Ice Tethered Profilers (ITPs) have been moored into ice in the Arctic Ocean since 2004 and lower instruments into the ocean every day to record water properties at different depths. The majority of profiles taken in the Canada Basin contain structures with many layers, each several meters thick and separated by thin interfaces. These are called double-diffusive staircases and have significant impacts on the vertical transport of nutrients and heat. However, they can be difficult to find in data. Here, for the first time, we identify staircase layers in data using the HDBSCAN clustering algorithm. Our approach is objective and requires significantly less prior knowledge than previous detection methods. Using this method, we reproduce the results of two previous studies. We track individual layers in staircases across time and space between different profiles. We quantitatively show that the difference in salinity between neighboring layers is greater than the salinity variance within a layer, while the opposite is true for pressure and temperature.
Detecting Arctic Ocean Staircases using a Clustering Algorithm
Host: Christian DiMaria