A Greenland-wide empirical reconstruction of paleo ice sheet retreat informed by ice extent markers: PaleoGrIS version 1.0

Abstract

Abstract. The Greenland Ice Sheet is a large contributor to global sea level rise, and current mass losses are projected to accelerate. However, model projections of future ice sheet evolution are limited by the fact that the ice sheet is not in equilibrium with present-day climate but is still adjusting to past changes that occurred over thousands of years. While the influence of such committed adjustments on future ice sheet evolution remains unquantified, it could be addressed by calibrating numerical ice sheet models over larger timescales and, importantly, against empirical data on ice margin positions. To enable such paleo data–model interactions, we need Greenland-wide empirical reconstructions of past ice sheet extent that combine geomorphological and geochronological evidence. Despite an increasing number of field studies producing new chronologies, such a reconstruction is currently lacking in Greenland. Furthermore, a time slice reconstruction can help to (i) answer open questions regarding the rate and pattern of ice margin evolution in Greenland since the glacial maximum, (ii) develop a standardised record of empirical data, and (iii) identify new sites for future field campaigns. Based on these motivations, we here present PaleoGrIS 1.0, a new Greenland-wide isochrone reconstruction of ice sheet extent evolution through the Late Glacial and early- to mid-Holocene informed by both geomorphological and geochronological markers. Our isochrones have a temporal resolution of 500 years and span ∼ 7.5 kyr from approximately 14 to 6.5 kyr BP. We describe the resulting reconstruction of the shrinking ice sheet and conduct a series of ice-sheet-wide and regional analyses to quantify retreat rates, areal extent change, and their variability across space and time. During the Late Glacial and early- to mid-Holocene, we find the Greenland Ice Sheet has lost about one-third of its areal extent (0.89 million km2). Between ∼ 14 and ∼ 8.5 kyr BP, it experienced a near-constant rate of areal extent loss of 170 ± 27 km2 yr−1. We find that the ice-sheet-scale pattern of margin retreat is well correlated to atmospheric and oceanic temperature variations, which implies a high sensitivity of the ice sheet to deglacial warming. However, during the Holocene, we observe inertia in the ice sheet system that likely caused a centennial- to millennial-scale time lag in ice extent response. At the regional scale, we observe highly heterogeneous deglacial responses in ice extent evident in both the magnitude and rate of retreat. We hypothesise that non-climatic factors, such as the asymmetrical nature of continental shelves and onshore bed topographies, play important roles in determining the regional- to valley-scale dynamics. PaleoGrIS 1.0 is an open-access database designed to be used by both the empirical and numerical modelling communities. It should prove a useful basis for improved future versions of the reconstruction when new geomorphological and geochronological data become available.

Automatic identification of streamlined subglacial bedforms using machine learning: an open-source Python approach

“…Several efforts in the last few years have focused on using previous work to develop a single map of glacial features across Britain (BRITICE: Clark et al 2018), the Arctic (GlaciDat: Streuff et al 2022), and Greenland[]( Leger et al 2024) Even so, it is not yet always customary to make developed mapping tools and geomorphic datasets from deglaciated landscapes fully open and publicly available. There is an importance for geoscientific data, including geomorphic analyses, to meet FAIR data standards (Wilkinson et al 2016) as recommended by NASA's Transform to Open Science Initiative (Gentemann 2023) and emphasized at NCAR's recent "Community Expectations for a Geoscience Data Commons" workshop, which had wide international representation.…”

“…When tools for open and FAIR geomorphic datasets are paired with data from ICE-D, we will have a powerful tool for establishing ice histories across large spatial scales (e.g., Leger et al 2024, Stoker et al 2024.…”