sealPurdue News

Study rethinks atmospheric chemistry from ground up

WEST LAFAYETTE, Ind. -- Don't be snowed by everything the textbooks have to say about atmospheric chemistry.

Download Photo Here
Photo caption below
A Purdue University research team studying natural processes that affect ozone in the Arctic atmosphere has discovered that snowpacks not only absorb chemicals from the atmosphere, but also can help produce them.

The findings, published in the March 18 issue of the scientific journal Nature and the March 15 issue of Geophysical Research Letters, cast a new light on scientists' perceptions of how atmospheric gases are processed, says Paul Shepson, professor of atmospheric chemistry at Purdue.

The new findings also may affect the way that scientists view data from ice core studies, because researchers have assumed that the air trapped in ice provided representative samples of atmospheric conditions at the time the ice was formed.

"Ice core studies designed to look at reactive species such as nitrates may have to be revisited, as the air bubbles found in these ice cores may not be the mirrors of atmospheric composition that we suspected they were," Shepson says.

This is not a concern for more stable greenhouse gases such as carbon dioxide and methane, which have been extensively studied in ice cores, because these stable gases are less likely to react with other compounds in snow or ice, Shepson says.

His group studies the chemistry of ozone in the troposphere, the lowest part of the atmosphere. Ozone, a beneficial component of the earth's upper atmosphere, is a pollutant at the ground level.

Last winter, Shepson led a research group to the Canadian Arctic to observe how sunlight interacts with various gases in the atmosphere to reduce near-surface ozone levels.

"It has recently been observed that, at polar sunrise, which occurs in March or April after several months of complete darkness, ozone in a thin layer of air over the Arctic ocean is completely removed," Shepson says. "This was a big surprise to us, and it indicates that our understanding of atmospheric ozone is poor."

From the Environment Canada research site at the Canadian Forces base at Alert, the group tracked levels of atmospheric compounds, including formaldehyde, over a two-month period. Formaldehyde is an important part of the atmosphere's self-cleaning mechanism because it is a major source of free radicals, Shepson says.

"The atmosphere acts to clean itself of pollutants through reactions involving free radicals. When formaldehyde absorbs light, it falls apart to produce these free radicals."

Previous studies of formaldehyde in the Arctic had shown concentrations up to 10 times higher than expected, so graduate student Ann Louise Sumner spent two months at the Alert laboratory measuring formaldehyde in the snowpack and in the atmosphere.

These measurements, published in the Nature article, suggest that formaldehyde is produced through photochemical reactions at the snow surface.

"The data account for much of the discrepancy between the high concentrations of formaldehyde found in the Arctic and the amounts predicted by our models," Shepson says.

The second paper, published in Geophysical Research Letters, reports on studies at the ice core site at Summit, Greenland, where the Purdue group participated in an experiment led by Richard Honrath of Michigan Technological University.

The studies found further complexity and importance in photochemical processes that occur at the snow surface, Shepson says.

Specifically, the team found that concentrations of nitric oxide and nitrogen dioxide -- collectively known as NOx -- were actually higher within the snowpack than in the atmosphere.

The findings suggest that nitrate ions in the snow can interact with sunlight to produce NOx, a pollutant derived largely from the combustion of fossil fuels and a critical precursor to the production of ozone the atmosphere, Shepson says.

"This observation changes the way we look at atmospheric chemistry in a fundamental way, in that deposition of nitric acid to the snow was previously regarded as the final fate of NOx," he says. "Now it appears that nitric acid in the snow can be reprocessed by interactions with light, causing re-release of a variety of pollutants back into the atmosphere."

In addition to forcing a re-evaluation of data from ice core studies, the new findings call into question some models that are used to predict long-term changes in the composition of our atmosphere.

"Specifically, models of atmospheric chemistry need to do a better job of treating interaction of gases with surfaces," Shepson says. "Although we are starting to do better with atmospheric particles, it is important to remember that a potentially important atmospheric surface is the surface of the earth."

Shepson and his group are working with another group at Purdue to develop new computer models that incorporate the chemical reactions that occur in snowpacks into the current models of atmospheric chemistry and transport.

Working with Shepson on the studies are graduate students Bryan Splawn, a native of Spartanburg, S.C., Sumner of Lake in the Hills, Ill., and Brian Michalowski of Racine, Wis.

Shepson's studies at Purdue are funded by the National Science Foundation and BASF.

Sources: Paul Shepson, (765) 494-7441;
Ann Louise Sumner, (765) 496-2404;
Writer: Susan Gaidos, (765) 494-2081;
Purdue News Service: (765) 494-2096; e-mail,

NOTE TO JOURNALISTS: Copies of the journal articles are available from Purdue News Service, (765) 494-2096.

Photo caption:
Ann Louise Sumner, a doctoral student in atmospheric sciences at Purdue, collects a sample of snow and ice crystals from the frozen surface of the Lincoln Sea, located in the Canadian Arctic, to measure formaldehyde levels in the snow. (Purdue News Service Photo courtesy of John Deary of Environment Canada )
Color photo, electronic transmission, and Web and ftp download available. Photo ID: Shepson.Arctic

Download Photo Here

ABSTRACT: March 18 Nature
Snowpack production of formaldehyde and its impact on the Arctic troposphere
Ann Louise Sumner and Paul B. Shepson

Efforts to understand the details influencing the dramatic surface level ozone (O3) depletion events in the Arctic at polar sunrise continue since the first reported observations of the phenomenon. Ozone is destroyed, as proposed by Barrie et al, through a chain reaction with bromine atoms (Br). Since the chain terminates by Br reaction with formaldehyde (HCHO), the influence of HCHO on the cycle must be understood. Previous reports of ambient HCHO concentrations, which are larger than current gas phase models can simulate, are perplexing, given the short HCHO lifetime after sunrise. However, our recent measurements of snow and gas phase formaldehyde in the Arctic suggest photochemical HCHO production at the air-snow interface, resulting in a substantial flux out of the snowpack. This accounts for much of the discrepancy between ambient data and model outputs. Production of free radicals and carbon monoxide (CO) from photolysis of HCHO from this newly discovered source is potentially important to the oxidative capacity of the polar troposphere. This observation will force the re-evaluation of our current understanding of Arctic ground-level ozone depletion events. Snow surface photochemistry may complicate the interpretation of ice core records for reactive species including hydrogen peroxide (H2O2), HCHO, and CO.

ABSTRACT: March 15 Geophysical Research Letters
Evidence of NOx Production Within or Upon Ice Particles in the Greenland Snowpack
R.E. Honrath, M.C. Peterson, S. Guo, J.E. Dibb, P.B. Shepson and B. Campbell

NOx and NOy were determined in the interstitial air of surface snow and in ambient air at Summit Greenland. NOx levels in interstitial air were 3 to more than 10 times those in ambient air, and were generally greater than ambient NOy levels. [NOy] in interstitial air varied diurnally in a manner consistent with photochemical generation within the snowpack. These observations imply that photochemical reactions occurring within or upon the ice crystals of surface snow produced NOx from a N-reservoir compound within the snow. Average [NOx]:[HNO3] and [NOx]:[NOy] ratios in ambient air above the snow were elevated relative to other remote sites, indicating that NOx release within the snowpack may have altered NOx levels in the overlying atmospheric boundary layer. We suggest that the observed release of NOx may have been initiated by photolysis of nitrate, present in relative abundance in surface snow at Summit. Such a process may affect levels of nitrate and other compounds in surface snow, the overlaying atmosphere, and glacial ice, and its potential role in cirrus cloud chemistry should be investigated.

* To the Purdue News and Photos Page