We review successes and challenges from five recent subglacial bedrock drilling campaigns intended to find evidence for Antarctic Ice Sheet retreat during warm periods in the geologic past. Insights into times when the polar ice sheets were smaller than present serve as guiding information for modeling efforts that aim to predict the rate and magnitude of future sea level rise that would accompany major retreat of the Antarctic Ice Sheet. One method to provide direct evidence for the timing of deglaciations and minimum extent of prior ice sheets is to extract subglacial bedrock cores for cosmogenic nuclide analysis from beneath the modern ice sheet surface. Here we summarize the lessons learned from five field seasons tasked with obtaining bedrock cores from shallow depths (<120 m beneath ice surface) across West Antarctica since 2016. We focus our findings on drilling efforts and technology and geophysical surveys with ground-penetrating radar. Shallow subglacial drilling provides a high risk, high reward means to test for past instabilities of the Antarctic Ice Sheet, and we highlight key challenges and solutions to increase the likelihood of success for future subglacial drilling efforts in polar regions.

In-situ cosmogenic 10Be (and 26Al) concentrations in alluvial sediments provide a spatially averaged signal of the erosion rate of the catchment area. Catchmentwide erosion rates reflect the production rate of the entire basin, and their calculation requires knowledge of the complete production rate model. Available calculators determine production rates on a pixel-based approach and achieve computational efficiency by relying on a scaling method that ignores geomagnetic field strength variations. Here we introduce a new python-based tool that determines erosion rates based on the hypsometry of the catchment. The method relies on the fact that production rates are much more sensitive to changes in elevation than latitude. Our tool has two main advantages: (1) computation time is short (<30 seconds) and independent of the scaling method; there is no need to neglect magnetic field variations, and (2) because production rate scaling is performed by a widely used online calculator, the results are fully comparable to exposure ages or point-based erosion rates determined with the same calculator; future updates to production rate scaling are immediately effective for catchmentwide erosion rate calculation. We demonstrate in two case studies that (1) for similar scaling methods, our calculator reproduces pixel-based results within a few percent, and (2) erosion rates determined with different scaling methods may differ by >20%, differences can vary systematically with erosion rate, and using a time-constant scaling method may result in a bias in the interpretation of catchmentwide erosion rates.

We used measurements of radar-detected stratigraphy, surface ice-flow velocities and accumulation rates to investigate relationships between local valley-glacier and regional ice-sheet dynamics in and around the Schmidt Hills, Pensacola Mountains, Antarctica. Ground-penetrating radar profiles were collected perpendicular to the long axis of the Schmidt Hills and the margin of Foundation Ice Stream (FIS). Within the valley confines, the glacier consists of blue ice, and profiles show internal stratigraphy dipping steeply toward the nunataks and truncated at the present-day ablation surface. Below the valley confines, the blue ice is overlain by firn. Data show that upward-progressing overlap of actively accumulating firn onto valley-glacier ice is slightly less than ice flow out of the valleys over the past ∼1200 years. The apparent slightly negative mass balance (-0.25 cm a-1) suggests that ice-margin elevations in the Schmidt Hills may have lowered over this time period, even without a change in the surface elevation of FIS. Results suggest that (1) mass-balance gradients between local valley glaciers and regional ice sheets should be considered when using local information to estimate regional ice surface elevation changes; and (2) interpretation of shallow ice structures imaged with radar can provide information about local ice elevation changes and stability.

Riverbluff Cave, in Greene County, Missouri, is a short single passage between the James River and its direct tributary, Ward Branch. Before stream incision the cave functioned as a spillway/piracy between the two streams during high-discharge events and accumulated a sequence of stratified fluvial sediments within the cave. Five cosmogenic-nuclide burial ages for these sediments range from 0.984 to 0.570 Ma. These ages are consistent with both the stratigraphic order of the samples and the inferred position of the Matuyama/Brunhes paleomagnetic boundary. These ages indicate that sandy channel-facies deposits derived from Ward Branch entrances began to accumulate within the cave as early as 0.984±0.065 Ma. This facies is capped by highly fossiliferous gravel beds dated at 0.658±0.065 Ma, which contain abundant mammoth bones (possibly Mammuthus trogontherii) and other vertebrates. The high concentration implies that this deposit may record some type of mass-mortality event. By 0.570±0.072 Ma, all Ward Branch entrances had been abandoned because of incision, and a laminated red clay derived from backflow from flooding along the James River capped the older channel sediments.

Estimates of regolith degradation in the McMurdo Dry Valleys of Antarctica are currently based on indirect evidence and ancient ashes at or near the soil surface that suggest excellent preservation of surfaces. On the other hand, the existing cosmogenic-nuclide surface exposure ages from many parts of the Dry Valleys are younger than the age of surface deposits inferred from stratigraphic relations. This suggests some combination of surface erosion or past ice cover, both of which would reduce the apparent exposure age. This paper quantifies the regolith degradation and/or past ice cover by measuring10Be and26Al from a landslide deposit that contains 11.3 Ma volcanic ash. The surface sample yields an apparent exposure age of only 0.4 Ma. However, measurements of the subsurface nuclide concentrations show that the deposit has not been shielded by ice, and that the age of the ash does not conflict with the apparent exposure age when slow degradation of the deposit (2 m Ma−1) is taken into account. Soil creep, which is a common degradational process in a wide variety of environments, is non-existent at this field site, which likely reflects the persistent lack of bio- and cryoturbation.

The Whippoorwill Formation is a gleyed diamicton that is present locally within bedrock depressions beneath the oldest glacial till in northern Missouri, USA. Stratigraphy, paleomagnetism, and cosmogenic–nuclide burial ages show that it was deposited between the Matuyama–Gauss magnetostratigraphic boundary at 2.58 Ma and the first advance of the Laurentide ice sheet into Missouri at 2.47 ± 0.19 Ma. High cosmogenic–nuclide concentrations also show that the constituents of the Whippoorwill Formation experienced long exposure at a stable landscape surface with erosion rates of 1–2 m/Ma. However, cosmogenic–nuclide concentrations are invariant with depth below the Whippoorwill Formation surface, indicating active mixing of the soil profile shortly before burial by till. The Whippoorwill Formation retains numerous features indicative of cryoturbation. Therefore, we interpret it as a buried Gelisol, a soil formed under periglacial conditions in the presence of permafrost. At the onset of Northern Hemisphere glaciation, climate cooling established permafrost conditions and accelerated erosion by inducing landscape instability. Thus, weathered regolith materials were mobilized and redeposited by gelifluction shortly before the ice sheet overrode the landscape.

Reconstructions of glacioisostatic rebound based on relative sea level in Maine and adjacent Canada do not agree well with existing geophysical models. In order to understand these discrepancies better, we investigated the lake-level history of 40-km-long Moosehead Lake in northwestern Maine. Glacioisostasy has affected the level of Moosehead Lake since deglaciation ca. 12,500 14C yr B.P. Lowstand features at the southeastern end and an abandoned outlet at the northwestern end of the lake indicate that the lake basin was tilted down to the northwest, toward the retreating ice sheet, by 0.7 m/km at 10,000 14C yr B.P. Water level then rose rapidly in the southeastern end of the lake, and the northwestern outlet was abandoned, indicating rapid relaxation of landscape tilt. Lowstand features at the northwestern end of the lake suggest that the lake basin was tilted to the southeast at ca. 8750 14C yr B.P., possibly as the result of a migrating isostatic forebulge. After 8000 14C yr B.P., water level at the southeastern end was again below present lake level and rose gradually thereafter. We found no evidence suggesting that postglacial climate change significantly affected lake level. The rebound history inferred from lake-level data is consistent with previous interpretations of nearby relative sea-level data, which indicate a significantly steeper and faster-moving ice-proximal depression and ice-distal forebulge than geophysical models predict.

Recession of the Laurentide Ice Sheet from northern New Hampshire was interrupted by the Littleton-Bethlehem (L-B) readvance and deposition of the extensive White Mountain Moraine System (WMMS). Our mapping of this moraine belt and related glacial lake sequence has refined the deglaciation history of the region. The age of the western part of the WMMS is constrained to ~14.0–13.8 cal ka BP by glacial Lake Hitchcock varves that occur beneath and above L-B readvance till and were matched to a revised calibration of the North American Varve Chronology presented here. Using this age for when boulders were deposited on the moraines has enabled calibration of regional cosmogenic-nuclide production rates to improve the precision of exposure dating in New England. The L-B readvance coincided with the Older Dryas (OD) cooling documented by workers in Europe and the equivalent GI-1d cooling event in the Greenland Ice Core Chronology 2005 (GICC05) time scale. The readvance and associated moraines provide the first well-documented and dated evidence of the OD event in the northeastern United States. Our lake sediment cores show that the Younger Dryas cooling was likewise prominent in the White Mountains, thus extending the record of this event westward from Maine and Maritime Canada.

This paper describes measurements of concentrations of cosmogenic 26Al, 10Be and 21Ne in quartz from bedrock surfaces in the Transantarctic Mountains where stratigraphic and geomorphic evidence shows that the surfaces were covered by ice in the past, but were not glacially eroded during periods of ice cover. It then explores to what extent this information can be used to learn about past ice sheet change. First, cosmogenic nuclide concentrations in sandstone bedrock surfaces at two sites in the McMurdo Dry Valleys near 77°S are consistent with an equilibrium between nuclide production and loss by surface erosion and radioactive decay. They are most easily explained by a scenario in which: i) sites more than c. 100 m above the present ice surface were almost never ice-covered and eroded steadily at 0.5–1.5 m Ma-1, and ii) sites near the present ice margin experienced similar erosion rates when ice-free, but have been covered by cold-based, non-erosive glacier ice as much as half of the time during the past several million years. Nuclide concentrations in granite bedrock at a site in the Quartz Hills near 85°S, on the other hand, have not reached production-erosion equilibrium, thus retaining evidence of the time they were first exposed to the cosmic ray flux. Nuclide concentrations at these sites are most easily explained by 4–6 Ma exposure, extremely low erosion rates of 5–10 cm Ma-1 during periods of exposure, and only very short periods of cold-based, non-erosive ice cover.