Water where there should be ice. New data on high clouds forces scientists to rethink `greenhouse effect'

When scientists try to estimate the climatic impact of polluting gases, such as the carbon dioxide (CO2) released by burning coal and oil, they also have to take account of the role of clouds. It now seems that they may have misjudged the influence of some of the most climatically important clouds of all -- the high clouds known as cirrus. CO2 acts to warm the planet by absorbing heat radiated outward by Earth's surface and reradiating much of it back downward. This is the so-called greenhouse effect. Clouds also affect the planet's radiative energy budget by reflecting sunshine from their tops, by trapping outgoing heat radiation, and by radiating energy downward.

In estimating the effect of cirrus, scientists had assumed as a matter of course that the clouds are so high and so cold that they consist entirely of ice crystals. Meteorologists used mathematical approximations of the standard hexagonal ice crystal shape to simulate the influence of the cloud particles on Earth's energy budget. Now it appears that this assumption may have led them astray.

Last year, Kenneth Sassen and Kuo Nan Liou of the University of Utah reported finding significant amounts of water in cirrus. Computer simulations they have subsequently made now show that it is important to take the water into account to estimate correctly the role of cirrus in regulating Earth's climate.

This work, which they carried out together with Stefan Kinne and Michael Griffin, is yet another illustration that computer models that attempt to project the climatic effect of CO2 or other pollutants have been oversimplified.

The Utah scientists detected the water drops in cirrus cloud with ground-based laser radar, called lidar. Then an aircraft from the National Center for Atmospheric Research in Boulder, Colo., confirmed the water drops' presence in concentrations as high as 100 to 130 drops per cubic centimeter. They tended to occur in the bottom of the cloud layer in a zone about 100 meters thick.

The measurements, which were made on cirrus over Boulder on Oct. 17, 1983, were probably the first detection of cloud water drops as cold as -36 degrees C. Water can remain liquid well below the nominal freezing point, especially if there are no nuclei, such as the salt particles used in efforts to seed clouds, around which ice crystals can form.

At -40 C., however, water drops generally freeze spontaneously even without nuclei. Cirrus clouds, typically forming 6,000 to 12,000 meters (20,000 to 40,000 feet) high, have been thought to be too cold to have supercooled water drops. As the Utah scientists observe, ``it seems implicit in the definition of cirrus clouds [as the highest and coldest lower atmosphere clouds] that liquid water, in any measurable amounts, should be absent.''

But surprising or not, the water is there. And it significantly changes the clouds' impact on Earth's climatically important energy balance.

Reporting its analysis in the Jan. 25 issue of Science, the Utah team notes that ``water-containing cirrus clouds can significantly enhance the atmospheric warming between the surface and the cloud base and, at the same time, can produce a cooling of the surface'' by blocking some of the downward flux of solar energy. Their computer simulations show an almost twofold increase in the atmospheric heating rate at the cloud base when a thin (100-meter-thick) water-drop layer is added at the bottom of their 2-kilometer-thick, mathematically simulated cirrus cloud. This addition also enhances the cooling of the cloud top by 3 degrees C. per 24 hours, because of its effect on the cloud's overall energy balance.

``It is clear,'' the researchers say, ``that the presence of even a thin layer at the cirrus cloud base will modify the fundamental radiative properties of the cloud . . . and will therefore have an impact on attempts to model the climate of the earth-atmosphere system.''

As the Utah scientists point out, ``it is now of paramount importance to obtain climatological data on the frequency and distribution of liquid water in cirrus clouds.''

Is this a major, and hitherto unrecognized, climatic factor? Or is it merely a localized and sporadic influence?

A research program using ground-based lidars should be able to provide answers. Meanwhile, computer simulations of climate and forecasts of future trends should be recognized for what they are -- a technological tour de force that still has too many uncertainties and simplifications to provide a reliable basis for assessing the long-term effect of CO2 or other atmospheric pollutants.

A Thursday column. Robert C. Cowen is the Monitor's natural science editor. -- 30 --