Arrhenius wasn’t worried by this prospect. He thought it would take thousands of years to reach 11°F, if that ever happened. Meanwhile, he lived in frigid Sweden; rising temperatures seemed like a fine idea. In the future, Arrhenius predicted, “our descendants [will] live under a warmer sky and in a less harsh environment than we were granted.”
His colleagues were even less worried, because they thought he was talking through his hat. A chief source for the disbelief was one of Arrhenius’s longtime acquaintances, Knut Ångström. The son of a prominent physicist, Ångström attended Uppsala at the same time as Arrhenius and arrived at the think tank with him. But his long relationship with Arrhenius did not stop him, in 1900, from tearing into the other man’s work.
Ångström focused on Arrhenius’s depiction of the interactions of light and carbon dioxide. As children learn in high school chemistry, atoms are, so to speak, picky about which wavelengths of light they absorb and emit. They will interact with some wavelengths, but not others. In the nineteenth century, physicists discovered that every substance’s pattern of absorption and emission identifies it as surely as a fingerprint.*4 Examining the patterns for carbon dioxide and water vapor, Ångström realized that they took in many of the same wavelengths. Because the atmosphere contains so much more water vapor than carbon dioxide, any absorption of those wavelengths would be almost entirely due to water vapor. Which meant that Arrhenius’s whole carbon dioxide scheme was wrong—“the observations cannot be treated as Mr. Arrhenius did.” Just as Tyndall had thought, water vapor controlled atmospheric heat and carbon dioxide was irrelevant.
With Arrhenius’s carbon dioxide hypothesis seemingly disproved, scientists advanced other ideas about the origin of the ice ages. They were due to fluctuations in the brightness of the sun. To the uplift of mountain ranges. To continents moving about Earth’s mantle. To volcano dust. To the solar system passing through colder bits of space. To ocean currents. To collisions with extraterrestrial icebergs. Arguments were long, bitter, inconclusive.*5 One of the few points of agreement was stated in 1929 by George Clarke Simpson, director of the British Meteorological Office: it was “now generally accepted that variations in carbon-dioxide in the atmosphere, even if they do occur, can have no appreciable effect on the climate.”
Outsiders
Guy Callendar didn’t agree. Born in 1898, Callendar was the son of Britain’s leading steam engineer and grew up in an ivy-covered, twenty-two-room manse in a fashionable part of West London. He became his father’s assistant and then his successor. Intrepid and curious, he was willing to delve into fields he knew nothing about, atmospheric science among them. Nobody knows why climate interested him. Possibly he simply wondered why winters had become warmer since his boyhood. Callendar himself attributed it to ordinary curiosity: “As man is now changing the composition of the atmosphere at a rate which must be very exceptional on the geological time-scale, it is natural to seek for the probable effects of such a change.”
In the early 1930s Callendar began collecting measurements of the properties of gases, the structure of the atmosphere, the sunlight at different latitudes, the use of fossil fuels, the action of ocean currents, the temperature and rainfall in weather stations across the world, and a host of other factors. In effect, he was trying to redo Arrhenius’s calculations. But Callendar had an advantage. Arrhenius had been forced to guess at many values, because they had not been measured. His articles were therefore full of hedge words: “it is, therefore, justifiable to assume”; “I do not know if this [factor] has ever been measured, but it probably does not differ”; “it seems as if”; “I have convinced myself that by this mode of working no systematic error is introduced” (emphasis mine). Callendar had four more decades of data. He produced the first rough draft of the huge climate models familiar today.
Guy Callendar, 1934 Credit 68
Central to his work were more-precise measurements of carbon dioxide and water vapor. As the reader will remember, the ground absorbs solar energy and then sends most of it back into the air in the form of infrared radiation. The majority of this outgoing infrared energy is caught by water vapor and transferred to the rest of the atmosphere. Just enough leaks out to prevent the atmosphere from heating to unbearable levels.
Two mechanisms are responsible for the escape. The first is that the water vapor releases some of its absorbed energy as infrared radiation, and some of that released infrared beams into outer space (it is re-absorbed and re-emitted by water vapor many times along the way, but eventually passes beyond the atmosphere). The second is that water vapor doesn’t absorb all of Earth’s infrared radiation—it is effectively transparent for certain wavelengths. New measurements, Callendar learned, showed that the most important of these “windows” occurs at wavelengths around 10 micrometers—that is, water vapor lets through light waves that are about ten-millionths of a meter long. Another prominent window occurs at 4 micrometers.
Callendar also learned that scientists had more precisely measured the wavelengths absorbed by carbon dioxide. Contrary to Ångström, they did not overlap precisely with water vapor. Quite the opposite: carbon dioxide absorbs some of the wavelengths that water vapor lets through—it shuts the windows. The more carbon dioxide in the air, the more of this radiation it can absorb.
The infrared spectrum, from wavelengths of about 1 µm (1 micrometer) to those of about 10 µm. Water vapor (H2O, bottom half of chart) in the air absorbs most of the infrared radiation from the surface. But there are gaps (light shaded areas) in the spectrum that it lets through—a safety valve that prevents Earth from overheating. Carbon dioxide (CO2, top half of chart) absorbs only a few bands of infrared. By chance, two of them (dark shaded areas) sit in the water vapor gaps. They absorb some of the infrared radiation that water vapor lets through. The effect is to reduce—by just a bit—the amount of infrared radiation that escapes into space. Credit 69
In Callendar’s scenario, the atmosphere is like a bathtub. Water pours into the tub in the form of infrared radiation. In the tub are small holes—the “windows” through which water vapor allows infrared radiation from the surface to pass. Because the outflow from the holes is approximately equal to the inflow from the spigot, the water level in the tub is constant. Now block a hole or two with chewing gum. That is like adding carbon dioxide to the air. Inevitably, the water rises.
From the human point of view, this is stupid bad luck. If the physical properties of carbon dioxide and water vapor didn’t intersect in this way—if carbon dioxide didn’t happen to absorb the infrared radiation that water vapor lets through—then burning fossil fuels would be of little interest to climate researchers. The carbon dioxide rise would be viewed as a dusty corner of atmospheric science, the province of pedants. Coal and oil could be burned without worry (after removing pollutants). Industrial civilization would not be facing an existential challenge.
Neither Callendar nor anyone else understood the stakes as we think of them today. That carbon dioxide blocks the outflow of infrared radiation from the earth—Callendar, like Arrhenius, thought this was a good thing. “Small increases in mean temperature” would help farmers in cold places, he said. Better yet, they would “indefinitely” postpone “the return of the deadly glaciers.”
In making this argument, Callendar was effectively telling climate professionals that he, an outsider, had made a breakthrough in their discipline—one that they had wholly missed. This did not go over well. He had enough academic status to be allowed to present his ideas to the Royal Meteorological Society in 1938, and to have six professional climate scientists comment on them. But he did not have enough standing to stop the commenters from being condescending (they praised his “perseverance”). Years before, British Meteorological Office head George Clarke Simpson had stressed the consensus that carbon dioxide had “no appreciable effect on the climate.” Now he was one of Callendar’s commentators. The problem with Callendar’s work, he sniffed, was that “non-meteorologists” simply didn?
??t know enough about climate to be helpful.
Two criticisms from the panel were more substantial. The first was that Callendar hadn’t shown why carbon dioxide from fossil fuels wouldn’t be absorbed by the ocean rather than remain in the atmosphere. The second was that all measurements of atmospheric carbon dioxide were unreliable, because scientists’ instruments could be affected by nearby car exhausts, factories, farms, and power plants. Although Callendar had spent years gathering data, the assembled meteorologists were “very doubtful” that the data meant anything.
Undiscouraged, Callendar kept working on climate until his death in 1964. And slowly—very slowly—climatologists began to give his ideas a hearing.
They were almost forced to, because after the Second World War a torrent of new researchers poured into atmospheric science, most of them funded by the U.S. military. During the war, the military had used unfamiliar types of radiation to great effect—infrared light for signaling and sniping, microwaves for radar detection of enemy aircraft. To develop these techniques further, the armed services wanted to know everything about unusual types of light and how they interacted with the atmosphere. Strategic Air Force commander General George C. Kenney laid out his dreams in a speech at MIT in 1947. “Below the infrareds and above the ultraviolets,” he said, “there may be weapons of future warfare as devastating as the atomic bomb.” For example, “An airplane equipped with a sort of super dog whistle could fly around a city for a while and upset the nervous systems of the whole population.” Even more exciting to Kenney, the interaction of radiation and the atmosphere suggested the possibility of understanding and directing the weather: “The nation that first learns to plot the paths of air masses accurately and learns to control the time and place of precipitation will dominate the globe.” Buoyed by visions of “climatological warfare,” the military funded the computer pioneer John von Neumann’s plan to create the first digital simulation of the atmosphere.
The gush of Pentagon money shifted the center of climate science from Europe to the United States—an impact that could be seen in 1957, when three California-based scientists launched two projects that answered Callendar’s critics. The first project came from Hans E. Suess and Roger Revelle, of the Scripps Institution of Oceanography, in San Diego. Like Callendar, they were climate outsiders: Suess was an Austrian physicist who had worked on the fringes of the failed Nazi effort to make an atom bomb before immigrating to the United States; Revelle, an oceanographer, was director of Scripps, the most important institution in his field.
Thanks to his physics background, Suess had figured out that the recently invented technique of radiocarbon dating could be used to distinguish fossil-fuel carbon from other types of carbon.*6 In turn Revelle realized that Suess’s technique could be used to learn whether the ocean, as Callendar’s critics had suspected, was taking in the carbon dioxide from fossil fuels. Working together, the two men determined that the oceans indeed were absorbing most of it. But only after they drafted a paper to this effect did they grasp that pumping carbon dioxide into the sea set off other interactions that ended up with the water quickly releasing much of the gas it had initially absorbed. In the end, the ocean didn’t soak up the carbon dioxide emitted by coal, oil, and gas. Hastily tacking on a paragraph to the conclusion of their article, Revelle and Suess conceded that burning fossil fuels amounted to “carrying out a large-scale geophysical experiment of a kind that could not have happened in the past.”
The second project, complementary to the first, was executed by Charles D. Keeling, another Scripps researcher. Born in 1928, Keeling was the son of a Chicago banker who became convinced that bankers like him had caused the Great Depression. The father quit his job and became a proselytizer for banking reform, Keeling wrote in an autobiographical essay, “thereby plunging our family into the poverty he was distressed about.” The son studied chemistry at the University of Illinois, but switched to liberal arts to avoid a required course in economics—“I felt quite passionately that my exposure to economics at home had been enough.” After further confusion, Keeling returned to chemistry, obtained a Ph.D., fell in love with geology, and wangled a geochemistry fellowship at the California Institute of Technology, in Pasadena. An offhand remark from his supervisor in 1956 set up the rest of his life.
Keeling’s supervisor speculated that the carbon dioxide in water was in equilibrium with the carbon dioxide in the air—if you increase the amount of one, the other will quickly compensate to balance it. Keeling decided to find out if this was true. He was bored with his desk job and liked the idea of performing measurements outdoors, in the mountains near Pasadena. To his surprise, his data jumped all over the place—sometimes there was more carbon dioxide in the air, sometimes less. Keeling realized that “emissions from industry, car exhaust, and backyard incinerators” were blowing by his instruments—exactly the problem Callendar’s critics had identified. To determine the actual level, he would need to collect long-term data in a contaminant-free place. The U.S. Weather Bureau agreed to sponsor Keeling in an effort to measure carbon dioxide on Mauna Loa, a thirteen-thousand-foot volcano in Hawaii. Because prevailing winds were from the west, the nearest carbon dioxide sources were thousands of miles away in Asia.
Keeling’s measurements—CO2 levels in parts per million on the vertical axis, time on the horizontal axis—were precise enough to observe slight seasonal fluctuations of carbon dioxide. As Northern Hemisphere plants grow in the summer, their increased photosynthesis sucks some of the gas from the air, regularly reducing its atmospheric concentration. Credit 70
In the interim, Revelle had learned about Keeling’s initial work and invited him to Scripps. Keeling moved there and assembled the world’s first high-precision gas-analysis system. He exasperated Revelle by constantly tinkering with the design and asking for more money; Keeling, Revelle thought, was fixated on achieving a meaningless degree of precision, pointlessly piling on the decimal points. Measurement began in February 1958. Within two years his instruments showed that the world’s store of airborne carbon dioxide had increased in that period from about 313 parts per million to about 315 parts per million.
Keeling worked on Mauna Loa from 1958 until his death in 2005, during which time the proportion of carbon dioxide in the air rose to 380 parts per million. Combined with the work by Revelle and Suess, Keeling’s meticulous, decades-spanning measurements convinced climate researchers that carbon dioxide was accumulating in the air.
But what, if anything, would be the effects? On an absolute scale, the carbon dioxide rise was minuscule—a few parts per million. Any rise in temperature, most researchers thought, would occur only in the distant future, and could be beneficial. Revelle and Suess had concluded that humanity was conducting an experiment, but thought it unlikely to amount to much. In a few decades, they wrote in their article, enough data might accumulate to “allow a determination of the effects, if any, of changes in atmospheric carbon dioxide.”
“The world is growing slightly warmer,” The New York Times reported one Sunday in 1959, noting that most scientists believed that “the warming trend” is not “alarming or steep.” Befitting the lack of concern, the article appeared on page 112.
Moral Interlude
In the spring of 2016, my friend Rob DeConto published an article in the scientific journal Nature. By chance, our families had dinner together a day after its release. Rob was both cheerful, because he had just finished a long project, and uneasy, because of the implications of that project. With a colleague, David Pollard, he had spent years examining the Antarctic sheet, by far the world’s largest ice mass. Past researchers had thought that because of its size it would respond slowly to rising global temperatures. To their dismay, DeConto and Pollard had realized that Antarctica might be more vulnerable than previously thought.
Increasing temperatures would attack the ice in two ways: warmer air would melt it from above, forming pools on the surface, and warming ocean currents would eat at the underside
of the sheet, creating large cracks. The pools on the surface could drain through the cracks, widening them and splitting the ice sheet into unstable pieces that would fall apart under their own weight. The remaining chunks, surrounded by warm water and air, would melt quickly, like the ice cubes in a cocktail. If the two men were correct, melting Antarctic ice could by itself raise the world’s oceans more than three feet by 2100, enough to swamp Miami, Tokyo, Mumbai, New Orleans, and many other cities. By 2500 the rise could be as much as fifty feet.
Every few years the Intergovernmental Panel on Climate Change (IPCC), a United Nations–sponsored international scientific consortium, issues a set of lengthy, multivolume reports that attempt to portray the state of the science on climate change. In 2013 it put together a consensus projection of how sea level would be affected by a variety of factors, including the melting of glaciers in Greenland, the shrinking of the Arctic ice cap, and the fact that water expands as it gets warmer (the increase is slight, but it adds up when extended across the whole globe). Taken together, the IPCC estimated, these would push up the sea roughly twenty-eight inches by 2100. Antarctica, the biggest ice mass, would remain more or less intact, contributing only one to three inches to the total. This idea seemed borne out by the facts—in the previous few decades, Antarctica had shown little sign of melting. Indeed, parts of the ice sheet were growing. But the IPCC also thought that scientists needed to look more closely, just to be sure. Now DeConto and Pollard had performed that analysis, and they were saying that there was a chance that Antarctica could fall apart by 2100, and that therefore the seas could rise higher and faster than the IPCC had imagined possible.