Douglas R. MacAyeal

Binge/purge oscillations of the Laurentide Ice Sheet as a cause of the North Atlantic's Heinrich events

Ice-rafted debris in sediment cores from the North Atlantic suggests that the Laurentide ice sheet (LIS) periodically disgorged icebergs in brief but violent episodes which occurred approximately every 7,000 years. Here, I propose that Heinrich events (i.e., what these episodes are called) were caused by free oscillations in the flow of the Laurentide ice sheet which arose because the floor of Hudson Bay and Hudson Strait is covered with soft, unconsolidated sediment that forms a slippery lubricant when thawed. The proposed Heinrich event cycle has two phases. The growth phase occurs when the sediment is frozen and the LIS is stranded (immobile) on a rigid bed. The volume of the LIS slowly grows during this phase at a rate dictated by snow accumulation. The purge phase occurs when the basal sediment thaws and a basally lubricated discharge pathway (i.e., an ice stream such as those which occur in West Antarctica today) developes through Hudson Strait. The volume of the LIS rapidly equilibrates to the reduced basal friction during this phase by dumping icebergs into the Labrador Sea. The periodicity T=π/κ(−kθsl2G∼)2≈7000 years of the proposed Heinrich event cycle is a function of the thermal conductivity and diffusivity of ice, k and κ, respectively, the atmospheric sea level temperature θsl (in degrees Celsius), and the excess geothermal heat flux defined by G∼=G−kΓ where Γ is the atmospheric lapse rate, and G is the geothermal heat flux. Agreement between the predicted T and the apparent periodicity implied by the marine record is the main virtue of the free oscillation mechanism I propose. An alternative mechanism in which Heinrich events are forced by periodic variations in external climate is implausible, because periodic atmospheric temperature perturbations are strongly attenuated with depth in an ice sheet.

Ice‐rafted debris associated with binge/purge oscillations of the Laurentide Ice Sheet

The North Atlantic sediment record suggests quasi-periodic (7000- to 12,000-year period) ice-rafted debris (IRD) depositions during at least the last glacial period. The cause of these Heinrich events, as they are commonly known, is not fully understood; however, they may point to surges of the ice stream that drained the Hudson Bay/Hudson Strait region of the Laurentide Ice Sheet. We investigate a simple conceptual model of ice stream instability (the binge/purge model) to suggest ways in which the ice stream could have entrained sufficient debris to account for the estimated mass of IRD associated with a typical Heinrich IRD layer in the North Atlantic (1.0 ± 0.3 × 1015 kg). We find that freezing of debris-laden ice at the bed of the ice stream during the brief (≈ 750 years) surge phase of the ice streams hypothesized binge/purge cycle can incorporate up to 5.1 × 1015 kg. This amount is sufficient to meet the constraints of the North Atlantic sediment record but by no means verifies the binge/purge model as the cause of Heinrich events.

Acceleration of Pine Island and Thwaites Glaciers, West Antarctica

Abstract Recent satellite investigations revealed that in the 1990s the grounding line of Pine Island and Thwaites Glaciers, West Antarctica, retreated several km, the ice surface on the interior of the basins lowered 10 cm a–1, and Pine Island Glacier thinned 1.6 ma–1. These observations, however, were not sufficient to determine the cause of the changes. Here, we present satellite radar interferometry data that show the thinning and retreat of Pine Island Glacier are caused by an acceleration of ice flow of about 18 ± 2% in 8 years. Thwaites Glacier maintained a nearly constant flow regime at its center, but widened along the sides, and increased its 30 ± 15% mass deficit by another 4% in 4 years. The combined mass loss from both glaciers, if correct, contributes an estimated 0.08 ± 0.03 mma–1 global sea-level rise in 2000.

Catastrophic ice-shelf break-up by an ice-shelf-fragment-capsize mechanism

Two disintegration events leading to the loss of Larsen A and B ice shelves, Antarctic Peninsula, in 1995 and 2002, respectively, proceeded with extreme rapidity (order of several days) and reduced an extensive, seemingly integrated ice shelf to a jumble of small fragments. These events strongly correlate with warming regional climate and accumulation of surface meltwater, supporting a hypothesis that meltwater-induced propagation of pre-existing surface crevasses may have initiated ice-shelf fragmentation. We address here an additional, subsequent mechanism that may sustain and accelerate the ice-shelf break-up once it begins. The proposed mechanism involves the coherent capsize of narrow (less than thickness) ice-shelf fragments by rolling 90° in a direction toward, or away from, the ice front. Fragment capsize liberates gravitational potential energy, forces open ice-shelf rifts and contributes to further fragmentation of the surrounding ice shelf.

A low‐order model of the Heinrich Event Cycle

If Heinrich events result from free oscillations in the size and basal melting conditions of the Laurentide ice sheet, a rigorous quantitative analysis of ice sheet physics should describe their dynamics. To explore this possibility, I exploit two characteristic timescales which arise from ice sheet physics to construct a relaxation oscillator model of the North Atlantics Heinrich events. The numerical implementation of this model confirms the notion that the periodicity of Heinrich events (approximately 7,000 years) is determined by the gross properties of a steady glacial climate (e.g., an annual average sea level temperature of −10° C and an adiabatic lapse rate).

A tutorial on the use of control methods in ice-sheet modeling

Control methods are recommended as an efficient means to estimate undetermined physical parameters and boundary conditions and, in so doing, to improve the fidelity of a given ice-sheet flow model to specific ice-sheet velocity observations. To accomplish this task, the underlying dynamics of the model are inverted. This permits model-tuning adjustments to be represented explicitly in terms of model/observation misfit

The basal stress distribution of ice stream E, Antarctica, inferred by control methods

The irregular spatial distribution and velocity independence of basal friction derived from Landsat measured surface velocity suggests that ice stream flow is not controlled by the properties of a deformable basal till alone. Rigid bedrock substrata may contact the base of the ice stream in small (<100 km2) areas where the velocity field displays strong vorticity and where the ice stream surface appears rumpled in Landsat images.

DEGLACIATION OF A SOFT-BEDDED LAURENTIDE ICE SHEET

We present a series of numerical reconstructions of the Laurentide Ice Sheet during the last deglaciation (18–7 14C ka) that evaluates the sensitivity of ice-sheet geometry to subglacial sediment deformation. These reconstructions assume that the Laurentide Ice Sheet flowed over extensive areas of water-saturated, deforming sediment (soft beds) corresponding to the St. Lawrence lowland, the Great Lakes region, the western prairies of the U.S. and Canada, and the Hudson Bay and Hudson Strait regions. Sediment rheology is based on a constitutive law that incorporates experimental results from late Wisconsin till deposited by the Laurentide Ice Sheet which suggest only mildly nonlinear viscoplastic behavior. By varying the effective viscosity of till, we produced four reconstructions for the ice sheet during the last glacial maximum 18 14C ka, and two reconstructions each of the ice sheet at 14, 13, 12, 11 and 10 14C ka. We also produced one reconstruction for 9, 8.4, 8, and 7 14C ka. Reconstructions that assume a low effective viscosity for all areas of deforming sediment show a multidomed ice sheet with a large bowl-shaped depression over Hudson Bay and thin ice (<1000 m above modern sea-level) over the western and southern margins. Those reconstructions that assume a higher effective viscosity of till in the Hudson Bay region than for the western and southern margins also show a multidomed ice sheet but with considerably thicker ice over Hudson Bay and a more northerly position of the central ice divide. These two different geometries may represent ice-sheet orographic changes associated with a Heinrich event. Further increases in effective viscosity of till, approaching the effective viscosity of ice, would result in a high, monolithic ice dome centered over Hudson Bay, reinforcing the notion that a multidomed ice sheet reflects the distribution of substrate geology. Modeled ice-surface geometry at the last glacial maximum shows many of the same general features as previous reconstructions that incorporate deformable beds. Our reconstructions with higher effective till viscosities in Hudson Bay also agree with the ICE-4G reconstructions (Peltier, 1994), which are based on inversion of relative sea-level data, for the early part of the last deglaciation (18–13 14C ka), but then depart significantly from ICE-4G beginning at about 12 14C ka due to differing assumptions of the history of deglaciation. Modeled ice volume for the last glacial maximum suggests a glacioeustatic change of 50–55 m by a soft-bedded Laurentide Ice Sheet; this would increase as the effective viscosity of till increases. Subsequent ice-volume changes through the last deglaciation generally parallel the trend of eustatic rise recorded at Barbados, New Guinea, and Tahiti, but suggest that the Laurentide Ice Sheet was not the source of meltwater pulse 1A.

Numerical reconstruction of a soft-bedded Laurentide Ice Sheet during the last glacial maximum

We used a numerical ice-sheet model to reconstruct the North American Laurentide Ice Sheet during the last glacial maximum. Our model simulates ice-sheet conditions that can be specified experimentally as either a rigid substrate (hard bed) or a wet, deformable till (soft bed); basal sliding is excluded. We use geologic records of former basal ice-sheet processes to prescribe the distribution of hard and soft beds. Our reconstruction of the Laurentide Ice Sheet is significantly lower in ice-surface height and contains less ice volume than the CLIMAP (maximum) reconstruction. In contrast, our reconstruction agrees well with the ICE-4G reconstruction, both in height and volume. Because the ICE-4G reconstruction is based on the inversion of relative sea-level data, whereas our reconstruction is based on glacial geology and ice mechanics, this agreement suggests that soft beds provide a glaciological mechanism to explain the shape and volume of the Laurentide Ice Sheet that is most consistent with observations of relative sea-level change and other geodynamic considerations.

Basal friction of Ice Stream E, West Antarctica

We use surface velocity derived from sequential Landsat imagery and a control method to estimate the basal-friction distribution of a major West Antarctic ice stream. The area-averaged basal stress is approximately 1.4×10 4 Pa, or about 29% of the area-averaged driving stress of 4.9×10 4 Pa. Uncertainty of the derived area-averaged basal stress is difficult to assess and depends primarily on spatial variation of the flow-law rate factor in the constitutive law. Spatial variation associated with depth-averaged temperature variation gives an uncertainty of approximately ±10 3 Pa. Approximately 60% of the ice stream has a basal-stress magnitude less than 10 4 Pa, and approximately 30% has less than 10 3 Pa. These characteristics suggest the presence of a mechanically weak, water-charged subglacial till. Small-scale sticky spots where basal friction exceeds the area-averaged driving stress are scattered irregularly across the subglacial regime and comprise approximately 15% of the ice-stream area. Sticky spots cluster in regions where Landsat imagery suggests structural features in the underlying bedrock.