Last season at Davis we showcased the innovatively named ‘Big Blue Box’ that was used to collect ground penetrating radar data over the Sørsdal Glacier. As mentioned it can difficult to collect this kind of data over large areas especially on a fast flowing outlet glacier with a tendency to crevassing. Crevasses, cracks within the ice, occur when the ice cracks due to stress caused by movement. Stress can be caused by flowing over rocky and uneven ground beneath or to the side of the ice, by stretching ice due to speeding up or slowing down as a glacier flows and generally areas of the ice moving at different speeds. Crevasses can pose a hazard to anyone travelling on ice as they can become covered with fresh or windblown snow at the surface. Sometimes the snow is thick enough so that you can’t see the crevasse but not thick enough to take your weight. There are many techniques and safety practices used to detect and avoid crevasses and I’ve done many crevasses rescue training courses before travelling on glaciers for both work and leisure.
The ground penetrating radar data that we got from the Sørsdal Glacier last year provided us with such an amazing view of the drainage channels within the ice, we were excited to develop this for other applications. One such application I’m very passionate about is the detection of crevasses and this is done very successfully using ground penetrating radar. Ground penetrating radar is a well established technique for research on glacier, the earliest measurements were carried out in 1929 to estimate the thickness of a glacier. It is increasingly used for crevasse detection, often operated at the front of an over ice convoy or remotely using an autonomous rover. By developing the helicopter-borne radar system we can detect crevasses beneath the snow surface from the air. Our aim is to test and develop the system we’ve been using so that it can be easily mounted to the helicopter, not only for specific surveying but also for routine operation. The raw data that is collected is visible on the screen of the controller in realtime and unlike some techniques structures such as crevasses are clearly visible in this raw data.