Mickelson, D.M., Cutler, P.M., and Colgan, P.M., 2001, Southern Laurentide Ice Sheet drumlin and flat till plain landform assemblages and the distribution of permafrost: field observations and model results: INQUA Commission on Glaciation, 6^th International Drumlin Symposium, Torun, Poland, Abstracts of Posters and Papers, p. 19-20.

Ice-lobe behavior, and the landform assemblage that is left behind, is a result of interaction between ice, climate, geology, and topography; the relative contributions of which vary greatly through time and space. Flat till plains must be formed where there are uniform depositional and erosion processes acting at the glacier bed during the whole period of ice cover. Basal motion, through some combination of sliding, ploughing, and bed deformation is likely to be a major contributor to ice movement at low driving stresses.Whatever the processes, they must be fairly uniform over areas of 100s of km². In contrast, the formation of drumlins seems to require heterogeneity of the bed, and therefore heterogeneity in processes and shear stresses, to explain their existence. These heterogeneities result in the preferential accumulation of till and the preferential erosion and transport of sediment from inter-drumlin areas toward drumlins and toward the ice margin. In the case of the Green Bay and adjacent Lake Michigan lobes of Wisconsin, more than 5000 drumlins formed during the last glacial maximum. Although many drumlins are composed of homogenous basal till, and therefore have an uncertain origin, some are composed of sand and gravel or lake sediment and were clearly deposited prior to the last glacial maximum. This indicates that part of the drumlin forming process is erosional and part is deformational. It has been previously suggested that drumlins may have formed in a zone behind the ice margin where part of the bed was frozen and part thawed. In the drumlin area ice wedge casts and other evidence clearly shows that ice advanced over permafrost. Results of a 2-dimensional numerical ice-flow model driven by climate also indicates that along the Green Bay lobe flow line thick permafrost was overridden by ice and that permafrost beneath the glacier slowly decayed in a zone 100 km wide behind the margin. This decaying permafrost likely produced the heterogeneities required for drumlin formation. Climate was evidently cold enough for permafrost to be re-established between some readvances, but eventually warmed sufficiently to have a sliding bed that produced striations before ice retreat. The flow line along the main axis of the Lake Michigan lobe produced a flat till plain, but no drumlins.

Pat and Paul-Here is the problem (see Paul's comment below). I gather that the LML model run didn't go all the way to the LGM maximum because it wouldn't advance that far? I guess we can't say what I thought we could. This also creates problems because Kelly's LML sediment flux is all based on the LML flowline south of Chicago, so it won't be very meaningful as it relates to the model.

Any suggestions on where to go from here?

ORIGINAL VERSION

The flow line along the main axis of the Lake Michigan lobe, which produced a flat till plain, but no drumlins, extended further south and model results and the presence of extensive spruce wood indicated that the ice did not overrun permafrost, but had a wet bed all the way to the margin. Limited permafrost probably did develop after the glacial maximum in this area, but the bed of the glacier remained wet. Thus, sliding with a uniform depositional regime existed throughout the period of ice cover in

this area.

PAUL"S COMMENT-- [DAVE - model runs are only wet-bedded for LML because simulated

extent does not get south of the lake in these runs.

Under the climate conditions in the model, ice would have overrode some

permafrost at the south of the lake. However, the

climate in the model assumes modern temperature gradients with respect to

latitude. The gradient may well have been steeper at

LGM, so conditions allowing permafrost in Wisconsin might have been compatible

with much warmer conditions in IL than are

represented in the model.]

This flow line extended further south and model

results and the presence of extensive spruce wood in tills indicates that the

ice did not overrun permafrost, but had a thawed bed all the way to the

margin. Limited permafrost probably did develop after the glacial maximum in

this area, but the bed of the glacier remained wet. Thus, sliding with a

uniform depositional regime existed throughout the period of ice cover in this

area.