JUNGFRAUJOCH, Switzerland -- On a clear day at the Sphinx, a legendary atmospheric observatory 11,000 feet up in the snowed-in peaks of the Bernese Alps, the blue sky runs down green hills and white glaciers toward seemingly all of Europe beyond.
On a lucky day here, though, there's only gray. There are only clouds.
From the Sphinx's rooftop terrace, scientists are conducting one front of a long-running campaign to sample cotton puffs of atmosphere. Their prey is shifty, yet tauntingly present. And much of the planet's future depends on what they find.
There are few places better than the 75-year-old Sphinx lab -- swaddled in gray nearly half the year, to the dismay of millions of people who have made the Jungfraujoch one of Switzerland's busiest tourist attractions -- to study clouds and the microscopic particles they form around, said Urs Baltensperger, an atmospheric scientist at the Paul Scherrer Institute who has worked at the lab nearly 30 years.
"We can just sit there and wait until the cloud shows up," Baltensperger said.
Baltensperger has seen cloud research move from an academic sideshow to the heart of climate science. Far from the domain of poets and dreamers -- though many scientists are a bit of both -- researchers are probing much that remains mysterious about clouds and how they've been changed by humanity's touch. They are testing clouds' resilience to the Anthropocene.
It is a demand driven as much by politics as necessity. A durable uncertainty remains in the science of climate change. Due to human-released greenhouse gases, global warming is happening. There is no serious disagreement. But the exact speed of that warming is a matter of debate. Under doubled CO2 concentrations, does the world simmer up by 2 degrees Celsius, or does it broil up by 4.5 degrees or more? Scientists can't truly say.
This uncertainty is driven, in large part, by clouds.
"This is the main source of uncertainty in the predictions of how much warmer it's going to get," said Ray Pierrehumbert, a climate scientist at the University of Chicago. "We can virtually rule out the idea that clouds could cancel out any global warming. ... But on the other side, there's almost unlimited possibilities for making it bad."
The range of warming estimates will be the next point of contention in the U.S. debate on climate change if the Republicans move off their reluctance to accept human-caused warming. It's a pivot already made by former Republican presidential nominee Mitt Romney, who cited doubt about the warming rate in September. And across the aisle, activists calling for action on climate change have grown weary waiting for certainty on the warming rate. Scientists, mulling over feedbacks, haven't moved quickly enough for their liking.
The cloud problem, as scientists call it, also makes specific warming targets, such as the 2 degree Celsius limit touted by leaders of the industrial world, a matter more of rhetoric than reality. But it's a message policymakers don't want to hear.
"We don't know even know the sign of how much [more] CO2 we can put into the atmosphere without committing ourselves [to 2 degrees]," said Stephen Schwartz, an atmospheric scientist at Brookhaven National Laboratory. The world could already be past that threshold, or it could have 40 years to go, he said. "No one really wants to come to grips with what we're saying."
Whatever the warming rate, it is a consequential global problem, he hastened to add.
"If it's at the low end, the consequences will be serious," Schwartz said. "If it's medium, they will be severe. If it's at a high end, they will be catastrophic. And I can say that with a lot of confidence."
The pressure to resolve the cloud problem has kicked what was once a sleepy corner of meteorology into the center ring of climate science. Ahead of next year's science report by the Intergovernmental Panel on Climate Change (IPCC), the influential U.N. summary, researchers have scrambled to advance their understanding. But behind each puff and wisp, they have only uncovered more questions. Which is why, much to their consternation, next year's report will not reduce the uncertainty in any tangible way.
"It's a major, major problem right now that we can't constrain it better," said Trude Storelvmo, a climate scientist at Yale University. "It's not given that we'll be able to constrain it any better. It's not obvious how we can go about constraining it better."
It's a reality that climate scientists have hinted at for several years. In 2010, Kevin Trenberth, a prominent climate modeler, reported that efforts to improve the predictive abilities of climate models would result in greater uncertainty. And this June, a pair of British scientists warned that more, not less, uncertainty is expected in the next U.N. report, presenting a serious public-image problem for scientists.
Some cloud scientists have grown impatient with the need to frame their work through the global warming rate, which is known in scientific circles as climate sensitivity. It's a reality that Graeme Stephens, the head of climate science at NASA's Jet Propulsion Lab, railed against in a lecture last year. The error in the sensitivity might not be reduced, but if you look past that, real progress is happening, he said.
Beyond the frame imposed by the political debate, scientists now understand clouds, and their response to air pollution, better than ever. Unknown unknowns are now known unknowns. Looking into the sky from observatories like the Sphinx, or down from new satellites above, new contours are visible. But they are not yet in focus.
And as any picnicker knows, a cloud can be interpreted in many different ways.
'The largest unsolved problem in physics'
Any story about the clouds must begin higher -- much higher -- with the sun.
Pouring energy on our waterlogged planet, the sun is the ultimate cause and arbiter of cloud behavior. Its energy arrives as visible light, and returns to space as reflected light and heat. But along the way, it manifests in many forms. To scientists, all weather is simply a matter of the sun's energy flows. This includes clouds.
Clouds arise as a way of returning the sun's heat upward. When light strikes the Earth's surface, some of it goes into evaporating water. This vapor rises in patches of shifty humid air -- thermals, plumes and vortices -- that mingle with atmospheric particles as they go up. Swimming through cooling air, the water vapor then reaches a saturation point, phase shifting back to liquid water in a burst of energy. Voila, cloud.
It almost seems simple. But this brave cloud -- call it a cumulus, the popcorn cloud that has decorated many a nursery wall -- rises in a mixed, turbulent world. The same solar energy that created the cloud strikes the tropics and poles in uneven rates, an imbalance corrected by planetwide patterns of circulation. These winds flow over the Earth's craggy, uneven surface -- mountains here, plains there -- in the boundary layer, the rough-and-tumble bottom of the atmosphere that holds most of the weather. It's a realm familiar to most modern travelers. Jetliners fly just above it.
The turbulence of the boundary layer lies at the heart of cloud science. Unlike the global uniformity of greenhouse gases, clouds are unpredictable and heterogeneous. Studies made in one system can't be applied to another. All clouds, in short, are local.
"Historically, scientists have called turbulence the largest unsolved problem in physics," said Joao Teixeira, the deputy director of the Jet Propulsion Lab's climate science unit. "That problem is compounded with clouds."
Still, there is much scientists do understand. At any point in time, about 70 percent of the planet's surface is shadowed by clouds. They come in many forms. There are low decks of stratus and stratocumulus clouds, reflecting more energy than they retain. There are puffy cumulus clouds, buoyed by convection, which can strengthen into stormy weather; they may be a wash for the climate. And strung high in the atmosphere are cirrus clouds, wisps of suspended frozen water that likely warm the planet.
On their way to winking into existence, clouds often have help. At lower elevations, the atmosphere is suffused with suspended microscopic particles called aerosols. Smog, smoke, haze -- these are all aerosols. Humans are often a source of the particles, through the burning of fuels and forests, or the dust of desertification. Aerosols reflect sunlight on their own, and the water often attached to them provides a kernel for clouds to form around. One could even call them something evocative, like cloud seeds.
Scientists, of course, call them nuclei.
It's unclear how much sunlight clouds and aerosols reflect back into space. That dual uncertainty is why scientists have not been able to extrapolate from measures of current warming to the climate change awaiting the future. (The uniform physics of greenhouse gases is a much easier problem.) It could be aerosols blocking the light. It could be clouds. Two variables are missing, and there's only one equation to solve.
It's a problem neatly sketched out by Paul Zieger, a postdoc under Baltensperger, on the three-hour train ride to the Jungfraujoch one sunny morning this summer. It's a packed near-vertical ascent, standing room only, flush with Japanese and Indian tourists. The train passes cows seemingly placed by Swiss tourism. The air grows thinner.
Among scientists, Zieger continued, there's a famous bar chart in the IPCC reports. It depicts the current human influence on the climate, both warming and cooling. Almost all the error comes from aerosols and their influence on clouds.
"It is always our motivation," he said. "[There's] no Ph.D. defense without this plot."
Solving the aerosol-cloud interaction won't end the uncertainty in climate sensitivity stemming from the response of clouds to higher temperatures. Aerosols are more a derivative of the larger problem. But the field presents tractable problems, rather than a fundamental question of physics, which is perhaps why it has attracted a lion's share of cloud work over the past decade, including at the Jungfraujoch.
Perched on top of the Sphinx, a heated inlet, barely taller than a scientist on tiptoe, sucks in air day and night. In the lab below, it splits like a mutant octopus, carrying cloud remnants to a host of instruments, all part of an effort to resolve at least one of these aerosol variables. It's careful, meticulous work that takes advantage of the lab's position above the turbulent boundary layer in pristine air.
Pristine, that is, when the railway company isn't blowing up the mountain for its tourists.
White wine, cheese fondue and Nutella
Before there was the Sphinx, there was the Jungfraubahn.
Chasing the mountaintop sublime for the common man, more than 100 years ago Adolf Guyer-Zeller, a Swiss industrialist, proposed to blast a railway through the Bernese Alps to the Jungfraujoch, an anticline between the peaks of the Mönch and Jungfrau.
Since the mid-1800s, when Alpine tourism took off, the Swiss had puzzled over plans for the region -- one engineer proposed a 15-minute ascent through a series of pneumatic tubes -- only to be foiled by economics or human frailty to pressure changes. Guyer-Zeller eventually won the permit, but with a concession that his firm, the Jungfraubahn, must support research. It was the start of what would be a troubled but productive marriage.
In the 1930s, scientists found a permanent home on the mountain, with the construction of a 10-bedroom research station and then, up an elevator, the Sphinx, topped with a metallic astronomic cupola. Tended by two couples, the station has a rustic charm; much of its wooden furniture remains from the 1930s. Black-and-white portraits of grim bearded men, past custodians and scientists, hang on the walls. Several of the framed scientists fell to their death down to the glacier below, but none since the 1950s.
On a recent visit, the station was quiet, removed from the bustle of the Top of Europe tourist site attached to it. During research campaigns, though, the station is packed. Young researchers haul experiments, Feldschlösschen beer and fondue up on the train. Work on the Sphinx runs 24 hours a day, with laptops scattered across the station's antique furniture. (It also has one of country's fastest Internet connections.) There may be occasional inner-tube sledding, but mostly it's frantic work in rough conditions.
Thank goodness for the fondue, Zieger said.
"You can eat as much as you want and you'll never become fatter, because we burn so much calories," he said. The station is stocked with white wine, ancient sugar and Nutella. "It's so exhausting being up there in that air. ... You're always very hungry. So you eat tons and tons of chocolate and cheese fondue, whatever."
It seems an idyllic place for cloud science, and often it is. But not always.
Long a tourist attraction, Jungfraujoch has seen visits explode in recent years.
One of the largest employers in the region, the Jungfraubahn made extensive expansions to its tunnels and restaurants in advance of its centennial. There are slick video rooms; an Indian restaurant for Bollywood fans; a series of ice caves filled with sculptures; and even a human-sized snow globe seemingly transplanted from the "It's a Small World" ride at Disney parks, complete with gondolas and alphorn blowers.
It all sounds innocuous enough, but it takes a toll. The railway detonated a new tunnel in the mountain for its anniversary, and the Sphinx researchers had dozens of false alarms for dust blowing in from the Sahara, for example. Helicopters shuttle wealthy tourists up the mountain, releasing air pollution. And the rare tourist will even sneak a smoke, which is about the worst thing a visitor could do to the scientists' records.
"It's a working collaboration, but it's quite difficult," Baltensperger said.
Two glass elevators lift tourists to the Sphinx, where they visit the lab's ground floor and buy high-end Swiss watches from glass kiosks. It's a crowded ride for Zieger; elevators are routinely overloaded, bleating with electronic dismay. Swerving through the crowd with practiced ease, Zieger opens a gray side door, leading to the real lab above. Up past a toilet armed with perhaps Europe's best view, there is the typical scientific clutter, and then a reinforced door, where the inlet's pipes unfurl. The lab is overpressurized, in an attempt to keep out pollution from the elevators below.
There's a blast of mechanical whine as Zieger opens the door. Cloud science isn't quiet.
Zieger slides from machine to machine. Here an aerosol mass spectrometer. There a scanning mobility particle sizer. (The former measures aerosol composition, the latter the size distribution.) On a screen, Zieger sees that a colleague is remotely testing an instrument. He moves the computer's cursor around with glee, a ghost in the machine.
The Sphinx is one of the leading stations of the World Meteorological Organization's Global Atmosphere Watch program. Few, if any, sites record a greater diversity of data about clouds and aerosols. It's difficult to keep it all going, Zieger said, pausing in his tour.
"The most challenging thing is keeping all these [machines] running," he said. "It's looking for losses. It's looking for good calibration. And this costs a lot of effort. And it costs a lot of money."
At the lab's corner, there's a rectangular white box attached to several bottles: the cloud condensation nuclei counter. It's made by a company called Droplet Measurement Technologies.
"The aerosols get in here, and then you have a wetted tube that is heated and cooled," Zieger said. "Then you expose particles to supersaturated air. And then particles grow and they're activated. And then you count them."
Just five years ago, there was no standard device for counting cloud nuclei. There was a hodgepodge of home-built devices. Now the commercial instrument is becoming common across research stations, allowing global comparisons. At the Jungfraujoch, it's enabling tests of the maximum saturation a cloud can experience, helping to determine the critical particle size for a cloud's formation.
It all seems promising, at least on a lucky day.
But the Jungfraujoch is just one site. Alone, it won't have all the answers.
'Is the lifetime effect even real?'
Far from the Alps, at sea, the pollution's influence on clouds has long been clear as day.
Look at any photo of the low, flat clouds that run along America's Pacific coast, and bright cotton-ball streamers appear. These are the tails left by ships, steaming across the ocean and releasing pollution -- aerosols -- as they go. The higher number of cloud nuclei allows smaller, brighter cloud droplets to form, reflecting more sunlight into space.
It's called the albedo effect by scientists, or the Twomey effect, after the man who popularized it. (For true wonks, it's also called the first indirect effect.) There's little doubt it's real. Early on, it was easy to add into climate models: Simply make the cloud droplets over aerosol-emitting continents smaller and see what happens. The models have improved recently, getting actual physics into them. It remains the only cloud effect that the IPCC has tried to estimate.
If there's certainty in aerosol and cloud science, it's that the albedo effect exists, at least in some cloud systems, said Graham Feingold, a scientist at NOAA's Earth System Research Lab who has spent the past decade knitting together theories on cloud formation, informed by the surge of data from satellites, radar and aircraft, including the first ground-based observation of the albedo effect.
"Nobody argues with the fact that more particles means smaller droplets," he said.
Still, there is plenty of debate -- and "effects" -- to go around. For instance, 15 years ago, NASA's James Hansen identified a new aerosol role, the semidirect effect. These aerosols live in clouds but don't act as nuclei. They can heat clouds and essentially burn them off. The vaporized clouds could then stabilize the atmospheric column and reduce cloudiness. Or they could destabilize the column and induce cloudiness.
"It depends on the height," Yale's Storelvmo said.
But the hot debate now, she added, is over the science's other pillar -- the lifetime effect.
"That's a controversy right now," she said. "Is the lifetime effect even real?"
The lifetime effect connects back to the cloud seeding experiments of the 1940s and builds off the albedo effect in a simple way. If more aerosols cause smaller water droplets in clouds, then won't these droplets collide less, causing less rain and, ultimately, more clouds to persist in the sky? It's an elegant theory, and one that was picked up by much of the modeling community without question.
That has started to change in recent years, partially due to an influential paper published by NOAA's Feingold. It's time to stop looking at clouds as single entities, he said. They are part of a system. When aerosols are added, sure, there may be less rainfall, at first. But that cloud may grow deeper and darker, predisposing it to even more rain than before. In effect, the cloud buffers its response to human pollution.
"One could think about this as the resilience of the cloud system," Feingold said.
Feingold's theories echo what has been a deficiency in how climate models rendered aerosols. In the past, the programs simulated how pollution could change a cloud, but not vice versa, said Andreas Muhlbauer, a climate scientist at the University of Washington.
"One thing that's been neglected is the impact of clouds on aerosols," he said. This was a topic of much discussion at a global modeling workshop this summer in Poland. "If precipitation kicks in, it can clean out the atmosphere."
There have been reports this summer out of the newest suite of climate satellites that bolster Feingold's case, if not necessarily his theory. For the first time, these satellites, collectively known as A-Train, allow scientists to make nearly instantaneous snapshots of rain, aerosols and clouds. Combining A-Train data with a traditional satellite, French researchers found support for the albedo effect, but not much for the lifetime effect. Similarly, a study published in August found evidence for a weaker lifetime effect than estimated by climate models alone, a result echoed by NASA's Stephens, one of A-Train's leaders, in remarks earlier this year.
There's still much more to come from the satellites, Stephens added in his lecture last year.
"Raining clouds are a hell of a lot brighter than nonraining clouds," he said at the American Geophysical Union's annual meeting last December. "Precipitation has an important radiative signature that we haven't really considered in feedbacks."
Still, as these studies admit, satellites and climate models alone won't crack the aerosol and cloud conundrum. It remains maddeningly difficult to quantify aerosols on a global level. Each method of study has its drawbacks. The Jungfraujoch captures a cloud only as it passes; a plane catches only a pencil line of its prey. Or take satellites. They may snap only two pictures a day of a cloud.
"Someone once likened this to someone completely unfamiliar with the rules of soccer getting snapshots twice a game," Feingold said. "And after the fact trying to figure out what the rules of the game are."
To get at aerosols, scientists are going to have to find a way to unify this data: to link the small scales crucial to cloud dynamics with the large scales of the climate. Much of this is a question of statistics and scale, and gets arcane quickly. But Feingold and Allison McComiskey, his co-worker and a geographer, are getting at the question, developing a method to represent strong but variable aerosol effects at coarse global scales. If it is successful, it could be reproduced by scientists across the planet.
Clearly, it's going to take some Herculean science to sort it all out, Storelvmo said.
"It may look like we've made no progress, but that's far from being true," she said. "It's a huge community working on cloud and aerosol interactions. We've made a lot of progress. But it hasn't resulted in a narrower spread of estimates."
'The biggest uncertainty'
Back in Switzerland, if there's one thing cloud scientists love, it's a good YouTube video.
The other giant of aerosol research in the country, Ulrike Lohmann of ETH Zurich, has a favorite video she shares with people curious about clouds. It's from a past research trip to the Arctic Ocean. Outside on the icebreaker Oden, her researchers steep tea in a mug. Despite the tea's heat, there's no steam. The Arctic air is so pristine -- for now -- that there are no particles to help the water condense.
"If I have a cup of tea, I don't see anything at all," she said.
Flick a lighter next to the Arctic tea, though, and presto, instant cloud.
There will be no lighter to flick on and solve the aerosol uncertainties, but the next IPCC report will show progress, Lohmann said. It will consider a new way of calculating all the warming and cooling forces caused by humanity, moving from calculating only instant forcings on the climate to fast feedbacks that respond within a couple of years. It's an idea she's championed, and one that will allow the report to estimate, for the first time, the lifetime and semidirect effect, among other influences.
There will also be, for the first time, an estimate for clouds made of liquid water and ice.
"The ice is coming," said Lohmann, who has migrated much of her work to such clouds.
Baltensperger, the longtime Jungfraujoch researcher, is equally excited by these clouds. In January and February, the Sphinx will host its next scientific campaign, with researchers from across Europe hauling several tons of temporary instruments on the train and breaking out blowtorches to melt ice formed overnight on their tools. These instruments will be specifically trained on water-and-ice clouds.
"The biggest uncertainty in this whole indirect effect is the mixed-phase clouds," he said.
Yes -- more uncertainty. There's a stumbling point any scientist, or anyone, can reach with clouds and climate change. It seems intractable. It's tempting to move on. Many do.
"It's almost a truism that we need to understand clouds better," Chicago's Pierrehumbert said. "People have known that clouds are a problem for 40 years. It's been incremental. It's not that exciting. It's frustrating."
But still, there are moments of deep reward to the work.
Clomping up the stairs to the Sphinx's terrace, his winter jacket zipped high for the summer chill, Zieger, Baltensperger's postdoc, steps out into the sun on a clear, windy day. The Mönch and Jungfrau tower above, and clouds lick off their peaks. It's a sea of cumulus over Switzerland. Wind bites skin, and fades. Turbulence comes and goes. The inlet sucks in empty air, waiting for the next cloud. The wind starts again.
Zieger will return for the winter campaign, he said, even if he's turning his work toward the Arctic. It's just not a time to be missed. On those nights, when the mountaintop empties and the 5:30 train departs, it's just a few scientists and their clouds.
"There's no possibility for tourists to stay overnight," he said. "There's a hut, for Alpine climbers, which is about 45 minutes from the station. After a while you're absolutely alone up there."
Next: Turbulence, sensitivity and a bit of MAGIC.
Clarification: An earlier version of this story stated that the planet's future warming is uncertain under "doubled emissions" rather than uncertain under CO2 concentrations doubled from the preindustrial era.