As the days wax longer, few of us care to dwell on the snow and ice that dominated the landscape only weeks ago. But for the scientists trying to explain and predict the movement of chemicals in our environment — from contaminants like polychlorinated biphenyls (PCBs) to atmospheric gases like ozone — snow and ice are a bit of an obsession, important factors in the cycle that can’t be ignored. As they freeze, melt and re-freeze, snowpacks and sea ice provide rapidly-evolving interfaces between the solid, liquid and gas phases, and have a powerful impact on chemical transport both locally and globally. So it makes sense that researchers in Toronto and Winnipeg have spent millions of dollars constructing ‘cold labs’ which allow them to probe the complex chemistry of snow and ice in carefully controlled conditions; their findings are key to managing the chemicals that inevitably make their way into the environment.

“If you go into a northern landscape in the winter, snow is really all you see, yet until about ten years ago, there had been very little work done,” says Frank Wania, a professor in the department of physical and environmental sciences at the University of Toronto Scarborough. An expert in computer modelling, Wania was one of the first environmental scientists to propose the idea of cold-trapping, whereby semi-volatile organics in the atmosphere condense once they reach the polar regions. Without enough heat to release them back into the air, they become trapped and start to accumulate. This explains why substances like the insecticide DDT are often found at surprisingly high concentrations in the Arctic, thousands of kilometers from where they were actually used.

But it’s not just cold temperatures that concentrate organic pollutants: falling snowflakes can scavenge such chemicals from the air. “They’re actually quite efficient at this,” says Wania. “A snowflake has tremendous surface area, so chemicals that are condensing out of the gas phase will adsorb very effectively to the snow surface.” If the snow remains on the ground permanently, it can provide a record of contaminant deposition going back decades. Torsten Meyer, who recently completed his PhD in Wania’s lab, knows this effect very well. From 2006 to 2008, Meyer was part of a team that dug seven-metre deep pits on Devon Island in Nunavut, collecting snow samples that dated back to 1993. Analysis by gas chromatography showed concentrations of the flame retardant decabromodiphenyl ether (decaBDE) as high as 100 nanograms per litre. While still at relatively low levels, it’s a demonstration of just how effective snow can be at concentrating organic pollutants in remote areas.

In most of Canada, however, snow doesn’t last more than a few months, and undergoes important changes in its structure during that time. Tracking the fate of chemicals in aging snow had been an important part of Wania’s computer models, but in science, nothing beats a controlled experiment. “I thought it would be really cool to have a dedicated facility with all sorts of controls and programmable temperature changes,” says Wania. About ten years ago, Wania received a grant from the Canada Foundation for Innovation (CFI) to build a two-chambered, state-of-the-art walk-in freezer for his lab at the University of Toronto Scarborough. Once it was finished, he hired Meyer to figure out how to fill it with snow.

On ski slopes, artificial snow is made by forcing compressed air and compressed water through a tiny nozzle, producing a fine mist that, if the temperature is low enough, will freeze into tiny ice crystals. Meyer built his home-made snow gun out of a length of tubing attached to a water tap and an air compressor from Home Depot. “The nozzle size was very important: if it’s too large, the water won’t freeze, but if it’s too small, you don’t generate snow quickly enough,” says Meyer. Another problem was that since freezing is an exothermic process, the heat generated would quickly overload the cooling system; Meyer had to alternate from chamber to chamber in order to give each half of the facility enough time to cool down. Finally, Meyer designed an injection system that would squirt the organic compound of interest into the freezing water jet, ensuring a predictable contaminant concentration. “All of those things had to be figured out by trial and error,” he says, with a hint of frustration. “But it was very satisfying once I got it to work.”

Using the facility, Wania and Meyer obtained detailed information about how the aging of urban snow affects the pollutants contained within. “Even at minus 20 degrees, the snow can metamorphose,” says Meyer. “In general, the grains grow and become more uniform, and the surface area per unit volume decreases.” That shrinking of the snow surface leaves less room for any chemicals that are adsorbed onto it. While the more volatile ones can be released back into the atmosphere, the non-volatile ones have only one option: to clump together with motes of dust and dirt, forming particles that visibly darken the snow.

Once the snow starts to melt, the process accelerates. “As soon as there is meltwater in the snowpack, the grains grow much faster; you can literally see the pure ice and snow grains increase in size overnight,” says Meyer. The meltwater can also dissolve any chemicals that are hydrophilic. This means that the first trickle of spring runoff can carry a whole winter’s worth of water-soluble chemicals — for example, the pesticide chlorothalonil — into streams, soil, or groundwater, depending on the permeability of the underlying ground. The dirt particles with their adsorbed hydrophobic chemicals — for example, polycyclic aromatic hydrocarbons (PAHs) from car exhaust — can’t dissolve in the meltwater, so they grow larger, forming the characteristic black stain often seen in urban snow. These contaminants will also be released all at once, but only when the last of the snow melts.

One of the key results of Meyer and Wania’s experiments has been a series of chemical space plots. These are visual representations of where a given substance is likely to end up in the snowpack based on its chemical properties: dissolved in the meltwater, adsorbed to the ice surface or incorporated into particles of organic matter. “We tried to come up with answers that are generally applicable, so that if you have a new chemical that you don’t know anything about in terms of snow behaviour, you could try to estimate it via these partition coefficients,” says Wania. That in turn could help with environmental management: for example, deciding where to dump the snow removed from city streets so as to minimize the impact of contaminants released when it melts.

* * *

For most of us, snow characterizes the Canadian winter, but in the Arctic, the dominant features year-round are water and sea ice. One of Canada’s experts in this area is Feiyue Wang, an environmental chemist and biogeochemist at the University of Manitoba. Over the last decade, his team has logged countless hours aboard the Canadian research icebreaker Amundsen, making measurements of sea ice and correlating them with data obtained from satellites. Whereas in the 1970s and 1980s the Arctic was covered with sea ice that lasted all year round, today ever-larger areas melt and re-freeze each year. Among other things, Wang is trying to track the interaction between chemicals in the lower atmosphere — for example, bromine oxide, ozone, and mercury vapour — with those in newly forming sea ice. “It’s good to work in the field, but there are many different processes that could be responsible for what you see,” says Wang. “As a chemist, I want to be able to hold certain variables constant.” It was this drive that led him, along with three other principal investigators, to create the Sea-ice Environmental Research Facility (SERF) which just completed its second research season.

SERF consists primarily of a 400 cubic-metre concrete pond filled with artificial seawater. Despite the $1.6 million budget (also from CFI) the team had to be thrifty with their raw materials. “We needed more than 10 tonnes of salt, which could have been quite expensive,” he says. “Luckily, the City of Winnipeg donated some of the sodium chloride brine they use for defrosting roads, which we then topped up with secondary salts. The chemical composition is as close to seawater as it can be.” The facility has a heating system and a retractable roof, which ensure that each experiment starts from the same carefully controlled conditions. However, it’s up to the famous Winnipeg winter to provide the cold temperatures the system needs to run, meaning it can only operate for about three months of the year. Learning from their experience on the Amundsen, Wang and his colleagues have designed their experiments such that up to 30 workers from different research groups can be taking measurements at once. At the end, they combine their data to give a more complete picture of sea ice formation, growth and melting.

Surprisingly, some of the most basic features of sea ice are still relatively unstudied. For example, Wang’s group recently published a paper in Marine Chemistry that contains the first pH measurements of newly forming sea ice. “We found that the exterior part of the ice — the surface and the bottom layer — is highly alkaline, whereas the interior is more acidic,” says Wang. That finding may not seem important, but in fact it has powerful implications for the overall flow of carbon in the Arctic Ocean. Most climate models treat the Arctic Ocean as a net sink for CO2, which lowers ocean pH by forming carbonates, bicarbonates and carbonic acid. However, as ice crystals form, they exclude these dissolved ions, pushing them to the outside of the ice pack in a process known as freeze rejection. This creates a concentrated brine on the surface, which could act as a source of CO2, rather than a sink. If carbon is indeed flowing from this brine to the atmosphere, one of the results would be increased alkalinity, exactly what the SERF researchers observed.

Another phenomenon that impacts the ice-atmosphere interface is frost flowers. These beautiful structures are formed not from sea water, but from atmospheric moisture that deposits on a newly-forming sea ice surface, similar to the hoarfrost often seen on windows and windshields. However, due to their porous structure, they can act as wicks to expose even more brine to the atmosphere, further increasing the exchange of salts. Part of Wang’s work focuses on the role of frost flowers in so-called ‘bromine explosions,’ an atmospheric chemistry phenomenon that often follows large Arctic storms in the springtime. Bromine oxide, presumably from seawater, initiates a complex cascade of reactions that sharply reduces levels of atmospheric mercury. The hunch is that this might be one of the mechanisms that brings mercury from the atmosphere into the water column, where it eventually finds its way into Arctic mammals like seals and whales.

Both SERF and Wania’s snowmelt lab have already done important work, but the researchers aren’t content to rest on their laurels. Of the emerging issues that warrant investigation, oil development looms large. “To me it’s really just a matter of time before we will see oil drills in the Arctic,” says Wang. “And let’s face it, there will be spills.” Wang dreams of an expansion for his facility that could simulate an Arctic oil spill and provide a testing ground for methods to clean it up. He’s already got a design, and WANGis now looking for funding to make it a reality. For his part, Wania is interested in another symptom of oil development: atmospheric pollutants like PAHs. “I have a student who is modelling the fate of PAHs in an area such as the oil sands region, which is snow-covered for half of the year,” he says. “It’s an entirely theoretical study, but of course these models rely on the sorts of things we learned from our laboratory work.”

Untangling the mysteries of environmental chemistry is tough work; it requires a combination of laboratory studies from these unique facilities as well as improved computer models and careful fieldwork like that conducted on the Amundsen. But when it comes to snow and ice, there’s little doubt that Canadians are leading the pack.