Modeling Thwaites Glacier shear margin stability by slip initiation due to variation of effective pressure

The loss of Thwaites Glacier is the most important factor in determining the future stability of the West Antarctic Ice Sheet (WAIS), which contains enough ice to raise global mean sea level by up to 3.4 m. All ice that will terminally calve from a glacier can be shows as a flux, that is a velocity through a cross-sectional area. This flux is partly controlled by the distance between the two lateral shear margins that bound an ice stream. For glaciers unconstrained by topography, margin migration can significantly influence ice discharge to the ocean. This is well illustrated in the case of the eastern shear margin (ESM) of Thwaites.

The initiation of basal slip is determined according to a sliding law. Although regularized Coulomb-style sliding laws can resolve slip over both hard and soft bedded glaciers, they depend on knowledge of the effective pressure (ice overburden minus water pressure) of the ice. Most modeling studies that examine Coulomb style slip laws limit themselves to a constant floatation percentage, a constant melt rate based on thickness, or other hard parametrizations, while inverting for friction coefficient values. Instead, I look to find the effect of the variation of basal hydrology with a constant friction coefficient, and quantify the change in slip initiation under differing upstream water flux scenarios.

To examine hydrological shear margin controls, I implemented two models. First, a regional hydrology model which couples the Glacier Drainage system model (GlaDS) with a shallow shelf flow approximation (SSA) forced by an ITS_LIVE annual velocity mosaic. This model yields a high resolution, near-steady state, regional-trending hydrologic system across the Amundsen Sea Sector of West Antarctica which informs boundary conditions for a finer scale model. The second model is local to the Thwaites ESM. It is a hydro-thermomechanical 3D flow model which melt in the basal elements of the domain from enthalpy. In the local model, I couple this melt and hydrology to slip through a regularized Coulomb-style sliding law to calculate the spatial slip distribution over the shear margin.

This work finds good agreement between flow solutions and current velocity observations at the local model scale. Additional upstream water sources divert excess water to the Thwaites catchment rather than to neighboring Pine Island Glacier, and these increased fluxes drive a widening of Thwaites’ main trunk (a margin drift outwards). Future work should expand this local model to regional and WAIS-wide domains to resolve many ice stream dynamics in prognostic ice flow models to more accurately predict sea level rise contributions.