Variational inference of ice shelf rheology with physics-informed machine learning

Abstract

Floating ice shelves that fringe the coast of Antarctica resist the flow of grounded ice into the ocean. One of the key factors governing the amount of flow-resistance provided by an ice shelf is the rigidity of the ice that constitutes it. Ice rigidity is highly heterogeneous and must be calibrated from spatially-continuous surface observations assimilated into an ice flow model. Moreover, realistic uncertainties in calibrated rigidity values are needed to quantify uncertainties in forecasts of future shelf flow. Here, we present a physics-informed machine learning framework for inferring the full probability distribution of rigidity values for a given ice shelf, conditioned on surface velocity and thickness fields derived from remote sensing data. We employ variational inference to jointly train neural networks and a variational Gaussian Process to reconstruct surface observations and rigidity values and uncertainties. Application of the framework to synthetic and large ice shelves in Antarctica demonstrate that rigidity is well-constrained in areas where deformation of ice is measurable within the noise level of the observations. Further reduction in uncertainties can be achieved by complementing variational inference with conventional inversion methods. Our results demonstrate a path forward for continuous calibration of ice flow parameters from remote sensing observations.

Weak relationship between remotely detected crevasses and inferred ice rheological parameters on Antarctic ice shelves

“…Current ice flow models do not account for an evolving ice rate factor in transient simulations, thus the accuracy of their projections remains restricted. Recent advances adopting a novel physic informed machine learning framework have tested a complementary approach to calibrate uncertainties in the ice rate factor and sea level forecasts, by inferring a posterior distribution of the ice rate factor (Riel and Minchew, 2023), providing some insights on better and continuously updated calibrations of ice flow parameters. To further reduce these errors in the ice rate factor field when adopting more conventional inversion procedures, it remains crucial for future efforts to prioritize the monitoring of ice shelves' velocity, geometry, and calving front positions, which are essential parameters for accurate estimates of sea level forecasts.…”

Widespread slowdown in thinning rates of West Antarctic ice shelves

“…These artifacts are amplified by conventional (explicit) spatial derivatives and map onto the melt-rate estimates. Opticalderived velocities lack continental coverage and contain artifacts in places (Riel and Minchew, 2023). For this reason, we derive velocity fields by blending multiple products.…”