Monday 22 October

Ice, oceans, forests and deserts; in fact everything on the planet, even penguins, reflects light to some degree. And according to University of Washington glaciologist Professor Stephen Warren, how much of the earth’s surface it covers will determine its effect on climate.

Stephen and University of Washington PhD student Maria Zatko are here on the SIPEX-II voyage to measure the ‘albedo’ — the fraction of light reflected — of snow and different types of ice in the spring sea ice zone of East Antarctica. These measurements, in combination with estimates of the area covered by each type of ice in different seasons, will contribute to the development of climate models.

'For climate, the albedo of everything is important, but the polar regions have the most variable albedos,' Stephen says.

Stephen and Maria use a ‘spectral radiometer’ to measure light in the ultraviolet, visible and near infra-red wavelengths (from 350nm to 2400nm). The pair look for areas of young ice and newly forming ice in leads between ice floes, either on foot or from a rubber boat.

The radiometer’s fibre-optic guide is mounted on a tripod and extends out over the ice of interest. The instrument has a ‘diffuser plate’ to ensure that light entering its sensor is evenly distributed. It then measures the amount of light at different wavelengths coming from the sky and bouncing back from the ice.

'First we turn the diffuser plate up toward the sky so we get a reading of the spectrum of light that’s falling on the ice surface. Then we turn the diffuser over and get a reading of what’s being reflected back from the ice. By dividing the two results we get the fraction of light reflected, or the albedo,' Stephen says.

But it’s not quite that simple. Clouds absorb some light wavelengths, so Stephen and Maria have to account for that, and on clear days the angle of the sun can influence the reading if the diffuser plate is not precisely level.

The pair have measured the albedo of newly forming ‘nilas’ ice, slush and thin ice with a covering of frost flowers — feathery ice crystals coated with a salt solution that significantly brighten the ice surface. They take about 10 measurements over five minutes and then average them for the final result. If the light or cloud cover changes while they’re out, they’ll take another 10 measurements.

'The dark ocean has an albedo of about 7%, nilas ice is 10 to 20%, depending on how thick it is, and a covering of frost flowers increases this to 30%,' Maria says.

Not surprisingly, snow has a very high albedo of 80%, with most of the light reflected in the visible part of the spectrum. However, impurities in the snow, particularly soot and dust, can affect the albedo.

'Impurities are significant for climate in the northern hemisphere but measurements on the Antarctic continent show the impurity level is far too low to affect snow albedo,' Stephen says.

As there have been no measurements of light-absorbing impurities in snow on Antarctic sea ice, Stephen and Maria have been collecting big bags of snow for analysis. The snow is melted back in the ship’s laboratory and then passed through a filter, which collects any traces of soot or dust. The filters will be analysed by Stephen back in Seattle, but already he has a good idea of what the result will be.

'We have some standard filters with known amounts of soot on them so we can get a visual estimate,' he says.

'On the Antarctic continent we have values of 0.1 to 0.7 parts per billion of soot and we’re seeing that range in the snow on ice floes here.

'I think there’s probably more soot fallout in the sea ice zone than on the Antarctic continent, but it’s being balanced by deeper snowfall. It’s still the cleanest snow in the world though.