Tunnel Channels and Associated Fan Deposits in Wisconsin, U.S.A.: Insights into the Plumbing of the Southern Laurentide Ice Sheet

Cutler, P. M.1, Clayton, L.2, Mickelson, D.M.1, Colgan, P.M.3, and Attig, J.W.2

 Department of Geology and Geophysics, University of Wisconsin-Madison, Wisconsin 53706

2Wisconsin Geological and Natural History Survey, Madison, Wisconsin 53705, USA

3Department of Geology, Northeastern University, Boston, MA 02115, USA

Abstract

Tunnel channels breached the last glacial maximum margin of the Laurentide Ice Sheet at over 150 locations in Wisconsin.Of a subset of 80 tunnel channels examined in detail on the western margin of the Green Bay Lobe, most are 0.15 to 0.45 km wide, typically between 2 and 7 km long, and 5 to 30 m deep.These dimensions reflect the influence of post-formational sedimentation and slumping, however.The tunnel channels rise up adverse bedslopes of 1 to 20 m/km towards the margin, and most terminate in proglacial outwash fans, confirming that sediment-laden water exited the tunnel channel.We are interested in the origin of these channels, and what they reveal about the subglacial drainage system under this portion of the southern Laurentide Ice Sheet.Clues to their origin lie in the outwash fans, where coarse boulder gravels contain clasts with an average intermediate axis for the largest 10 clasts of about 1.6 m.

Suitably located gravel pits within the outwash fans are rare, but three such examples have been found.Each pit extends to within 100 to 200 m of the glacial maximum ice margin. Exposures are typically 5-7 m high, and consist of laterally extensive (up to 150 m wide), clast-supported cobble and boulder gravel units, each capped with planar and trough cross-bedded sand.At all three fans, up to four such discrete glaciofluvial packages exist below a strikingly coarse clast-supported massive boulder gravel that lies within 2 m of the fan surface. Due to its lateral continuity and extent, and the size of clasts, this boulder unit is interpreted as a jökulhlaup deposit from a sediment-rich proglacial sheet flood emanating from the associated tunnel channel. One puzzle is the apparent lack of evidence for pitted outwash on the fans, as is often observed in jökulhlaup deposits.

Paleodischarge estimates from boulders within this unit are tentative because of uncertainties in the original fluid density, channel width, and the undoubtedly unsteady nature of the ‘event’. Nonetheless, peak discharge was probably between 700 and 5,000 m3/s at all three pits. At the lower end of this discharge range, three water sources are viable: an extreme precipitation event, drainage of a supraglacial lake, or drainage of a subglacial lake. For most of the discharge range, however, the large volume of stored water required implies that the only reasonable source is a subglacial-lake outburst.

Environmental factors were favorable for formation of subglacial lakes behind the southern margin of the Laurentide Ice Sheet in Wisconsin. For example, proglacial permafrost was present as ice advanced to its glacial-maximum extent, and the hydrogeologic setting of the ice margin resulted in inconsequential drainage through the groundwater aquifer. Numerical ice-sheet modelling of the Green Bay Lobe from 55,000 to 21,000 yr B.P. reveals that the frozen-bed zone, formed as ice overrode the permafrost, probably extended on the order of 100 km upstream from the ice margin at the glacial maximum. Permafrost was up to 200 m deep, and the bed gradually thawed from the upstream end until trapped subglacial meltwater was able to drain to the margin. There was ample stored water to move the biggest boulders.

The challenge remains to determine the significance of the jökulhlaup deposit to overall tunnel-channel genesis. Hundreds to a few thousand years probably elapsed before this final large drainage event, based on the existence of multiple units of glaciofluvial gravels beneath the coarse boulder facies, and the fact that at least 30 m and possibly more than 70 m of sand and gravel lies in two of the fans. During this time the majority of the fan was constructed, possibly with additional coarse boulder units at depth, and apparently with the fan apex in a stable position. Proglacial permafrost was concurrently degrading, and supraglacial meltwater was probably able to penetrate to the bed within a few km of the terminus. The final jökulhlaup most likely took advantage of a pre-existing drainage route, enlarging the channel and flushing out the subglacial boulder lag that accumulated during the preceding less competent flow regime. Evidence for jökulhlaups at stratigraphically similar locations in all three tunnel-channel fans raises the possibility that a major release of water occurred just before ice retreated from its glacial-maximum position.