Speaker: Anne Voigtlaender, Earth & Environmental Sciences Area (EESA), Lawrence Berkeley National Laboratory (LBL), Postdoctoral Researcher in Basic Energy Sciences (BES) – Geosciences program with Professor Yida Zhang

Seminar:  Why firn quakes - a continuum mechanics theory with granular legacy

Abstract
Compacting snow becomes firn then ice. While the snowpack consolidates it transitions from a non-homogeneous granular material to a more elastic continuum material. The granular legacy produces spatial variations in density, stiffness, and pre-stress. This creates an internal structure of supports in unconsolidated snow at depth. Firn can quake, when these supports collapse. By combining granular with brittle fracture mechanics, and making use of statistical percolation theory, we can explain the conditioning, triggering, and progression of firnquakes in a bulk homogeneous material, with near constant boundary conditions. Based on minimum assumptions and data, we address a general phenomenon in compacting granular medium, in which, unlike firnquakes, ruptures can have hazardous consequences, like landslides, avalanches, powder tailing failure.
 

Combining granular and continuum mechanics provides us with additional conceptual pathways to explore firnquakes. The granular legacy allows us to hypothesize that structures of denser layer resting on solid like supports in the unconsolidated layer exist at depth. As these supports carry the overburden, they are pre-stressed. Dynamic triggering, we show would be feasible to cause local collapse of the supports and can propagate by a sort of domino effect. Based on bulk density and elastic properties we find the propagation speed of flexural waves by collapsing supports matches the firnquake velocities in the order of 100 m/s. The analytical solution also provides us with a length scale of maximum spacing between pre-stressed supports is on the order of decimeters, to trigger the next one. While we present a consistent theory to explain why firn quakes, more observational data and dedicated experiments are needed to test our theory and hypotheses.