Tuesday 16 March 2010

musings on the origins of ice rings






Having been overwhelmed with responses to my earlier observations of ice rings on New York's Central Park reservoir and being unable to bear the suspense of awaiting for all your theories as to how they might have formed, I feel compelled to advance some of my own hypotheses. Certain of the ideas expressed below are related to views I have been trying to encourage as to the thermo-mechanical origins of much wider classes of periglacial morphologies. And it has partly been some of the frustration of not having been able to find a satisfactory outlet for these views, or being able to have them taken seriously by the experts in these fields, that has encouraged this blog. But more about some of these wider musings in a later blog.

As was shown in the 2 photographs posted on the previous blog, there appear to be two distinct types of ice-ring. Each would seem to be the result of distinct thermal-mechanical processes responsible for their development.

Type 1 ice-ring (are shown in the lower of the above 2 pictures): It was interesting that most of this class of ice-ring had formed in zones of lake ice that were within a 50m distance of the footpath following the reservoir edge. Perhaps of significance, this zone was also about the maximum distance over which it is possible for someone to lob a snow ball or some other object onto the ice sheet. In other pictures not shown here a series of almost spherical balls of ice could be seen to be resting on the surface of the ice sheet. Most were observed to be sitting in locally depressed areas of the sheet ice. Other ice balls were observed to be in more advanced depressions where the balls and the supporting ice had all but disappeared. Having a relatively large surface area exposed to fluctuations of solar radiation these balls of snow melted fairly rapidly each time the temperature exceeded 0oC. The latent heat contained within this melt water seemed to concentrate and accelerate the surface melt of the ice sheet in the zone surrounding the surface ball. Eventually the ice balls melted completely leaving circular areas of either very thin or at times fully melted ice. Why these circular depressions would then give rise to the build-up of crushed ice around the periphery of the circular depression to form the distinctive ramparts could have been the result of subsequent cycles of heating and cooling. [Incidentally, I did try a few tests of my own. Unfortunately, the weather had changed the day after I had lobbed my own snow balls onto the ice. Heavy intervening cloud cover meant that my snow balls had not experienced sufficient radiant energy to cause the anticipated melting. Because by the time of my next visit the ice sheet had melted in this area my own little test was never completed!]

But getting back to the first hypothesised origin of the ice rings, each time the temperature increases the surrounding ice would want to expand into the thinned area and the thinned ice within the depression would want to expand outwards. The total relative movement would be constrained and the comparitively high strength of the surrounding thicker ice would cause the outer boundary of the thin central disc of ice to be crushed. Over a number of cycles of heating and cooling this would have the effect of heaping-up the crushed thinned ice around the edges. Since the thin ice layer would crack when subject to tensile stress during cooling the compressive crushing strains would not be recovered after a full cycle of heating and cooling. With many such cycles occurring over the period of a single day it becomes possible to conceive of a mechanism capable of building up the peripheral circular ramparts.

It is possible too that where the local melting had exposed a circular pool of water a further contribution to the build up of the rampart rings might have been wind. Small wind generated ripples would cause the free surface water to splash over the edge of the hole. Some of this splash water would be refrozen and assist in the build-up of the ring ramparts. This process can often be observed at locations where lake ice has experienced crack fractures. Splashing over the edges of the ice on both sides of such cracks induce the build-up of ice ridges. Upon warming these open fractures often reclose leaving a characrteristic double ridge following the line of the previous crack.

On the day following the above observations there was heavy cloud cover and little variation in the sub-zero temperatures. It was interesting to observe over this period a gradual flattening of the ice-ring ramparts and an associated thickening of the ice within the circular regions.

Type 2 ice-ring (are shown in the upper of the above 2 pictures): The ice-rings shown here have quite different forms to those shown in the lower picture. The border rings comprise discrete slabs of thin ice that have clearly fractured from what had earlier been a flat sheet of ice. How they got into these elevated and rotated positions is possibly due to a somewhat different mechanism.

Around the edge of the cluster of ice rings shown in the upper picture can be seen a few smaller embryonic rings for which the ice sheet is still largely intact but has been thrust up to form small diameter domes of ice. One of these is visible at the lower left of the top picture. The top has been partly melted away by the heat of the sun to reveal the cavity beneath. The cause of these ice blisters is likely to be a form of thermal uplift buckling like that recently described as being responsible for the blistering of asphalt pavements(1,2) and at larger scale frozen ground to form features such as seasonal frost mounds, palsas and perennial pingos(3,4). When the ice sheet is heated diurnally the ice being restrained will develop horizontal compressive stresses. For the thin ice at the edge of the Central Park reservoir, it would not take large increases in temperature within the ice to induce a thermal uplift buckle. The diameter of the characteristic dome shape depends inter alia upon the ice thickness and the level of compressive stress reached during the diurnal warming period. That the uplift buckle is not fully recovered during nocturnal cooling would again be the result of the ice cracking under the tensile stresses developed during the restrained shrinkage. It would take just a few cycles of heating and cooling for the ice domes to reach the heights and diameters typical of the ice rings shown in the top picture above. Fracturing of the dome to leave the debris rings shown in the picture would have occurred once the melting of the top portion of the dome had spread far enough that the then open shell of ice would be no longer capable of developing the very efficient load carrying capacity provided by the membrane stresses available for the complete dome. At this stage the open water beneath the dome would again start to ice over, leaving the remains of the ice dome in the form of the debris ice-rings shown in the top picture.

With the ice beneath the relic dome being thinner than the surrounding ice, alternations in solar energy, caused by breaks in the cloud cover, could see the mechanism described above for the development of the type 1 ice ring starting to add to the piling up of ice around the peripheral ring.

Then again, the type 1 ice rings could just be the result of a longer term melting process operating on the discrete edge accumulations characterising the type 2 ice rings.

Anyway, these were some of my thoughts while wandering around New York during the cold snap in February. Still waiting to hear from anyone who has observed similar features and perhaps given some thought to why they might occur.

ps Shortly after these observations I prepared the paper ice-rings for which it transpired I could not find a suitable publication outlet which would encourage discussion of this curious phenomenon. It was for this reason that the blog was started.

4 comments:

  1. I suggest that next time you find something like this, you quickly set up a time-lapse video recorder so that you can observe the whole process of development of such rings. You might also like to artificially create such thermal cycles in lab in front of a time-lapse camera and investigate your theories.

    Wish you the best investigating and blogging...

    ReplyDelete
  2. Dear nogeek. You are quite correct, this would indeed be a very satisfying means of testing the hypothesised thermal-mechanical processes being suggested for the origin of the ice blisters. Indeed, a few years ago in an application for research funding to investigate what I believe is a closely related process in the formation of asphalt pavements (see a later blog on this subject) we included a proposal to incorporate monitoring using time lapse photography with suitable markers on the evolving blisters along with continuous temperature measurements to enable quantitative modelling of the growth rates. Needless to say the application for funds was not successful. But given the prevalence of the problem in asphalt pavements I am very hopeful that someone else with time on their side will eventually take up this challenge.

    Whether I will just happen to have time lapse photographic equipment next time I am walking around an ice bound lake is of course quite another matter!

    ReplyDelete
  3. I was greatly intrigue to read your reports on the curious rings that had formed on the ice covering the reservoir in Central Park and have to confess that despite often running around the reservoir (even when it is covered with ice) I have not observed any of these ice rings. But that probably says more about my rather single minded pursuit of fitness than the absence or presence of ice rings!
    I was also interested in your possible explanations for the formation of the ice rings. Is it possible that the mechanism could actually be a combination of those you describe? A melting snow ball or some dark object thrown onto the ice surface, such as a stone, might be anticipated to melt a small patch of ice beneath. The snowball for the reasons you suggest associated with the concentrated transfer of latent heat, and a dark object as a result of its higher thermal conductivity and "black body" greater absorption of radiant energy. Diurnal melting of the thin ice sheet to form a small effectively circular hole would be anticipated to ice-over in the subsequent nocturnal cooling. The result would be a small circular disk of relatively thin ice surrounded by the older and rather thicker ice layer. During the next period of diurnal warming the relief of the expansion strains experienced by the thick surrounding ice sheet at the now thin disk of ice covering the hole, would induce high compressive stresses in the thin disk. It is easy to imagine these compressions to be sufficient to cause a buckle much like those shown in some of your photographs.
    Just a thought - but thanks for bringing these curiosities to my attention.

    ReplyDelete
  4. My thanks sottovoce uno, I was beginning to think that my blogs were being sucked into that great black hole lurking somewhere out there in the cosmos!

    Your observation on the capacity of black bodies to drill their way through the ice sheets to create essentially circular holes was something I too have often observed. In the conditions that prevailed on the Central Park reservoir, stones thrown onto the ice would during any sunny period fairly quickly develop depressions, which would become filled with melt water. When eventually the thinned ice disk could no longer support its weight the stone would break through and sink to the bottom of the reservoir leaving a small ice free circular hole. Provided the weather patterns were appropriate these circular holes would no doubt freeze over nocturnally (or even when heavy cloud cover caused temperatures to drop sufficiently) forming the thinned disks of ice which became the locations where blister buckling sometimes occurred. Collapse of these blisters seemed to then provide the origin for one of the forms of ice ring.

    That the inward radial displacements at the boundary of a hole, occurring when the average temperature of the ice sheet increases, could be enough to cause a buckling failure of the relatively thin ice disk is I think fairly clear from what happens when the ice sheet undergoes cooling shrinkage. When the average temperature of the ice sheet drops sufficiently the relief of the radial tensions at the hole boundary will be accompanied by a build-up of hoop tensions around the inner edge of the hole. Photos reproduced in my blog "more evidence of thermal ratchetting in lake ice" of 16 March, 2010, show that these tensions can often result in a series of radial cracks penetrating into the ice sheet. Outward radial displacements of any thin disk of ice occupying the circular hole would in all probability suffer cracking to relieve the associated tensile stresses. These cracks would inevitably fill with water which would freeze to prevent the cracks from closing during any subsequent warming, thereby ensuring that the increased temperature would induce compressive stresses in the thin disk - possibly sufficient to buckle it or in other circumstances cause it to crush at its boundary thereby adding to the development of the ice rings.

    I am wondering if what we are discussing here might also be a useful analogy for some of the later discussions I intend to post relating to the possible growth of pingos (see posting "from asphalt to permafrost blisters" of 25 March, 2010), and indeed other forms of surface mound that emerge from seasonally or perennially frozen ground - an interesting thought!

    ReplyDelete