Where was I?

I am horrified to find that it has been nearly 2 years since my last entry on the blog. To the vast numbers of you avidly following this blog (sic) my apologies, but it has been a busy 2 years in areas that perhaps I might at some time in the future have a chance to elaborate.


My last posting dealt with the difficulties experienced by Henry Moseley in having accepted his ideas on what could and could not be powering the motion of glacial ice, and how through a weird echo I found myself quite independently putting forward a very similar explanation for at least some forms of glacial motion some 150 years later. There is I am sure a lot more to be said on this subject, and I am sure to return in later postings to some of the issues raised. But for the moment I would like to turn to some of the events which encouraged another of my little excursions into the realms pavement engineering.

Soon after my initiation into the subject of pingos I started what has been a lengthy correspondence with Professor Ross Mackay, undoubtedly the world’s foremost authority on the subject of pingos (see some of the earlier postings). I cannot recall now whether it was his suggestion or whether the subsequent correspondence with Professor Chris Burn (an ex graduate student of Ross) was initiated independently. But one of the useful aspects to have emerged from the correspondence with Chris was his suggestion that as an analogue for the thermal ratchet model I had been arguing to form part of the thermal buckling explanation for the emergence of pingos, I would do well to look at the literature dealing with ice-wedge polygons. Again, this was a subject with which I was not at all familiar. But a quick reading of the fabulous book "geocryology” by Professor Washburn [8] soon made it apparent that there was indeed a direct link. For those of you who like me at the time are unfamiliar with what ice-wedge polygons are I have included a few pictures below. There genesis is
(a)
(b)
(c)
(d)

         Examples of ice-wedge polygons in permafrost regions. Note the relic formation in (c).

fairly well established as being due to cycles of warming and cooling of the upper few metres of permafrost as a result of the annual seasonal cycles of surface temperature. Initial formation of basically discrete ice-wedge cracks within permafrost is fairly quickly (at least in geological timescales) followed by the growth of bifurcated cracks that start to link-up with other discrete cracks and their branches [17-19]. Eventually, networks of polygonal cracks will form, that for fairly self-evident reasons are often referred to as “ice-wedge polygons”. Ice-wedge polygons are ubiquitous in current regions of permafrost. They are also common in relic form in areas that experienced permafrost conditions during some past extended period of colder climate; such a relic formation is shown in (c) above. Like pingos there is considerable interest in the geomorphic relics of ice-wedge polygons that remain after the retreat of permafrost. They provide significant indicators as to the extent and the intensity of previous fluctuations in climate. The existence of relics as shown in picture (c) demonstrates the distance reached by permafrost conditions during the most recent glacial period – or even an earlier glacial period if it has not been wiped-out by the advance of subsequent and even more extensive ice sheets. The characteristic diameter of the random, but in an averaged sense, polygonal shapes, indicates the intensity of the seasonal variations in temperature experienced at that latitude. It would seem that for a given form of ground material conditions the depth to which the thermal wave extends into the ground has a strong determining influence on the distance between adjacent cracks. For anyone brought up on St Venant’s theorem within the subject of elasticity this relationship between the depth of stress relief due to cracking and distance to which this localize stress disturbance propagates would make it clear as to why the crack separation is directly linked to the crack depth. And of course the depth of the crack has a direct relationship to the depth of penetration of the thermal disturbance which in its turn relates to the severity of the maximums and minimum seasonal temperatures. So that the study of the morphologies of crack wedge polygons has taken on great significance in attempting to understand palaeoclimates, both on earth and some of the other planets and their moons. For this reason there is growing interest in the research being carried out on the growth and decline of ice-wedge polygons.

The process whereby over periods measured in millennia – indeed over glacial periods – the ice wedges grow to become as wide as a few metres in width at the surface, is roughly as follows. During the extreme cooling of late autumn through to mid-winter the upper few meters of permafrost will undergo cooling. The restraint of the contraction that would otherwise take place when cooled (and permafrost has a very high coefficient of thermal expansion so the thermal strains are really quite high) induces high horizontal tensile stresses. Well they would were it not for the fact that permafrost like ice has a low tensile strength so that it cracks at a fairly early stage of the cooling. Moisture getting into these cracks quickly turns to ice. So that when the ice is warmed from early spring the cracks do not close, but instead there is left an additional vertical slither of ice with thickness variation vertically that reflects the original crack width. Over many years this repeated cycle results in the gradual widening of the ice wedge.

The above pictures show a few examples of both current and relic ice-wedge polygons. Current ice-wedge polygons often show-up as distinct contrasts in colour, sometimes as a result of the differences in vegetation supported by the very different ground conditions within the wedge areas compared with the rest of the ground. They, but especially the relic forms, also show up as differences in elevation between the central regions and the polygonal boundaries. For a more complete account of the different forms taken by active and relic ice-wedge polygons see for example [8,9,17]. While the permafrost literature is relatively silent on the subject, it would seem that many of the differences in elevation of the ground within the ice-wedge polygons could be accounted for in terms of the variety of compression related failures that occur during the constrained expansion that accompanies the subsequent warming.

As I have been arguing as the basis of a possible genesis of pingos it has even been suggested that these compressions are capable of causing the upheaval failure into the characteristic dome shaped mode. There is some evidence that at the rather smaller scale of typical ice-wedge polygons a similar form of general buckling uplift could be occurring. Washburn [8] reports cases where massive ground ice, in the form of lenses, exist beneath the centres of ice-wedge polygons. This would be consistent with a form of seasonally induced compressive uplift buckling at wavelengths consistent with the inter-crack spacing. In other cases there appears to be an accumulation of material either side of the troughs within which the ice-wedges occur. These so called “ramparts” have all the characteristics of a local form of compression related shoving failure, much like that forming when a rug on a smooth floor is pushed up against a wall. They would appear to have origins in the shoving action of the seasonally frozen, or even the perennially frozen, ground as it expands during warming. This outward expansion resulting in an effective outward transport of material is similar to the development of ice shove ridges around the edges of lake-ice [22,23,24]. Just how the compressive energy is released during the warming phase is complicated by the interactions between the permafrost and the overlying “active layer “ – the active layer being the soil above the top of the permanently frozen ground that is only seasonally frozen. Excellent long-term field evidence of this cyclical outward motion of the active layer is summarised by Mackay [17]. The mechanism at work in this outward mass movement would appear to be closely related to that recently hypothesised as being responsible for the mass movements involved in the formation of sorted and non-sorted, stone and rock, circles, polygons, nets etc [25]. There appear to be few explicit references to this form of compression related failure accompanying the release of compression energy generated when restrained expansion occurs during warming.

So while temperature induced fluctuations in tensile strain are widely used to account for the development of the ice-wedges of ice-wedge polygons, curiously, there appears to be relatively little acknowledgement that compressive straining during warming might account for some of the other observed features of ice-wedge polygons.

Anyway, the reason for this diversion into ice wedge polygons is by way of introduction to a return in my next posting to asphalt in order to explore whether similar processes might be responsible for what is one of the most common and costly forms of asphalt pavement failure – that of alligator (or in N America, crocodile) cracking.