Nautilus Members enjoy an ad-free experience. or Join now .

John Muir was an idiot. “Climb the mountains and get their good tidings,” he wrote in his 1901 book Our National Parks. “Nature’s peace will flow into you as sunshine flows into trees.” Muir is my hero, his eloquent defense of the natural world an inspiration and the reason the wild places I love are still around. But as I hang here, thousands of feet above the ground on a sheer cliff face in Yosemite watching the onset of an angry and fast-approaching storm, it is difficult to think charitable thoughts about the man. I climbed this mountain in search of the salvation Muir promised. Instead, I find buzzing mosquitos, unprotected traverses across uneven ledges, and salt-crazed marmots intent on eating my socks.

Worse, there are clouds. Friendly and “fleeting mountains of the sky,” Muir called them. Now they are a malevolent force. Sky that was blue 30 seconds ago is thick and gray and rumbling. Condensing air rises and spreads into an ominous anvil above our heads. Lightning encroaches. Rain—or is it hail?—rattles against my helmet. My climbing partner and I set up a rappel line for a hasty retreat. As we slither down the rope, I am keenly aware of the fallacy of Muir’s optimism.

Nautilus Members enjoy an ad-free experience. Log in or Join now .

Muir saw Nature as protector, mender—mother, even. “Earth has no sorrow that Earth will not heal,” he said. But he couldn’t have foreseen the scale of damage we are capable of inflicting on the natural world. Computer models predict—and abundant observational evidence confirms—that greenhouse gas emissions are profoundly altering Earth’s climate. Global warming, like the oncoming storm, is an imminent menace.

What we don’t know—what climate models can’t predict—is exactly how hot the planet will get. Certain models predict that a doubling of carbon dioxide levels will lead to catastrophe: a temperature increase exceeding 8 degrees Fahrenheit and attendant calamities, including floods, droughts, and heat waves. Other models indicate a more muted response: warming of less than 4 degrees.

Nautilus Members enjoy an ad-free experience. Log in or Join now .

We are beginning to put together a picture of how clouds behave in a warming world. It is scientific detective work.

The gap between these predictions is stubborn, persisting even as models have advanced to incorporate more complex and realistic physical processes. And the uncertainty suggests that perhaps there exists some natural phenomenon that climate researchers are simulating poorly, understand incompletely, or have simply overlooked. Where might this missing component reside? In the respiration of plants? The currents of the oceans? Will Earth, after all, wield some unknown power to help itself and us, as Muir would have us believe?

An answer may be found in the very clouds that are driving me off this cliff. Basic physics dictates that clouds affect the climate system in different ways depending on their height, composition, and thickness. Clouds tend to reflect more sunlight than the surfaces below them, preventing some energy from reaching Earth, and thereby cooling it. However, clouds can also act like blankets, trapping heat radiated from our planet’s surface. When light rays strike the earth, it absorbs and re-radiates them into the atmosphere, where clouds can absorb and re-radiate them once more. But because the top of a cloud is generally colder than the surface of Earth, clouds typically emit less radiation back into space. The higher the cloud, the colder it tends to be—and the more pronounced this effect. It follows that if greenhouse gas emissions increase cloud cover up high or decrease cover down low, warming may accelerate, or at least continue unabated. If clouds change in opposite ways, warming may slow.

So which is it? Will we escape the worst fallouts of climate change at the cost of more gloomy days and fewer tropical thunderheads? Or will clouds conspire to make our situation worse?

Nautilus Members enjoy an ad-free experience. Log in or Join now .
peace and quiet: John Muir contemplates nature, seated on a rock near a lake. This photograph was taken around 1902.Library of Congress

Stories about climate scientists often feature wholesome-looking, healthy people wearing harnesses and crampons. They rappel into active volcanoes, or frown at dwindling stocks of polar bears, or rescue each other from glacial crevasses. I am paid to do none of these things. I sit in a small room directly above Tom’s Restaurant in Manhattan and think unfriendly thoughts about my temperamental, recalcitrant computer, whirring in a data center more than 200 miles away.

I need more computing power than can fit in my office because I work with terabytes of data at a time. I sift through state-of-the-art climate models to understand why they say what they say, why their predictions sometimes conflict, and to confront them with reality: satellite output, ground-based measurements, even hand-written records of ocean temperatures measured with buckets. Lately, I’ve been trying to make sense of what these models tell us about clouds in order to understand how and why they will change as the world warms.

Nautilus Members enjoy an ad-free experience. Log in or Join now .

Studying clouds forces you to be comfortable with uncertainties, ambiguities, and the push-pull of competing outcomes. Clouds are the ultimate scale jumpers and shape shifters: Formed in microscopic processes, they cover more than half of the sky at any given time. But while models are good at simulating large-scale movements of air and water—some even capture smaller eddies that drive our everyday weather—they miss physical processes that happen on the tiniest scales, such as condensation around a grain of dust or air flowing over a hill. The upshot is that models poorly emulate real clouds. It simply costs too much time, money, and computing power to design a model that simultaneously captures the very large and the very small.

Yet despite these challenges, we are beginning to put together a picture of how clouds behave in a warming world. I work in a field known as climate-change detection and attribution—“D&A” in the local lingo. It is scientific detective work in three parts. First, we use theory and models to determine the effects, or “fingerprint,” of anthropogenic climate change on a particular geographic region or a particular variable, such as cloud cover. We then turn to data: Do we observe this fingerprint in the real world? Is it detectable beyond the natural variability of Earth’s climate system? Finally, we check whether our observations match our simulations.

John Muir was wrong. Nature has no feelings or urges toward our preservation.

To describe the fingerprints that human-induced climate change may leave on clouds, I am working with cloud scientist Mark Zelinka and his colleagues at Lawrence Livermore National Laboratory, in California. So far we have identified three well-understood physical processes that most climate models incorporate, despite their large variability in simulating clouds. We know, for instance, that warmer air holds more water vapor. So in the absence of other changes, global warming will likely make wet regions, including the tropics and mid-latitudes, wetter. But because that surplus water must come from somewhere, evaporation will increase in subtropical deserts and other dry areas to fuel distant rainclouds. We expect cloud cover to reflect these changes, and indeed models predict that, in a rich-get-richer arrangement, increased water vapor may increase cloud cover in already cloudy regions.

Nautilus Members enjoy an ad-free experience. Log in or Join now .

The second process concerns the large-scale dynamics of the atmosphere. The circulation of air around the planet mainly depends on the temperature contrast between the hot equator and the cold polar regions. As the planet warms, temperatures will rise fastest at the poles, particularly in the Arctic, where melting snow and ice exposes warmer, darker ground underneath. This disparity will flatten the pole-to-equator gradient, affecting air circulation. As a result, both wet (cloudy) and dry (clear) regions will likely shift, spreading out toward the poles in both hemispheres.

Lastly, as the planet heats up, the troposphere—the layer of atmosphere where weather happens—will likely expand to reach a higher altitude. And the tops of clouds should rise along with it.

In satellite data, we have found tantalizing evidence that, in many ways, clouds are behaving as we’d expect. Yet persistent and large uncertainties remain. There are two independently maintained satellite datasets—the International Satellite Cloud Climatology Project, or ISCCP, and the Pathfinder Atmospheres-extended dataset, or PATMOS-x—that span the mid-1980s to roughly the present day. Both datasets are plagued by inconsistencies, inhomogeneities, and spurious artifacts, which make most scientists wary of using them to estimate long-term trends. Happily for our work, however, these datasets agree on the locations of the cloudiest and clearest regions and, to a certain extent, on how these regions are changing. For instance, the data undoubtedly indicate that cloudy and clear regions are moving toward the poles. The datasets disagree, however, on how the amount of cloud cover is adjusting. They also diverge somewhat on the pattern of clouds’ vertical migration, although there is some evidence that high clouds are indeed rising—theoretically accelerating global warming.

Nautilus Members enjoy an ad-free experience. Log in or Join now .

What we have not found is a sign that clouds are re-organizing themselves in some way as to lessen the net temperature rise. That does not mean that no such negative feedback exists. I have too much experience wrestling with the quirks and oddities of real data to set much store by our flawed observations. A change in satellite viewing angle here, a nudge of a measurement station there—these oversights add up. And clouds—those changing, shifting, vanishing, maddening things—are almost as difficult to observe as they are to model.

It seems unlikely, though, that clouds will completely nullify the effects of our carbon dioxide emissions. They may at most buy us time and delay the consequences of our actions. But they will not bring salvation.

John Muir was wrong. Nature, like my faraway computer, has no feelings or urges toward our preservation. It does not love us. Clouds may bear tidings of our ultimate fate, but we would be foolish to assume they are friendly.

Nautilus Members enjoy an ad-free experience. Log in or Join now .

Kate Marvel is an associate research scientist at NASA’s Goddard Institute for Space Studies and Columbia University’s department of applied physics and applied mathematics. She blogs about science at All views expressed are her own.

close-icon Enjoy unlimited Nautilus articles, ad-free, for less than $5/month. Join now

! There is not an active subscription associated with that email address.

Join to continue reading.

You’ve read your 2 free articles this month. Access unlimited ad-free stories, including this one, by becoming a Nautilus member.

! There is not an active subscription associated with that email address.

This is your last free article.

Don’t limit your curiosity. Access unlimited ad-free stories like this one, and support independent journalism, by becoming a Nautilus member.