Learning Subglacial Bed Topography from Sparse Radar with Physics-Guided Residuals

TL;DR

This paper introduces a physics-guided residual learning framework that improves subglacial bed topography predictions from sparse radar data, outperforming existing methods and ensuring physically plausible, high-fidelity results for ice sheet modeling.

Contribution

The authors develop a novel residual learning approach that integrates physics-based constraints with deep neural networks to enhance bed topography reconstruction from limited radar observations.

Findings

Achieves high accuracy and structural fidelity in Greenland sub-regions.

Outperforms U-Net, Attention U-Net, FPN, and plain CNN in tests.

Produces spatially coherent, physically plausible bed maps.

Abstract

Accurate subglacial bed topography is essential for ice sheet modeling, yet radar observations are sparse and uneven. We propose a physics-guided residual learning framework that predicts bed thickness residuals over a BedMachine prior and reconstructs bed from the observed surface. A DeepLabV3+ decoder over a standard encoder (e.g.,ResNet-50) is trained with lightweight physics and data terms: multi-scale mass conservation, flow-aligned total variation, Laplacian damping, non-negativity of thickness, a ramped prior-consistency term, and a masked Huber fit to radar picks modulated by a confidence map. To measure real-world generalization, we adopt leakage-safe blockwise hold-outs (vertical/horizontal) with safety buffers and report metrics only on held-out cores. Across two Greenland sub-regions, our approach achieves strong test-core accuracy and high structural fidelity, outperforming U-Net, Attention U-Net, FPN, and a plain CNN. The residual-over-prior design, combined with physics, yields spatially coherent, physically plausible beds suitable for operational mapping under domain shift.