Multi-scale monitoring of sea ice parameters with passive seismology and deep learning – MSIM

In the context of the changing climate, sea ice extent and concentration are reducing at an accelerating rate. To integrate these changes, the next generation of sea ice and climate models will require a fine characterization, with an accuracy and resolution that cannot be obtained with current satellite-borne imaging methods. For example, not only the thickness of the ice, but also its stiffness and its density, are essential parameters for a better understanding and modelling of sea ice in general, and in particular in the marginal ice zone, which determines the flow size distribution.
The goal of this project is to provide measurements of these parameters with unprecedented accuracy, at scales ranging from tens of meters to a few kilometers, for deeper and more comprehensive analyses of pack ice dynamics. To this end, this collaborative research will focus on the following two main axes:
1) the development of new monitoring methods based on passive seismology, distributed acoustic sensing and deep learning, to generate high spatial and high temporal resolution maps of sea ice parameters (thickness, stiffness, and density), as well as those of the snow cover. These methods will require light instrumentation that is versatile and easy to deploy in the field with a drone, which will significantly reduce logistics, compared to other field methods such as AUV, drilling, or electromagnetic sounding, that are restricted to local measurements of the ice thickness.
2) the development of field and laboratory experiments with cutting edge measurement methods, in view of obtaining new upscaling laws for the physical processes at play during wave-induced sea ice breakup.