How Ice Age ice sheets and thickness are reconstructed

Evidence and modelling have different roles

Glacial landforms, sediments and dating evidence help constrain former ice extent. PaleoMIST uses reconstructed margins and a model of ice-sheet geometry to estimate grounded ice thickness. Assumptions about the bed and ice mechanics influence the result.

The white overlay uses published ICESHEET margin files for North America, Eurasia, Antarctica and Patagonia. It is not a comprehensive inventory of every small mountain glacier.

References: Gowan et al. (2021) — A new global ice sheet reconstruction for the past 80,000 yearsICESHEET 2.0 — published margin files and code

What happens when you click?

The atlas selects the closest 1° node from the published PaleoMIST global grounded-ice grid and the nearest 2,500-year snapshot. One degree spans about 111 km north–south; east–west distance shrinks toward the poles. The value is rounded to 10 m for display, which is a display choice, not a claim of 10 m accuracy.

Zero means no grounded ice in that model cell. It does not exclude a small glacier or floating ice. Missing data and failed downloads are shown separately. The nearest grid node can fall on the other side of a simplified ice edge.

References: Gowan (2019) — PaleoMIST reconstruction dataset, corrected April 2021 (CC BY 4.0)

Why this is one reconstruction, not a final answer

PaleoMIST is a peer-reviewed reconstruction. Its sea-level implications have also been challenged in a published commentary, followed by a reply from the authors. Both are linked here. Model publication does not make every estimated thickness a settled observation.

The margin/thickness model and the atlas’s Spratt–Lisiecki sea-level curve are separate inputs. Their combination is an educational visualization, not a single internally consistent ice–Earth–ocean simulation.

References: Gowan et al. (2021) — A new global ice sheet reconstruction for the past 80,000 yearsYokoyama et al. (2022) — Towards solving the missing ice problem and the importance of rigorous model data comparisonsGowan et al. (2022) — Reply to the missing ice commentary

Use thickness comparisons responsibly

Keep the same selected point while moving the timeline. Changes indicate differences between model snapshots. Do not infer a local melt rate from adjacent 100-year slider steps, or extrapolate this reconstruction into a future forecast.