Five Billion Years of Solitude (23 page)

After a particular atmospheric cocktail is chosen, the core of Kasting’s numerical approach kicks in, most of which he developed during his seven years at NASA. During all that time, he devoted himself to perfecting his models, hand-coding each important way that starlight interacts with an atmosphere. In the real world, and in Kasting’s models, a photon of a certain wavelength might simply bounce off the top of the atmosphere, while a photon of another wavelength might instead pass without incident all the way down to the planetary surface. Inside the atmosphere, real or virtual, a photon might be reflected by a cloud, or by bright ice on the ground. It might be absorbed by a greenhouse gas, or by the dark water of a sea. When a photon is particularly energetic—ultraviolet or higher on the electromagnetic spectrum—it might even create entirely new substances in the air and on the ground by knocking into molecules and splitting them apart—a process called “photolysis.” The photolytic products could then have their own secondary effects on the absorption and reflection of starlight, all of which must be taken into account. Over the years, Kasting accumulated all the necessary data he could find, building up a vast library of radiation-absorption tables, photochemical reaction rates, atmospheric lifetimes of different gases, and the global pace at which certain gases are emitted from volcanoes or absorbed by rocks. Collectively, all these various interactions and inputs have an enormous effect upon a planet’s atmospheric composition and average surface temperature—its climate.

If you naively calculated the average temperature of the modern Earth’s surface based only on the amount of sunlight it receives and its average reflectivity, or albedo, you’d obtain a value of -18 degrees Celsius, well below the freezing point of water. If you calculated it using one of Kasting’s climate models, you’d get a result of 15 degrees Celsius, which is, of course, what the Earth’s average surface temperature actually is. The discrepancy is mostly due to warming from several different greenhouse gases, each of which Kasting must account for in painstaking detail.

Water vapor, for instance, must be treated very carefully, as it is actually a much more potent greenhouse gas than CO
₂
, efficiently absorbing a much broader swath of the thermal-infrared portion of the spectrum. Further, its effect on climate is qualitatively different: unlike CO
₂
, which stays gaseous at typical Earth temperatures, water vapor is intimately affected by Earth’s temperature changes. Low temperatures will cause it to condense into clouds and fall out of the sky as rain, snow, and hail, which removes its greenhouse effect and drives temperatures even lower. Conversely, high temperatures increase the evaporation rate of surface water, pumping more water vapor into the air to raise temperatures even further. Water vapor thus acts in a positive feedback loop to amplify other climate changes, such as the steady heating forced by rising levels of atmospheric CO
₂
. If CO
₂
is the fulcrum about which Earth’s climate change pivots, water vapor is the lever.

The key output of one of Kasting’s climate models is something called a temperature-pressure profile—scientific jargon for how starlight shining on any given atmosphere will influence not only its warmth, but also its vertical structure. Earth’s atmosphere, for instance, reflects a quarter of the incoming sunlight and absorbs another quarter through greenhouse gases, allowing approximately half of the sunlight that strikes it to filter down to the surface. This means that, on average, Earth’s atmosphere is colder than its surface, and is warmed from the bottom up by convection, like a pot of water being heated on a stovetop. Most of the surface heating and convection occurs around the equator,
where, as a cursory examination of any globe will show, there is more surface area to absorb the sunlight that beats down from almost directly overhead. Convective cells of moist air undulate from the warm surface, cooling as they rise and expand, eventually growing cold enough to dump their moisture as condensed water vapor—that is, as clouds and rain. Atmospheric convection helps to explain why the tropics are hotter than the poles, why air around high mountaintops, though fractionally closer to the Sun’s radiance, tends to be thinner, colder, and drier than the air at sea-level plains, and why thunderstorms typically occur on torrid afternoons and early evenings, hours after the Sun’s zenith.

Earth’s temperature-pressure profile creates a feature in the atmosphere called the tropopause, a dividing line which runs above the warm, weather-filled troposphere and below the colder, thinner stratosphere. Since water vapor condenses when exposed to cold temperatures, it is effectively trapped beneath the tropopause by the colder overlying atmospheric layers. Just how important this “cold trap” effect is for Earth’s prolonged possession of water became apparent in the 1980s through a series of studies by Kasting, his colleague James Pollack, and a handful of their peers at NASA Ames. They were interested in understanding why Venus, our planet’s near twin, had developed such a dramatically different climate than Earth, despite evidence that in its very early history our sister planet had been hospitably tepid and wet, rather like our own world now.

“To someone like me, the single most interesting thing about Venus is what it says about the inner boundary of the habitable zone,” Kasting explained as we chatted in his office. “It sets a reasonable empirical limit on what you can expect for other planets outside the solar system—you don’t need to do much modeling to guess that something getting Venus’s amount of starlight probably won’t be habitable. So if you want to know what happens when an otherwise Earth-like planet forms too close to its star, or what can happen to a habitable planet as its star gets brighter over time, Venus can tell you a lot.”

Building on previous work performed by several other planetary scientists, most notably Caltech’s Andrew Ingersoll, Kasting modeled how the Earth’s atmospheric structure—Earth’s temperature-pressure profile—would react to increased intensity of sunlight, as would occur if the Earth’s orbit were moved in toward the Sun to a more Venusian orbit, or when the Sun slowly increases its luminosity over geological time. He found that with a relatively modest 10 percent increase in the starlight’s intensity, equivalent to moving the orbit of our planet to 0.95 AU, 5 percent closer to the Sun, the additional warming would saturate the troposphere with water vapor, pushing the tropopause up to altitudes of 90 miles or more.

As Kasting watched the tropopause soar in his numerical model, he knew he was witnessing what would lead to the end of that virtual world, and someday, our own: much of the water vapor that lofted to such heights would rise above the protective ozone layer, where it would be photolyzed by ultraviolet light from the Sun. A small percentage of the liberated atomic hydrogen would escape entirely into outer space, taking with it any potential to ever bond again with Earthbound oxygen to create water. Within a few hundred million years, enough hydrogen would be lost to space in this manner that Earth’s oceans would essentially boil away, leaving the planet lifeless and dry as a bone, with not a drop of water left upon its surface or in the air. In a billion years, long before it swells into a red giant and threatens to physically engulf our world, the Sun will have brightened by that crucial 10 percent, and the Earth will begin to rapidly lose its water and its life. This “moist stratosphere” mechanism is now thought to be how Venus began losing its oceans early in our solar system’s history, and its threshold of 0.95 AU for our own planet conservatively approximates the inner edge of Kasting’s habitable zone from his canonical 1993 paper.

As Venus lost its oceans, the rising temperatures baked CO
₂
out of the planet’s crust, and the gas began to fill the atmosphere. As a result, Venus’s atmosphere is now some 90 times denser than Earth’s and almost pure CO
₂
, creating a greenhouse effect so potent that the planet’s
surface temperature is hot enough to melt lead. In a second series of studies, Kasting and his coworkers modulated the CO
₂
content of Earth’s atmosphere to examine whether increased CO
₂
, rather than increased sunlight, could on a much faster timescale independently lead to the loss of oceans through a moistened stratosphere.

To his surprise, Kasting found that even as rising CO
₂
levels sent temperatures skyrocketing, the vast amounts of water vapor released acted like the lid on a pressure cooker, pressurizing the lower atmosphere to such an extent that the oceans never boiled, keeping Earth’s stratosphere relatively dry. For the stratosphere to become saturated with moisture, for the oceans to vaporize and escape into space, the numerical models indicated that Earth’s atmospheric CO
₂
would have to reach more than twenty-five times its present concentration—more than could be released by burning the entirety of our planet’s known “conventional” fossil fuel reserves of oil and coal, but just maybe within reach if all the planet’s unconventional sources, like the Marcellus’s shale gas, were burned as well. While humanity could readily give the planet a fever that could shrivel societies and severely diminish existing biodiversity, Kasting’s calculations suggested it would be very much harder—though not definitively impossible—for humans to create a moist stratosphere. By his reckoning, forcing the planet to give up its ocean to space by burning fossil fuels appears to be just beyond the reach of present-day civilization.

There are, however, significant uncertainties in Kasting’s considerations, such that science cannot yet entirely dismiss the possibility of a man-made moist stratosphere leading to a premature runaway greenhouse on Earth. Other greenhouse gases besides CO
₂
and water vapor play a role in Earth’s climate, and could potentially have significant future effects that are unaccounted for in Kasting’s models. And no one presently knows the exact amount of fossil fuels locked away within the Earth, or how much of that guesstimated total could be effectively extracted and burned based on future market conditions and potential technological development. Most fundamentally, no one fully
understands how wide variations in temperature and pressure can subtly affect water vapor’s absorption of thermal-infrared radiation. Nowhere is this haziness more evident than in considering the problem of clouds.

To the average person, clouds are simple things, pieces of cottony fluff in blue skies or ominous gray sheets portending dismal weather. To a climate modeler like Kasting, clouds are the most mercurial and beguiling form of water vapor, fickle creatures almost alive in their fiendish complexity. Depending on a cloud layer’s extent, altitude, and composition, it may either warm or cool a planet. A blanket of dense, low clouds can reflect a good portion of sunlight into space, potentially reducing temperatures. But throw a layer of thin clouds high above the low, dense ones, and much of that cooling effect will be undone, as the translucent upper layer now allows sunlight to stream down but traps the heat that subsequently tries to escape. What everyone agrees on is that as a planet like Earth warms, more water vapor steams into the air to form more clouds. But there is no consensus on where exactly those clouds would form and linger in the atmosphere, or the limits of their feedback effects. Both global-warming deniers and publicity-hungry planet hunters have found refuge in the resulting nebulosity: water-vapor clouds could, in theory, save an otherwise habitable planet from runaway global warming, whether induced by an overabundance of greenhouse gases or by the too-bright light of a nearby star. Farther out away from a star, where temperatures drop low enough for CO
₂
to condense into ice, an insulating blanket of dry-ice clouds could in some circumstances warm a planet enough to preserve liquid water at its surface. In 1993, Kasting conservatively estimated the habitable zone’s outer edge to lie slightly beyond the orbit of Mars at 1.65 AU, but it could in fact extend out much farther, depending in large part on the uncertainties associated with CO
₂
clouds.

There are two divergent strategies for numerically approximating clouds. One is to model them as accurately as possible in extremely detailed three-dimensional simulations. This approach requires reams of
data from Earth-observing satellites as well as state-of-the-art supercomputers, and risks losing the distinction between cause and effect in a flurry of variables and feedbacks. The other strategy is to model clouds much more simply in fewer dimensions, which carries the risk of overlooking vital behaviors that only emerge through complex interactions beyond the model’s boundaries. Kasting prefers simplicity. His models are one-dimensional, approximating the entirety of a planet’s atmosphere with a single linear sounding, something like measuring the average temperature and salinity of an ocean by sampling seawater through a very long seabed-to-surface soda straw.

“Clouds are pretty arbitrary in 1-D—you can get any effect you like in a 1-D model by playing with how you represent them. The ideal scenario for a 1-D model is a cloudless sky, which is obviously a huge weakness,” Kasting acknowledged when we discussed his models. “I try to get around it by basically painting the clouds on the ground, approximating their effect by tuning the surface albedo until it reproduces the average temperature of whatever planet I’m trying to look at—Earth, for example, or Mars. Some people don’t like that, and exactly what my method actually means in terms of real clouds is complicated, but I think of it as minimizing any cloud feedbacks that may occur as a planet’s temperature changes. To do any better than that, you have to go to 3-D, which is a very big step, and even there, clouds remain the biggest uncertainty—the 3-D guys don’t know how to do them, either.”