Amount of dust, pollen matters for cloud precipitation, climate change

The amount of ice crystals necessary to form precipitation in clouds is linked to the abundance of larger aerosol particles in the atmosphere, according to a study by Paul DeMott and Anthony Prenni, research scientists in the Atmospheric Science department at Colorado State. Their findings appear in this week's issue of the Proceedings of the National Academy of Sciences.

Using these new findings, a global climate model predicted that clouds have a stronger cooling effect on the globe than previously estimated. However, future increases in these ice nuclei for cold clouds would reduce the cooling impact on climate and vice versa, the scientists found.

Special particles called aerosols -- resulting from desert dust, some biological processes and possibly from pollution -- are needed as catalysts to form ice in clouds, which can influence precipitation and cloud dynamics. These particles can serve as the center, or nuclei, for cloud droplets that combine to form raindrops.

"The catalysts for most ice nuclei are primary emissions -- from pollution or sea spray or dust," DeMott said. "The bigger the particles, the better it is for ice nuclei."

At the same time, pinpointing a number of particles at a specific temperature is too simple for climate models to accurately represent what's occurring in nature, DeMott said.

Scientists have spent decades trying to understand the processes. The National Science Foundation, the National Oceanic and Atmospheric Administration, the U.S. Department of Energy and NASA have funded Colorado State's research in this area.

DeMott and Prenni analyzed data from 14 years of trips across the globe from the Amazon Rainforest in Brazil to the Arctic to Broomfield, Colo., to collect air samples in specially equipped National Center for Atmospheric Research planes. The Colorado State scientists also developed the first instrument -for use inside the plane -- to take continuous air samples from in and around clouds and measure in real time the ice-forming ability of particles. The instrument allows the researchers to sample air and detect the total number concentrations of ice nuclei without first putting them on a filter or other processing.

How capturing air from a plane works: CSU scientists take air samples into a small chamber through a special port on the side of a C-130 plane. A diffusion chamber cools and humidifies the air and particles between two plates of ice toward conditions where ice forms, essentially "growing" clouds by simulating the conditions in the atmosphere. Researchers then evaluate how many particles will form ice crystals for specific cloud conditions. The plane then passes through the wave clouds to measure, with other instruments, how much ice really forms.

Scientists also used specialized instruments to determine the chemical makeup of the particulates forming ice.

"Ice nuclei are hard to measure -- they're microns in size like the size of a bacteria," Prenni said. "They don't make haze -- there aren't enough of them. Of all the particles in the atmosphere, one in a million particles in the atmosphere can cause ice to form."

In March, Prenni and DeMott published an article in Atmospheric Environment that examined the role biological particles -- from plants, bacteria or other living things on Earth -- play in characterizing atmospheric concentrations and types of ice nuclei. They concluded that much more work needs to be done in tandem with biologists to determine numbers and sources of these particles as a function of season and temperature range.

"The people who look at snow and find these bacteria in it don't know if the bacteria were in fact the ice nuclei or how many of them there are floating around in the air in various places/seasons," said Sonia Kreidenweis, professor of atmospheric science who works with DeMott and Prenni. "There could be too few to matter. We are actually making these measurements in the air to try to nail this down."

"We don't know if we can identify all the biological particles," DeMott said. "What are the most effective ones? Their amounts matter as well. Is there any way that they play a role in cloud processes?"