Departmental Seminar

Marine-terminating glacier retreat is a key indicator of how ice sheets respond to climate forcing. Although atmospheric and ocean warming in the late 1990s has driven much of the observed retreat, individual glacier responses vary significantly. Even neighboring glaciers in similar fjord environments can behave quite differently, and there is currently no universally accepted law describing terminus change for use in ice sheet models. This highlights a fundamental gap in understanding which physical processes are being misrepresented or overlooked. In this talk, I demonstrate that long-term glacier terminus retreat can be predicted using a data-driven model trained on observations of climate, geometry, and ice dynamics—without relying on prescribed melt or calving laws. While geometry plays a dominant role in controlling retreat, its influence varies widely among glaciers, indicating that retreat is not a uniform process. To further investigate the role of geometry, I present findings from a recent survey of a Greenland outlet glacier that quantify topographic changes driven by erosion and sedimentation. Early results show that sedimentation occurs across a dynamic network of subglacial discharge outlets that evolve over time. This evolving structure alters near-terminus circulation and melt patterns, challenging common assumptions used in melt parameterizations. Overall, these findings suggest a critical limitation in current models: the assumption of uniform glacier response to climate, when in reality, these responses are spatially complex and structured.