Rock Circles and Polygons as thermal ratchets

In an earlier post it was suggested that the process responsible for stones, rocks and other inclusions of greater than average grain size, being gradually brought to the surface is a form of thermal ratchet, in the sense being developed here. The result is that stones that were present within the depth of soil equal to the maximum depth of the seasonal  

 

Side-shift and Thermal Ratchetting

While reasons have in the past been proposed for the horizontal movements needed to form such curious features as stone circle, polygons, etc, none of them appear to be entirely convincing. And while it is likely that no single explanation will be able to account for all such motions, the following would appear a feasible explanation for at least some features outlined in classic texts such as (Washburn 1979; French 1996).         

To illustrate how a thermal-mechanical ratchetting process could be at work in the horizontal component of motion of stones, consider a typical stone lying at surface level a distance h from the centre of an already forming stone circle having radius r. Once again stone is used as a generic descriptor for an object of size greater than the average, and circle will be assumed to cover the case of polygons, nets, etc. While fascinating, the genesis of the characteristic dimensions of the stone circles is not here at issue. What we are concerned with is the process that continues to drive the outward movement of the stones once the basic geometric patterns have been defined.

Following the above annual surface thermal cycle, suggested in Figure 6(a), the stone at a time (1) towards the end of summer might be sitting at location distance h from the centre of the circle. A short time later, time (2), frost will set-in at the surface. Because the outer edge of the circle is likely to be grounded within the accumulated debris and large mass stones at the edge of the circle, the horizontal contractions wanting to take place across the width of the circle will be restrained. The resulting tension stresses will at a fairly early stage induce surface cracks to form that will eventually extend some distance, related to the maximum depth of the seasonal frost layer, into the frozen soil. Wind-induced desiccation cracking could also help to relieve the tensile stresses otherwise developing during these restrained thermal contractions. This situation is suggested in Figure 6(b) by the section through the stone corresponding with time (3). Moisture and snow getting into these cracks will quickly freeze, so that when the warming period commences it is likely that the cracks will have been welded together so as to present a relative continuous sheet of frozen ground for the subsequent expansion induced compressions. By the time (4) when this expansion reaches its maximum in early summer, the relative strength of the frozen ground to compressive stresses will mean that the stone and its surrounding soil will have been forced out through a radial distance, proportional to the distance h and the temperature range T_max-T_min. Out at the boundary these outward expansions will have added to the accumulating debris and rocks. With outward expansion grounded at these locations one of a number of compression related failures could occur, just as for the case of ice-wedge polygons. The rock ramparts could be caused by shoving actions similar to the processes responsible for the formation of ramparts at the edges of expanding lake ice (Gilbert 1890; Hobbs 1911; Scott 1927). Once full thawing has taken place by the end of the summer the stone will be left at a distance further out from the centre than at the end of the previous summer. Over a period of years this stone and others will find their way to the edge of the circle, polygon, etc. Near the surface there is no reason why this cyclic thermal action could not be taking place on shorter time-scales, having even circadian periodicities.

 

 

The same mechanism could also explain the radial motions of stones within the soil, although the rate of outward motion would be less on account of the crack widths and the temperature increases giving rise to compression related straining both being attenuated with depth. It seems possible that this process could be responsible for the surface fissures, porous nature of the surface soil and the gaps around stones noted at say Figure 4.20 (Washburn 1979). In certain circumstances it is likely that summer surface water or wind action could see the accumulated soil at the outer edges being eroded back to fill the fissures and pores left by the melting of ice from the previous winter’s cracking action. This additional form of mass movement would accentuate the process of soil sorting whereby the centres of the circle are gradually colonised by finer particles and the outer edges by those of larger dimension.     

     All of this seems plausible but does not entirely explain the mechanics whereby the stone circles are formed in the first instance and what it is that determines for a particular set of circumstances the characteristic radii for the stone circles, polygons etc. It is possible that these characteristic dimensions are determined by the nature of the compression failures in the frozen soil at what will eventually become the outer circumferential boundaries of the stone circle. The compression related failure at the edge of the stone circle will depend upon the seasonal maximum thickness of the active layer. Were this compression failure to take the form of an overall uplift buckling, similar to that discussed in the next section, then surface stone movements could be the result of the sloping sides of the uplift deformation. It is also possible that the characteristic dimensions could be the result of tension crack patterns similar to those that determine the patterns exhibited by ice-wedge polygons. Whatever is the initiating cause for the defining geometric properties of the circles, polygons, etc, once started the ratchet depending upon the relative vulnerability of frozen soil to tension stress compared with its resilience to compression, could possibly provide an important ongoing motivating force. In arid regions wind blown fine particles might take the place of ice in filling the cracks caused by restrained thermal shrinkage. 

It is interesting to speculate as to why in seemingly more restrictive circumstances the larger stones accumulate not at the outer boundaries but at the centres of the graded stone circles, polygons, etc? It seems plausible, as suggested by Washburn (Washburn 1979), that these stone accumulations may be an advanced form of the stone polygon. As suggested in Figure 7 the corner nodes of the polygonal array may eventually attract through a process similar to that described above, the accumulations previously occurring at the edges of the polygon. These nodes would then become the centres of arrays of essentially high centred triangles; the triangular array representing the congugate of the original polygonal array. High centred sorted or unsorted circles may either be the result of collapse of the underlying ice-wedges at the boundaries or in active permafrost be the result of a compression failure at the centre of the circle. Such failures might involve a process of upheaval buckling similar to that described briefly in the following section in relation to the development of hummocks, frost mounds, palsas and pingos.