The Curious Case of Cloud Cover Consistency

Monday, February 8, 2010

Clouds Raise Many a Question, Including the Curious Case of Cover Consistency

"I don't know clouds at all" is the refrain that runs through a late 1960s popular song by Joni Mitchell. Her meaning is meant to be metaphoric—how little one knows of the seemingly familiar. But for scientists at the KITP program pioneering the "Physics of Climate Change," the meaning is far more literal than metaphoric. "Cloud physics is the largest source of uncertainty in the short term in predicting the climate," according to the program's principal organizer, Brad Marston of Brown University.

Clouds can function both as blankets and as mirrors with, respectively, either a warming or a cooling effect.

Wispy cirrus clouds, trailing high above the Earth, have a net warming effect because they reflect back down heat or infrared radiation emanating from Earth. Sunlight at visible wavelengths mostly passes through high cirrus clouds. Low-lying stratus clouds, the usual culprit of cool "cloudy" days, especially off the coast of California, have a net cooling effect because they reflect light back into space, and their behavior is among the hardest to model though modeling any cloud turns out to be very difficult.

What makes cloud modeling hard is the wide range of relevant length scales—from the microphysics of droplet formation and agglomeration that can lead to "rain" to the vast turbulent motions in the Earth's atmosphere which we call "wind."

Rain is the result of a two-fold process of condensation and collision inside a certain kind of cloud termed "cumulus" [from the Latin meaning "pile up"]. First is the process of nucleation whereby seeds of dust in the atmosphere augment the attraction of water molecules into drops. Turbulent movements within clouds foster collisions of drops that aggregate into droplets heavy enough to be gravitationally attracted to Earth.

How Raindrops Form

"It seems now in both these processes—condensation into drops and collisions between them—turbulence in clouds plays an important role," said Gregory Falkovich of the Weizmann Institute in Israel. "Small-scale inhomogeneities in the vapor concentrations determine the growth of properties due to condensation and strongly influence the collision rate of droplets."

Physicists attending the "Climate Change Program" tackled the problem of constructing equations that describe these processes of condensation and collision within the turbulent cloud medium.

"What is wanted is a theory that quantitatively describes these two phenomena," said Falkovich. "If we look at warm clouds and even if we know everything about them in the beginning, we cannot predict how fast those clouds will precipitate because the process of precipitation depends on turbulence, and turbulence in clouds is a function of both an internal process of convection and an external process of macro-turbulence, which happens on scales of 10s and 100s of kilometers."

In other words, there are two processes of movement within clouds: particle movement due to local phenomena, but also larger movements of air currents having to do with winds on larger scales.

"Clouds," asserts Falkovich, "comprise the biggest unknown of the various feedbacks in the process of global warming."

He points to another of the unsolved problems pertaining to clouds: the mysterious consistency of average global cloud coverage, according to recent observations, even though other observations indicate that the global climate is not only changing via Earth warming, but is likely changing relatively fast.

Cloud cover has changed very little since cloud cover has been being observed. That "little" change refers to an average cover area, not fluctuations from day to day because, as anybody knows, some days are cloudy in some places sometimes and not so in the same place at other times. And there are patterns of regional coverage, which can change year to year. But the average cloud cover over large portions of the globe is surprisingly stable, said Falkovich.

"We don't understand it at all, and it is very important because it affects the albedo [reflectivity index] of our planet and thereby the extent of the greenhouse gas effect. Cloud coverage seems to be stable, and we don't understand the physical mechanism behind this stability.

"There must be some negative feedback," he speculates, "while at every given part, the system is strongly fluctuating."
Because the mechanism for this cloud cover consistency is not understood, scientists do not understand if it could be broken and how stable it will be in the future.

Gregory FalkovichThe view from space indicates what percentage of the surface is cloud-covered, and the height of the clouds and therefore their albedo. "Those parameters," said Falkovich,  "stay surprisingly stable over a period of years. I as a theoretician don't understand why those parameters are stable over the scale of a few years. I don't really understand clouds."

Global circulation models constructed to enable an understanding of the effects of global warming take into account the effects of clouds in, according to Falkovich, a "hugely oversimplified" way. A more realistic model, he contends, would at least take into account the differing effects of clouds at different heights and the concomitant role played as blankets or mirrors.

"We don't know, he said, "the effect of cranking up the temperature on cloud cover. We don't even know if the sign is negative or positive." That is the difference between feedback that enhances or diminishes the effect.

Modeling global circulation requires "parameterization," a mathematical process involving the identification of a set of effective coordinates or, as physicists say, "degrees of freedom" because one cannot describe all the smallest scales.

But, according to Falkovich, a long-time turbulence expert, who has recently been applying its insights to clouds, "Turbulence resists scaling. The statistics of turbulence change as you change your resolution. Understanding turbulence on the scale of kilometers doesn't mean it is understood on a scale of meters. The behavior of fluid flows," he emphasizes, "depends on the scale of resolution." In other words, a larger scale reveals more flow; and a smaller scale, more detail but the two pictures cannot be superimposed on one another to yield a composite because each picture is a function of its scale.

Turbulent phenomena are both random, on the one hand, and show patterns, on the other hand. That dual oxymoronic character of turbulence is "precisely what makes it resolution dependent," said Falkovich. "When you blur the resolution, you stop seeing the patterns. The climate is like turbulence—patterns and randomness."