Snow studies examine causes of 2013 flood


Treading carefully on late season snowpack, which inevitably gives way to sporadic knee-deep post-holing, Nicolas Leroux, a PhD candidate from France, and his supervisor, Dr. John Pomeroy, director of the University of Saskatchewan’s Centre for Hydrology, approached a pit excavated from the bottom 25 centimetres of the snowpack.

Stooping to his knees, Leroux reached into the space between the forest floor and a wooden platform sitting under 40 centimetres of snow. Pulling out a plastic Nalgene bottle, he announced, “There’s water in it.”

“Yay,” Pomeroy exclaimed, “the experiment works.”

The two scientists had reason to be pleased, as this was their first attempt to collect rain water that had flowed through the snowpack, known as rain-on-snow. In fact, it is the first such study they are aware of anyone conducting, anywhere.

With rain in the forecast, a few days prior the two researchers had set up a snowmelt lysimeter at one of the several research sites at the Fortress Mountain Snow Lab in Kananaskis. The lysimeter consists of a large tray of 16 compartments, each fitted with a spigot attached to a one-litre bottle to catch rainwater that flowed down through the snowpack.

As Leroux retrieved the bottles one by one, he measured the contents as Pomeroy recorded results. A few bottles were empty, a few were full to the brim, and the remainder contained varying volumes. Those results proved the model Leroux has developed that presumes rain does not flow through the snowpack following a uniform pattern, but rather by following preferential paths, or “flow fingers,” to be valid.

“I’m ecstatic,” Pomeroy said. “I was optimistic we’d get something, and those differences (in quantities) are really good. It proves the basic concept of flow fingers is valid. It confirms the hypothesis that snowpacks can transmit water rapidly to their base during rain-on-snow events, and that they do it through flow fingers. And (this experiment) hasn’t been done before with rain-on-snow.”

It was the massive floods that deluged Canmore, the Bow Valley and Kananaskis Country in June 2013 that inspired the study. High above Canmore, the alpine slopes above Cougar Creek held a lot of snow when the heavy rainstorm stalled over the front range peaks on June 20.

At that time, nobody knew how quickly rainwater could pass through a late-season snowpack, and scientists wondered if the snowpack – particularly a soft, unconsolidated or isothermic springtime snowpack – would be able to absorb rainwater, or if it would be able to release the water quickly. This new rain-on-snow study will help explain how that happens.

“This can be very valuable for flood forecasting,” Pomeroy said. “All the evidence from 2013 is that it was released quickly.”

A paper Pomeroy co-authored with Xing Fang, one of the U of S Centre for Hydrology team members, and Danny Marks of the USDA Agricultural Research Service’s Northwest Watershed Research Center in Boise, Idaho, that explains the characteristics and diagnosis of the Canadian Rockies’ June 2013 cold rain-on-snow event was recently accepted by the international journal Hydrological Processes.

Earlier this winter when temperatures were colder, Leroux conducted experiments by placing a metal plate on top of the snowpack, then heating the plate to send melted snow downward. By adding dark coloured food dye, the researchers were able to see the paths taken by the water as it collided with icy layers that sent it flowing horizontally until it reached weaknesses that allowed it to flow downward again toward the ground.

“The model shows that water doesn’t flow evenly through the snowpack,” Pomeroy said. “These preferential paths can deliver water very rapidly to the bottom of the snowpack. Knowing how fast the water can move tells us how quickly a flood can be generated. Rain-on-snow was a component of that (2013 flood) in the alpine.”

On the infrequent occasions that rain falls on a cohesive mid-winter snowpack, meltwater moving through the snowpack can easily re-freeze and be retained in the snowpack. While previous theories held that the spring snowpack could also soak up rain like a sponge, the reality of the massive flooding that resulted in 2013 proved quite different.

“That’s not how it really happens,” Pomeroy said. “It’s a sponge full of big holes.”

The theory of flow fingers was initially proposed in the 1980s in Resolute, Nunavut by Dr. Phil Marsh and Dr. Hok Woo of McMaster University. Pomeroy later worked with Marsh as part of an Environment Canada research team that investigated the chemistry of the snowmelt north of Inuvik, NWT.

Using water coloured with rhodamine dye, or sometimes even tea, they were able to follow the erratic path water took as it flowed through the snowpack. They also learned that snowmelt water that travelled through the snowpack to the ground via the preferential paths was composed of a brine consisting of sulphuric acid, hydrochloric acid, nitric acid and ammonia, which flowed into streams and lakes in spring, causing acid shock.

The action of flow fingers is of ecological importance as well, Pomeroy added, as those paths perform a similar function to how algae such as “watermelon snow” (a species of algae that thrives in freezing water and turns pink), occurs when water seeps down through the snowpack.

When meltwater reaches the ground it activates the algae. Lured also by sunlight, the organisms grow little tails and swim up through the snowpack, following flow fingers until they reach the surface.

Leroux’s highly advanced background in complex numerical modelling and understanding of soil physics – he is a graduate of the prestigious École nationale supérieure d’électronique, d’électrotechnique, d’informatique, d’hydraulique et des télécommunications in Toulouse, France – is instrumental to the modelling behind the experiment conducted at Fortress Mountain, part of the largest and most sophisticated snow study site in Canada – the Canadian Rockies Hydrological Observatory.

“It’s the first time we’ve done it, and I think it takes many tries to get some good information,” Leroux said. “I’m impressed we got some water in some of the bottles. I think we could develop it a bit more for next winter.”

Leroux’s plans for next winter include installing the lysimeter earlier in winter, and inserting a flow meter into each spigot to allow immediate recording of the amount of water flowing by having the information relayed to a data logger.

After all the bottles were emptied at the site, Leroux dug a small snow pit to gain additional information. Measurements showed the snowpack was not uniformly wet; overall, it held higher water content than the day they set it up before the rain, with higher concentration in the middle third and the driest snow at the bottom.

“Not many people are studying wet snow metamorphosis,” Leroux said. “Wet snow metamorphosis involves the changing of grain size with liquid water content. In snow, there are a lot of physical processes that are combined. Snow grain metamorphosis is one of the bigger problems to solve. It has a big impact on water and heat flows.

“This is the first time the test has been done using actual rain on snow. For the plate experiment I had to do it six times to get it right. It took a month and a half. It’s difficult to combine modelling and fieldwork. But then, you learn a lot digging pits.”


About Author

Rocky Mountain Outlook