Friday, 26 September 2014

Glacial Loess: revisited, reviewed, reconsidered

Matti Seppala (2004, p.213) wrote that " Smalley (1966) proposed that glacial grinding provided loess material and the idea was supported by Boulton (1978, p.796)." Another sentence from Seppala (2004, P.117) "Butzer (1965) defined loess according to its origin as two types (1) periglacial loess, deflated from outwash deposits, from freshly exposed till, and from barren rock and tundra surfaces, and (2) desert or continental loess originated from desert areas."
What Butzer was saying was that loess material for loess deposits came from periglacial regions and desert regions. This is actually quite a reasonable statement provided that one realises that the loess material may have been introduced into the periglacial or desert areas- the actual source of the particles, the place where the particles were made, may be elsewhere. There have been long years of confusion when it was thought that material for desert loess deposits had to be made in the desert- rather than being simply stored there, or passing through. The loess derived from the Central Asian deserts is made in the mountains of High Asia, probably by the action of mountain glaciers.

The Butzer (1965) work was quoted by Smalley (1966). What Smalley (1966) did not quote or cite was Hardcastle (1889), a paper in which an approach to glacial loess was offered which was remarkably similar to that produced 70+ years later. Hardcastle pointed directly to the problem of producing loess material, and offered a solution in the form of the cold phase glaciers.

This blog is about two aspects of glacial loess; about the possibility that the loess in Western Pomerania, in Poland, is glacial loess; and about the results of some experiments with a Janet Wright (1995) glacier machine(in which the formation of loess stuff can be modelled).


Is the loess in Western Pomerania distinctive?  The map suggests that the material is glacial material, deposited by glaciers across mid-Poland, and then transported by west flowing rivers into the region of W.Pomerania- to form a loess deposit. In the simple deterministic view of loess formation this looks reasonable. Observations on the ground (vide KI) may support this contention.

The deformation results(vide KOHD et al) show the response of sand to shear stress in a Janet Wright (1995) glacier machine. This machine is a modified Bromhead ring shear testing machine, designed initially to test the shear strength of clay soils. It makes a passably acceptable model glacier- via a few simple modifications. The figure shows a typical result of sand deformation; the sample is placed in an annular chamber and a continuous stress can be applied. The height (thickness) of the specimen is measured (vertical axis in figure). The stress set up can be adjusted to be similar to that found in a real glacier system. Long term tests are possible (in figure up to 24 hours)- as the test proceeds the sand deforms. The stages can be explained: stage one is simple dilatancy- when a cohesionless granular material is sheared it expands (this was the basis of the Smalley-Unwin drumlin forming model). Then the macro-defects are activated, any major cracks in the sand grains allow rapid breakage so system height reduces quickly- this is stage 2. Stage 3 is the critical stage when the internal defects in the quartz particles are activated and silt sized material is produced. Here are the Moss defects controlling the size of loess particles. At the end of stage 3 material for a loess deposit has been produced. This is the key to the size of particles in loess deposits. The mode particle- the coarse silt sized quartz particle is controlled in terms of size and production by the defects in the quartz particles derived from the initial granitic rocks.


Seppala, M.  2004.  Wind as a Geomorphic Agent in Cold Climates.   Cambridge University Press 358 p.

No comments:

Post a Comment