Water flow under ice exerts a fundamental control on glacier and ice-sheet dynamics, with direct consequences for the pace of global sea-level rise and the frequency and severity of glaciological hazards. Despite its importance, progress in modeling the subglacial drainage system has been limited by challenges in upscaling the governing physics, the computational cost of resolving processes that operate across vastly different timescales, and to some extent, the persistent scarcity of direct observations. This seminar will sketch the development of physically based subglacial drainage models and describe how statistical emulators are being used to accelerate large-scale predictions of subglacial hydrological variables. We will reflect on the current state of the field and highlight the model deficiencies most in need of attention

Reading(s):

For those interested in a general introduction to physically-based subglacial drainage modelling:

Flowers, G.E. 2015. Modelling water flow under glaciers and ice sheets. Proceedings of the Royal Society A 471: 20140907, https://doi.org/10.1098/rspa.2014.0907

For those interested in subglacial drainage emulation (Gaussian Process model):

Hill, T., D. Bingham, G.E. Flowers, M.J Hoffman. 2025. Computationally efficient subglacial drainage modelling using Gaussian process emulators: GlaDS-GP v1.0, Geoscientific Model Development, 18, 4045–4074, https://doi.org/10.5194/gmd-18-4045-2025.

For those interested in an application to Antarctic drainage:

Hill, T., M.J. Hoffman, G.E. Flowers, D. Bingham. Submitted. Random Forest parameterization of Antarctic subglacial hydrology for coupled ice-flow modelling. The Cryosphere. Preprint: https://doi.org/10.5194/egusphere-2026-343