the smoking gun of human-caused climate change. (Other potential causes of global warming, like the sun growing brighter, would uniformly heat the atmosphere at every altitude.)
Finally, Manabe’s model implied that, as the upper atmosphere cooled, it would deform, causing atmospheric boundaries to pancake. The 1966 pencil-line graph was the first preview of the Earth’s future: the surface was going to cook, and the sky was going to collapse.
Syukuro Manabe was born in September, 1931, on the island of Shikoku, south of the main island of Honshu. His family lived in an isolated mountain hamlet, where his father was the village doctor. On the day Manabe turned three, the Muroto typhoon, then the deadliest storm in Japan’s history, made landfall on Shikoku, destroying thirty thousand homes and leaving three thousand people dead. Powerful cyclones captivated Manabe as a child. “I had a horrible memory and I was clumsy with my hands,” he told a Japanese newspaper. “My only good trait was to gaze at the sky.”
When Manabe was ten, the Japanese Navy bombed Pearl Harbor. In 1944, when he was thirteen, U.S. forces launched one of the largest bombing campaigns in history against mainland Japan. Shikoku was not a target, but bombing convoys would fly over the island on their way to Honshu. While his fellow-students hid out in bomb shelters, Manabe studied for his exams. “Fortunately the airplanes just passed over us, because we’re in the countryside in middle school,” he told an oral historian. Across the channel from Shikoku sat Hiroshima; one of the planes that flew over the island was the Enola Gay.
Manabe has downplayed the impact of his wartime upbringing. “The war didn’t bother me at all,” he said. “I just kept on preparing for the entrance examination.” But he has acknowledged one long-term effect. “I didn’t grow as much as I should have,” he said. “I was undernourished all the time.” The postwar American occupation brought prosperity. In 1955, the Toyota Motor Corporation introduced its first mass-market car. As the middle class recovered, Japanese households sought to acquire the “Three Sacred Treasures”: a television, a refrigerator, and a washing machine. In the decades following the introduction of Western foodstuffs, the average height of an adult Japanese man increased by nearly four inches. Such advancements were spurred by immense growth in the use of fossil fuels.
Manabe aced his entrance examination to the University of Tokyo. His brother, father, and grandfather were all physicians, but Manabe decided to be a physicist. “Then I realized, I’m not that good in math, to get into the difficult physics,” he said. “I’m not that good at measuring things, either. And I had dropped out of biology, because I’m not good at memorizing things.” Manabe ended up in meteorology.
He was a patient student. Struggling to follow some of his professors’ lectures, he learned meteorological physics at his own pace, and had to retake at least one exam. But, when Manabe and his fellow graduate students used these physics equations to predict the weather, he emerged as one of the stars of the department. Lacking access to a computer, the students made calculations on graph paper by hand. “I would spend hours drawing contour lines,” Manabe told me. He seemed nostalgic for the practice: “Drawing contours yourself, you can begin to notice things you’ve never noticed before. Maybe this primitive process is good, in some sense.”
Manabe’s research was inspired by the work of Joseph Smagorinsky, a pioneering American computer scientist. When Smagorinsky learned of Manabe’s efforts, he wrote a letter to the University of Tokyo, seeking to recruit him. Manabe’s adviser released him to Washington. “Not because I was better, but I was expendable,” Manabe said. (Manabe’s self-effacement can sometimes obscure the historical record: his colleagues often credited him with coming up with the fundamental ideas of their work, and Smagorinsky recalled asking for him by name.)
“I flew on a propeller plane,” Manabe told me. “It took about forty-eight hours,” he said, with stops in Anchorage and Seattle. Smagorinsky, wearing sunglasses, met him at the airport. Smagorinsky was in the process of transforming the Weather Bureau from a minor branch of the Department of Commerce, known for the inaccuracy of its crop-planting forecasts, into the leading atmospheric-research institution in the world. Smagorinsky’s parents had fled pogroms in the Russian Empire for the Lower East Side; he attended Stuyvesant High School. Skilled at mathematics, Smagorinsky was a disciple of John von Neumann, the architect of the modern computer. In the early nineteen-fifties, he had worked on the world’s first computerized weather-forecasting program, using the computer eniac. Soon eniac’s predictions were beating the weathermen.
Von Neumann believed that the weather-forecasting work could be extended to a complete, three-dimensional simulation of the planet’s climate, derived from elemental laws of physics. Smagorinsky assigned Manabe to finish the work. “This was impossible, given our tools,” Manabe told me. To make the predictions, Manabe had to solve nonlinear differential equations by brute force, a task that challenges supercomputers even today. But Smagorinsky relentlessly petitioned the government for more resources, and, to Manabe’s astonishment, received them, typically through brinkmanship. “He never walked into a budget meeting without a resignation letter in his pocket,” Manabe said.
Manabe never met von Neumann, who died in 1957, but, when I spoke with him in 2020, he remained astonished that the modern understanding of climate change originates with the man who also invented game theory, worked on the Manhattan Project, and designed the modern digital computer. “It’s amazing to me, to have worked on a problem proposed by von Neumann,” Manabe said of his climate model. Seeking to reduce the complexity of the problem, Manabe’s 1966 model used just one variable. Its success is due to its representation of the way carbon dioxide interacts with heat energy at different atmospheric layers. All other variables were simplified or ignored. Still, this primitive simulation would prove, over time, to have a startling amount of predictive power. When Smagorinsky saw the predicted rise in temperature, he compared it to the warning light on the dashboard of a car.
In 1968, Manabe and his team were relocated to Princeton University, and Smagorinsky’s fiefdom inside the Weather Bureau was soon folded into the newly created National Oceanic and Atmospheric Administration. In 1970, Manabe travelled to Stockholm, for a conference titled “The Study of Man’s Impact on Climate.” Despite Manabe’s findings, scientists there could not agree whether the Earth was more likely to heat up or cool down.
In the early nineteen-seventies, now using an exponentially more powerful computer, Manabe ran the numbers again. The second simulation included equations modelling the effects of snow cover and the exchange of water between the ocean and the clouds. Manabe’s 1975 paper “The Effects of Doubling the CO2 Concentration on the Climate of a General Circulation Model,” which he wrote with the meteorologist Richard Wetherald, once again predicted that the combustion of fossil fuels would lead to a rise in the Earth’s mean temperature of four degrees Fahrenheit.
Ronald Stouffer, an affable meteorologist who’d originally intended to be a TV weatherman, joined Manabe’s group at Princeton in 1977. Stouffer recalled that the group was collegial and informal—everyone called Manabe “Suki.” Manabe himself was enthusiastic and kind, and encouraged the researchers to challenge received ideas. “He also worked fourteen hours a day,” Stouffer said. Stouffer told me that, by the time he arrived, Manabe was convinced that anthropogenic global warming was real. “I think Suki knew it was real when he wrote his 1967 paper,” Stouffer said.
Manabe’s computer models were streamlined and elegant, with concepts represented simply. “There are many people who want to bring in the full complexity of the climate system,” Tom Delworth, another researcher who worked with Manabe, said. “But that creates possible interactions among components, and it becomes hard to really understand what’s going on.” Delworth told me that Manabe was a fan of the Princeton men’s basketball team—especially its longtime coach, Pete Carril, who developed the “Princeton Offense,” and led the school to thirteen Ivy League championships. “That offense consists of very simple but extremely efficient movements, repeated to perfection,” Delworth said. “I think Manabe liked that style, whether it was basketball or science.”
Manabe’s research prompted other scientific groups to run their own simulations. Most reached similar conclusions. By 1979, a consensus had formed among climatologists that the Earth was going to heat up. The temperature record, at the time, didn’t show concrete evidence of climate change, but the increasingly sophisticated computer simulations had won the climatologists’ trust. Manabe’s mathematical technique for showing how light responded to carbon dioxide was represented in every model. “Some of that code, from the early sixties—I’m still using it,” Stouffer told me.
In 1988, a prolonged drought in the western United States prompted congressional hearings about the possibility of climate change. Manabe testified, but he was mostly ignored. “My testimony, with this lousy Japanese accent,” he said. “I just had a terrible impact.” James Hansen, a researcher at nasa, testified the same day, captivating Congress and the public with his dire predictions of the effects of global warming. “Jim did a beautiful job communicating,” Manabe said, “which I don’t have any talent for.”
There was more than just a language barrier. Manabe, by all accounts, preferred scientific work to public advocacy. The Intergovernmental Panel on Climate Change first convened the same year that Manabe gave his congressional testimony. “He didn’t want to go to the meetings,” Stouffer told me. “So he sent me instead.” In interviews in the nineties, Manabe was non-committal about pressuring countries to reduce fossil-fuel consumption. “My personal feeling is, O.K., now if CO2 really quadruples by business as usual, then climate change is so large I think it approaches the days of the Cretaceous period,” he said, in 1998. “But, even then, maybe the human race can adapt to the environment.”
I wondered why Manabe hadn’t pressed harder for environmental action. He has claimed he isn’t comfortable with publicity, but when I put the question to Stouffer I got a different answer. “I’ve debated saying this now to you a couple times,” Stouffer replied. “But I can remember talking with Suki, where, from a scientific point of view, he didn’t want anybody to do anything. He wanted to see if his model was right.”
(Manabe denied that was how he felt. “Everyone agrees that we have to act, but I think the difficult question is in defining just how we should act, just what we should do about climate change,” he said. “My main interest is to understand these phenomena using a climate model.”)
Manabe’s model was right. The atmospheric concentration of carbon dioxide is now four parts per ten thousand and rising. Global mean temperature has increased nearly two degrees Fahrenheit since Manabe’s results were first published, and the ten hottest years on record have occurred since 2005. Earlier this year, nasa published the results of a thirty-year satellite survey of the polar mesosphere, some fifty miles up. Temperatures there are dropping about five degrees per decade, and the entire layer is contracting. The top has cooled, and the bottom has warmed. “Suki always wanted to be first, and he always wanted to be right,” Stouffer said. “He was both.”
In the latter part of his career, Manabe turned his attention to global warming’s second-order effects. As temperature increases, water evaporates more quickly. In the tropics, this means more frequent and more violent cyclones; in arid regions, longer periods of drought. When I spoke with Manabe about this, he was alarmed. “In India and Bangladesh, these storms are very serious business,” he said. “But the thing I worry about the most is drought. All these fires! And in Africa, particularly the Sahel, people keep waiting for the next rainfall. Maybe you can no longer do agriculture there.” I observed that cyclones and food insecurity were issues that affected him in childhood. “I’ve been alive for a long time,” he said.
On Monday, December 6th, Manabe received his Nobel medal at a ceremony at the National Academy of Sciences, in Washington, D.C. His Nobel lecture was broadcast in Stockholm on Wednesday. In it, Manabe repeated his warning. Computing power has increased by many orders of magnitude since the Stretch era—I.B.M.’s latest transistors are smaller than a strand of human DNA. Yet the forecasted range of warming effects have changed little since Manabe’s first prediction. “I would say, even with very primitive tools by today’s standards, he got us about eighty per cent of the way there,” Gavin Schmidt, a leading climatologist at nasa, told me. “Everything we’ve done since then is plus or minus ten per cent.” Modern climate models render the effects of global warming in astonishing detail, and we no longer need Manabe’s computer simulation. We’re living in it.