In February I travelled to the University of Manitoba’s Sea Ice Environmental Research Facility (SERF) with CPOM post-doc Rosemary Willatt to help with an experiment. The centerpiece of SERF is a eighteen meter long, concrete tank set into the ground. Looking a bit like an outdoor swimming pool, it’s filled with saltwater and surrounded by high-tech instruments. Winter temperatures at SERF regularly drop to -25°C, creating a covering of ‘artificial’ sea ice hundreds of miles from the sea itself.
Sunset at SERF. Two radar instruments overlook the tank and take automated measurements every eight minutes.
The view from the SERF office of a high-tech radar and a low-tech human. Photo: Rosemary Willatt
Ice that forms from seawater is different to the stuff that forms on lakes. When seawater turns to ice, the salt that was previously dissolved is mostly rejected from the structure in the form of super-salty brine. Some remains in millimetre-sized pockets within the ice structure, but most streams out the bottom or sits on top. This brine fundamentally changes the physical and electromagnetic properties of the ice. Scientists wanting to study the properties of sea ice must therefore grow it from saltwater, which requires very low, sustained temperatures.
This winter’s main SERF experiment focused on the snowpack that accumulates on top of sea ice. When it snows (and it always eventually snows), the cold snow soaks up some of the upward-rejected brine like a sponge. The height that the brine reaches in the snow is poorly understood, posing a problem for scientists interested in measuring sea ice thickness from space.
Scientists currently estimate sea ice thickness by using radar waves that are often assumed to penetrate the snow and bounce off the underlying ice surface. However, it’s also increasingly accepted that these radar waves won’t penetrate through brine-soaked snow, so if brine is present in the snow then our historical assumptions won’t hold. Where radar waves bounce off the briney-snow layer before making it to the ice, we measure the sea ice to be thicker than it really is. Understanding the height that brine reaches in the snow would be a big step towards improving our satellite-derived estimates of sea ice thickness.
Snow falls in large, highly complex shapes. The grains we saw sometimes exceeded 5mm in diameter, making them extremely difficult to model.
Sampling the surface of the ice with a chisel. Photo: Rosemary Willatt, CPOM UCL.
At SERF Rosie and I sampled and observed the snow cover and ice surface twice a day when it wasn’t snowing. When it snowed three days into the experiment, we sampled the snow every hour. Every two days we would drill an ice-core to measure the ice thickness and sample the full extent of the ice. By the end of the ten day experiment we’d collected more than a hundred samples, which we then took to the lab for salinity analysis.
While Rosie and I focused on our physical sampling program, Dustin Isleifson and Chris Fuller from University of Manitoba monitored the ice using radars similar to those carried on satellites. A third team made up of David Landry and Maddie Harasyn scanned the snow surface with a drone-mounted lidar. Our task is now to establish how the directly measured changes in the snow translate to what Dustin and Chris recorded with the radars and what David and Maddie scanned with the drone. Hopefully we can use the data to improve our understanding of what satellites like CryoSat-2 or Sentinel-3 are seeing from space when they’re pointed at the sea ice.
Rosie drills an ice core near the edge of the tank.
As well helping run the experiment, Rosie and I used our time in Canada to discuss our science and build collaboration with colleagues at the Centre for Earth Observation Science at UoM. The whole department was incredibly welcoming and friendly to us, with the scientists keen to show us around Winnipeg. By all accounts Rosie gave an excellent talk on her research, which sadly I had to sleep through after being up all night sampling the snow. As well as making Winnipeg feel like home for two weeks, Chris Fuller taught Rosie and me a lot about snow and ice sampling techniques and was always on hand to answer our questions. We’re now back in comparatively warm London and ready for some data analysis!
Chris Fuller downloads data from the C-band radar instrument.