Jw. Jenson et al., NUMERICAL MODELING OF SUBGLACIAL SEDIMENT DEFORMATION - IMPLICATIONS FOR THE BEHAVIOR OF THE LAKE-MICHIGAN LOBE, LAURENTIDE ICE-SHEET, J GEO R-SOL, 101(B4), 1996, pp. 8717-8728
We apply a numerical model of the late Wisconsin (circa 20,000 years B
.P.) Lake Michigan Lobe (LML), Laurentide Ice Sheet, to investigate ho
w fine-grained subglacial sediment might influence lobe behavior, part
icularly rapid millennial-scale marginal oscillations observed in the
geologic record. Over the Canadian Shield, we assume a rigid bed (''ha
rd bedded'') basal boundary condition. In areas overlain by fine-grain
ed sediment (''soft bedded''), the base of the ice is coupled to a def
ormable sediment layer, using a rate-dependent stress-strain law. Geot
echnical tests of clay-rich till deposited by the LML provide control
for sediment theologic parameters. Simulated cross-sectional profiles
are consistent with reconstructions from geologic evidence. Time-depen
dent simulations suggest that a soft-bedded lobe could have reached st
eady state in about 20,000 years or less, in contrast to 50,000 to 60,
000 years for an otherwise identical hard-bedded lobe. A soft-bedded l
obe with sediment viscosity at the experimentally determined value is
about twice as responsive to millennial-scale shifts in accumulation o
r ablation as a nonsliding hard-bedded lobe, but in both cases the res
ponse is slower than that indicated by the geologic record. Results su
ggest that while strong millennial-scale changes in accumulation and a
blation can produce responses in hard-bedded or soft-bedded ice that a
re consistent with the geologic record, changes in subglacial sediment
viscosity, even relatively modest changes (whether independent or in
conjunction with climate change), might more readily account for mille
nnial- and submillennial-scale fluctuations of the lobe margin. These
observations do not exclude a role for sliding, but they do provide so
me perspective from which to evaluate relative contributions of the va
rious processes that influence lobe behavior.