We date sedimentary deposits showing that the southeastern Laurentide Ice Sheet was at or near its southernmost extent from ~ 26 000 to 21 000 years ago, when sea levels were at their lowest, with climate records indicating glacial conditions. Slow deglaciation began ~ 22 000 years ago, shown by a rise in modeled local summer temperatures, but significant deglaciation in the region did not begin until ~ 18 000 years ago, when atmospheric CO₂ began to rise, marking the end of the last ice age.

Cosmogenic nuclides, such as ¹⁰Be, are rare isotopes produced in rocks when exposed at Earth's surface and are valuable for understanding surface processes and landscape evolution. However, ¹⁰Be is usually measured in quartz minerals. Here we present advances in efficiently extracting and measuring ¹⁰Be in the pyroxene mineral. These measurements expand the use of ¹⁰Be as a dating tool for new rock types and provide opportunities to understand landscape processes in areas that lack quartz.

Cosmogenic nuclide exposure dating is an exceptional tool for reconstructing glacier histories, but reconstructions based on common target nuclides (e.g., ¹⁰Be) can be costly and time consuming to generate. Here, we present a low-cost proof-of-concept ²¹Ne exposure age chronology from Lassen Volcanic National Park, CA, USA that broadly agrees with nearby ¹⁰Be chronologies but at lower precision.

Determining where and when the Antarctic ice sheet was smaller than present requires recovery and exposure dating of subglacial bedrock. Here we use ice sheet model outputs and field data (geological and glaciological observations, bedrock samples and ground-penetrating radar from subglacial ridges) to assess the suitability for drilling of sites in the Hudson Mountains, West Antarctica. We find that no sites are perfect, but two are feasible, with the most suitable being Winkie Nunatak.

We constructed a geologic record of East Antarctic Ice Sheet thickness from deposits at Otway Massif to assess directly how Earth’s largest ice sheet responds to warmer-than-present climate. Our record confirms the long-term dominance of a cold polar climate but lacks a clear ice sheet response to the Mid Pliocene Warm Period, a common analogue for the future. Instead, an absence of moraines from the Late Miocene-Early Pliocene suggests the ice sheet was less extensive than present at that time.