Drumlins and subglacial meltwater floods 

John Shaw Responds

The following is Dr. John Shaw's response to my comments on his subglacial meltwater flood theory, presented here with his permission. The text highlighted in red corresponds to subject headings in the article Drumlins and subglacial meltwater floods. 
Date sent:       Tue, 21 Jul 1998 
To:                 tcc@sentex.net 
From:             John Shaw <John.Shaw@ualberta.ca> 
Subject:          Responses to questions 

Hello Douglas, 

Here are my responses to your comments. 

How did melting occur under an ice sheet?  

Melting at a glacier bed is a result of several energy sources: 

i) Geothermal heat conducted from the Earth's interior; 
ii) Frictional heat generated by sliding at the ice bed; 
iii) Advected heat from surface water or groundwater; and 
iv) Frictional heat produced by viscous dissipation in flowing water. 

What prevents a sheet flow from being concentrated in channels?  

Nothing.  That's exactly what the field evidence tell us.  The sheet flows are transient and, after the formation of drumlin or Rogen fields or hummocky terrain, sheet flow is concentrated in tunnel channels. 

Convergent vs divergent flow patterns?  

Patterns of drumlins do converge into tunnel channels near the former ice margins (e.g., the Finger Lakes, N.Y.)  Elsewhere, patterns diverged as water flowed radially outwards towards distant ice margins (see Prest et al. 1968, Glacial Map of Canada).  In some locations flow diverged over high ground; e.g., the Livingstone Lake drumlin field (Shaw and Kvill 1984). 

Shifts in flow direction 

Flow direction in a pressurized system is determined by gradients in the hydraulic head (i.e. the pressure in subglacial meltwater). If the pressurized system bursts, flow will be towards ruptures, for example breaks in a seal around the periphery of the Laurentide ice sheet. Initially flow will occur through all outlets. Later, when the hydraulic head falls, the higher outlets will be resealed and flow diverted towards the lowest outlets. Hence drumlins show cross-cutting patterns related to the changing availability of outlets as outburst events evolved. 

Flow from out of the sea?  

If the ice were thickest over the sea then the hydraulic head would be highest there and flow would be towards what is now land.  For example, the Laurentide ice sheet flowed uphill from Hudson Bay to the Milk River Ridge in Alberta. 

Size of vortices forming drumlins and Rogen moraine  

In the meltwater hypothesis vortices forming cavity fill drumlins are presumed to be on the scale of the landforms.  Horseshoe vortices forming erosional drumlins, Beverleys, may be of much smaller diameter than the height of the obstacle (drumlin). 

Contemporaneous erosion of lake basins and formation of drumlins  

Yes this appears to have been the case, since lake basins, for example the Lake Ontario Basin, contain only thin or no surficial sediment on bedrock.  Areas adjacent to the present lake have thick deposits, for example the Oak Ridges Moraine, Ontario. 

Drumlin fields are about the same age  

Yes, this seems to be the case for the fields related to the maximum extent of the Laurentide ice sheet.  I have inferred from this that there was some external effect, such as a rapid climate change, on the ice sheet that caused build up of meltwater storage on, within or beneath it. 

What about the imprint of the ice sheet?  

This is a fascinating question.  Airphotos, satellite images and Digital Elevation Models (DEM's) illustrate regional-scale bedforms with very little subsequent reworking.  Consequently, if the meltwater hypothesis is correct, the ice sheet must have been let down gently on the bed as pressure fell in the meltwater system.  At that point, because the ice had been floating, there would have been insufficient slope on the surface to drive ice flow.  Thick ice must have stagnated and melted away. Hence recess. Where did the water come from? 

C. Warren Hunt's comments are out of date; I discuss the origins of the meltwater (Shaw 1996) including references to Shoemaker's papers on the storage and release of subglacial meltwater.  Furthermore, it is quite clear that we realize that the ice sheet could not be floating at the margin.  It is for this reason that I show tunnel channels in the marginal area of the ice sheet (Shaw 1996, Fig 7.45). 

As I have written before, the similarity of our conclusions regarding the mechanisms of formation of drumlins and Rogen moraine is remarkable, given that we worked completely independently of each other.  I appreciate your respectful questioning of the meltwater hypothesis and hope that you find my responses equally respectful. 

Best wishes, 



References Cited

  • Shaw, J. & D. Kvill 1984. A glaciofluvial origin for drumlins of the Livingstone Lake area, Saskatchewan. Canadian Journal of Earth Sciences 21, 1442-1459.

  • Shaw, J. 1994. A qualitative view of sub-ice-sheet landscape evolution. Progress in Physical Geography. 18,2. pp.159-184.

  • Shoemaker, E.M. 1995 On the meltwater genesis of drumlins. Boreas vol. 24 pp.3-10

  • Shaw, J., B. Rains, R. Eyton, and L.Weissling. 1996, Laurentide subglacial outburst floods: landform evidence from digital elevation models. Canadian Journal of Earth Sciences 33, 1154-1168.
 Copyright © 1998 by Douglas E. Cox
The Creation Concept | Controversy About the Glacial Theory