1Taken as a whole, almost any discussion regarding weathering processes within cold regions, which upholds the predominance of frost action as the weathering mechanism seems to satisfy the philosophical paradox of ‘Self-fulfilling Belief’ (Clark, 2002, p. 182) : “If I believe that I am believing this, where ‘this’ refers to that very belief, my belief necessarily makes itself true.” ; the second, key, part of the paradox is, though, judiciously avoided for it states “But I cannot hold this as a belief, since it has no genuine content.” As M. Clark (2002, p. 182) discusses in respect to the preceding quote, “Some beliefs are self-fulfilling - true because they are believed” and this, I would argue, is very much the case with respect to the assumed pre-eminence of freeze-thaw as the weathering agent in cold environments. For example, and these are but a couple among many, C.K. Ballantyne (2002) in his exceptional discussion regarding paraglacial environments refers almost exclusively to the role of freeze-thaw weathering. Indeed, C.K. Ballantyne’s figure 7 (2002, p. 1945) uses freeze-thaw as part of the foundation to explain rock wall retreat. Interestingly, C.K. Ballantyne does introduce a scale element in that same Figure where he refers to the role of (non-process specified) granular disintegration. This same theme recurs in A.M. Curry and C.J. Morris (2004) where once more rockwall retreat is associated primarily with freeze-thaw action but with some (p. 96) recourse to the formation of “...fines by flaking and granular disintegration..” Again, this granular weathering is frequently considered to take place by non-specified processes, although the Abstract refers to ‘microgelivation’ and this term is specified on page 96 where it states that the “...detachment of finer grains produced by microgelivation (weathering) of intact rock through splitting, flaking, granular disintegration and disaggregation”. Scale and process, apart from the implication of microgelivation, are overtly expressed where it is stated (Curry and Morris, 2004, p. 86) that the “...efficacy of frost weathering in dislodging blocks of rocks from the cliffs...” is a key element of talus slope production : in other words, frost weathering produces ‘blocks’. The corollary also exists whereby the implication is that gelivation (freeze-thaw), for both grains (microgelivation) and blocks (macrogelivation), is the only causative processes of all the debris. In a sense, these two papers (from amongst many) encapsulate the issues here, namely the assumption of freeze-thaw and the interaction of process and scale.
2Most discussions that uphold this supremacy of frost action frequently appear to be unconstrained by any attribute of scale. At best, there is referral to “microgelivation” and “macrogelivation” (e.g. Matsuoka, 2001), terms first suggested by J. Tricart (1970, p. 76) : “Macrogélivation exploits existing weaknesses in the rock ; microgélivation cuts into solid rock.” N. Matsuoka (2001, p. 300) provides a definition of these two terms to which, based on the descriptions (see below) of J. Tricart (1970), he also gives scale attributes : “...microgelivation, which involves granular disintegration and small flaking, the process leading to production of fine debris (μm-to-cm scale)” while macrogelivation deals with the “...opening (wedging) of pre-existing macrofractures (joints) that tend to produce pebble-size or coarser material (cm-to-m scale).” J. Tricart (1970, p. 76), in his discussion on these terms, also gives a sense of scale where he states : “Weathered crystals are at once less resistant to pressure and more permeable ; it is these that microgélivation exploits” and “The rate of macrogélivation is a function of the relationship between the spacing of lines of structural weakness and the depth reached by frost”. However, the adoption of these two levels and the application of scale to them assumes a validity which may be false ; all the more so when process is associated with scale. The problem is that process does appear to be linked to scale (Matsuoka, 2001, p. 300) : “The governing mechanism (between micro- and macro- gelivation) may also be different.” This will be the core of the discussion presented here.
3The problem seems to be that these notions of “micro” or “macro”, and any associated scale/process attributes, seem to be accepted and applied as ‘truths’ without condition or debate, and arguments and/or data collection made to fit within this framework. Although the original author of the idea (e.g. Tricart, 1970) may not have made such a claim, it appears that we give validity to these two scales by unquestioning repetition : “The story is one of casual empiricism gathering respectability by repetition until it attained the stature of an article of faith” (Thorn, 1992, p. 10). As R. Inkpen (2005, p. 129) states : “A common assumption is that scale is concerned with an absolute set of spatial and temporal dimensions.” It is this very assumption, with respect to freeze-thaw weathering, that is here being questioned. Thus, the aim here is to try and open the questions of process and scale with respect to micro- and macro- gelivation. Here it is worth stating that the endeavour is not to suggest that freeze-thaw weathering, in any or all guises, does not take place but rather to question the “...ill-founded but oft-cited notion that freeze-thaw weathering is ubiquitous in periglacial regimes (Thorn, 1988, p. 13).
4The two questions addressed here, at least in part, are (1) what is the nature of rock weathering in cold regions, and (2) what are the scale attributes of that weathering ? Rewritten, one might paraphrase the questions as (1) is frost weathering really so active, widespread and dominant, and (2) do we really have only two scales of rock weathering by frost action ?
5Perhaps this question will go against ‘mainstream’ acceptance and assumption within periglacial geomorphology where there is some almost ‘messianic’ quality to the validating of the role of ‘cold’ and its application to freeze-thaw weathering. Most studies describe how cold a site was, for how long it was cold, and/or how frequently it experienced “freeze-thaw cycles”. However, in how many of these descriptions can one find details regarding the high summer temperatures (some absolutely so, some relatively), summer rain (in some areas), and of how ‘pleasant’ the summer actually is during, in most cases, the period when the observer is actually there - as opposed to the vicarious description of the winter ‘cold’ ? This is, of course, all the ‘worse’ when trying to reconstruct former cold climates for which, at best, only proxy data are available - the rest is usually conjecture of extremes. In defense of this argument, consider from among the earliest (e.g. Losinski, 1909) through to recent (e.g. Gruber et al., 2003) quotes that expound this view : “...the temperature changes are less significant than frost action to such a degree that “freezing is not only one but the most important reason of weathering in the majority of cases” (Losinski, 1909, quoted in Evans, 1994, p. 120) to “...an increasing active-layer thickness will subject perennially frozen rock to freeze-thaw cycles and corresponding effects such as joint widening (macrogelivation)” and that “Effective frost weathering (is) by granular disintegration and small flaking (microgelivation)….” (Gruber et al., 2003, p. 325). It is as if, even under the warming climate scenario of S. Gruber et al. (2003), only the winter, and its associated cold, is of significance - the warm summers, and their extension under a warming climate, appear to play no role in weathering. It is not the aim to imply that cold-based processes are not of consequence, perhaps of dominance even, but that the obsession with them has been to the detriment of better understanding the holistic weathering synergies wherein the warmer summer conditions may play a sometimes significant role. Further, it is argued that this preoccupation with the cold-based processes has affected the manner of data collection, and interpretation, such that in many instances the data were inadequate to consider other processes and/or that in the evaluation of the data the effort was placed on the cold component at the expense of even recognizing that there was a warm element.
6 It is suggested that in many of the descriptions regarding cold environment weathering (such as those cited above) it is the general climate rather than the rock micro-environment which is used to exemplify the ‘cold environment’. As has been published (see Hall, 2004) and will be shown here, winter rock temperature maxima, during the day, need not, for sites equator-ward of their respective polar circle, be significantly different to those of the summer. Indeed, as has been argued elsewhere (Hall et al, 2002), surely it is the very availability of ‘heat’, rather than the all-pervading ‘cold’, that is actually the significant geomorphic factor in cold region weathering processes. If we fixate just on the cold, then the situation becomes akin to some of our hyper-cold, hyper-arid inner planets where processes such as freeze-thaw weathering are pre-empted by the inability of ice to thaw into water. Given an absence of solar radiation to heat rock, so even such as thermal stresses to cause rock breakdown may well be inhibited. Thus, the ‘driving force’, given a cold climate, must indeed be the addition of heat ; without the provision of heat there cannot be cooling. Thus, the key here is to consider the rock thermal conditions and their variability through time, and to base the determination of weathering at any given site upon those data. Of course, it is imperative to also consider the moisture conditions experienced by the rock for the thermal conditions alone tell but part of the story. Thus, in order to answer question 1 above, we need to have an unconditional (i.e. not to prove one or other process) monitoring of rock thermal and moisture conditions in a manner that will allow evaluation of any or all weathering processes.
7In respect of question 2, the ‘micro-/macro-‘ concept is a good starting point, but it should not be self-limiting and assumptive that it contains within it all that is required. Indeed, as M.F. Hochella (2002) has recently argued, perhaps many weathering processes, particularly chemical, should be considered at the nano-scale (10-9 to 10-7 m) ; attention is also drawn to the volume by J.F. Banfield and A. Navrotsky (2001) on “Nanoparticles and the Environment” where similar arguments are made. This nanoscale becomes quite a challenge when we consider that in mechanical processes of rock weathering we have not even really considered weathering at the micro-scale (10–5 to 10–7 m), indeed we have barely worked in the “meso”-scale of 10-2 to 10-5 m which would include sand-sized material. Interestingly, M.F. Hochella (2002, p. 739) sees this nano-scale as particularly important in ‘low temperature environments’ and he goes on to state : “...all aspects of weathering, soil and water/rock interaction science are inexorably linked to nanoscience. Within the Earth’s near-surface, materials, that are broken down… are often in the nanoscale regime.” The aim here is not to advocate studies at this level, although clearly this is a particularly significant scale for water/rock/bacteria interactions, but rather to use it as a medium to show the relevance of scale and its application, or not, to any meaningful study of weathering and the use of terms that constrain these boundaries (here the use of the terms micro- and macro-gelivation).
8The question arises, if the ‘microgelivation’ and ‘macrogelivation’ distinction is based on the latter “exploiting existing weaknesses in the rock” while the former “cuts into solid rock”, then what, exactly, comprises a “weakness” that is not a component of a “solid rock” ? Indeed, is such a dichotomy really possible ? Further, from a geomorphological perspective, is there such a thing as ‘solid rock’ other than as a colloquialism ? If these questions cannot be answered unequivocally in the affirmative then it is no more than a perception of scale. Even if the division is accepted, then the question still arises as to the scale boundaries for, in reality, grains (from sand to larger crystals within a polymineralic matrix) fall not within the “micro” scale but really within a “meso” scale (10-2 to 10-5 m) - a category between micro and macro. That said, if ‘meso’ is considered as scale break, does it imply yet another process connotation ? Really, at the end of the day, are these all truly ‘fuzzy’ concepts that perhaps we adopt as they provide ‘intellectual comfort’ and are so easily embraced because they appear to offer a framework of rigour, but which are, in actuality, just illusionary ?
9Consider any rock - what comprises the ‘existing weaknesses’ ? It is clear from both J. Tricart (1970) and N. Matsuoka (2001) that these are interpreted as being the major bedding and/or jointing within that rock mass. This must be so, otherwise we would not see the use of the term ‘blocks’ or the scale of ‘cm-to-m’ ; indeed, this is also the very usage and scale found in the other cited papers (Ballantyne, 2002 ; Curry and Morris, 2004). Thus, it is clear that ‘weaknesses’ here are only of a certain scale. But what of microcracks (in the range of 10-5 to 10-6 m) that would certainly preclude a rock as being “solid” or “intact” ? These cracks are described as being of three types (Bland and Rolls, 1998, p. 49) : (1) multigrain (crossing several crystals), (2) single-grain (involving a single mineral), and (3) stress fractures (larger than (1) or (2) and less influenced by mineralogy). This question as to the importance of microcracks may be highly significant for F. Schwarzl and A.J. Staverman (1956, quoted in Butenuth, 2001, p. 244) state “If the distribution of microstructure is known with respect to their strength then, on the basis of this and with the help of statistical failure theory, macroscopical fracture properties can be deduced also.” Thus, there is not really a scale ‘break’ but rather a ‘continuum’ between the levels, with each affecting the others. Indeed, as C. Butenuth (2001, p. 130) discusses, once more quoting F. Schwarzl and A.J. Staverman (1956), “A real understanding of breakage phenomena cannot be gained by macroscopic observations alone.” Perhaps the heart of the issue is our perception of rock where engineers, and, it is here argued, geomorphologists “...assume rock as a homogeneous and isotropic medium. However, most rocks are not sound ; hence they are neither homogeneous nor isotropic ”(Jumikis, 1983, p. 38). It is all a perception of scale. This perception is evident within studies of rock mechanics where in its natural environment “...rock is principally characterized by the fact that rock is not a continuum but a regulated discontinuum because of... discontinuities”(Jumikis, 1983, p. 74). That ‘regulated discontinuum’ is a product of the range, and scale, of discontinuities within any given rock, from (at least) nanoscale to macroscale, that impact, in some manner, the strength of that rock. Those discontinuities can operate discretely or synergistically. To bring the arguments full circle, it is apparent that rock properties can be grouped into two large groups, ‘macroscale’ and ‘microscale’ (Jumikis, 1983, p. 265) but that the “Macroscale rock mass properties are essentially properties of the component” while the microscale are “...properties of the substance.” ; where ‘substance’ refers to the material of which something is made. Hence, the macroscale properties are a product of the sum of the attributes of the substances of which it is made. In other words, as discussed above by C. Butenuth (2001), the ‘component’ is influenced by the ‘substance’. Thus, one is still left with the questions – is there any such thing as ‘solid rock’ and are we measuring appropriately for any meaningful determination of rock weathering as a function of scale ?
10By recourse to ‘micro’/ ’macro’ attributes and attempts to monitor and apply process within that framework, surely we negate the reality that all scales may be operative simultaneously. For example, the annual freeze penetration that goes to, say 3 m, also has, at some point in time, a surface effect - and the overall cooling within the rock may have, superimposed upon it, small surficial fluctuations - i.e. as the seasonal freeze occurs so there are still diurnal fluctuations occurring in the outer shell of the rock, and they all operate both independently and synergistically as a function of the spatial and temporal scales applied. Indeed, perhaps the small, superficial warming fluctuations on a rock undergoing a larger-scale freeze may promote even greater stresses in the outer shell of the rock. Further, even small-scale diurnal events must accrue some degree of strain to the rock beyond their zone of direct impact. There must be a measure of strain relief in the outer shell, resulting from surficial weathering (flaking), that impacts on the rock immediately below and this is further exacerbated by material loss at the surface resulting from the weathering ; the scale here may well be that of the nano-scale, but operative just the same. Thus, the application of either ‘micro’ and/or ‘macro’ attributes to monitoring and/or process excludes the reality that these are but levels within a continuum.
11As a further consideration regarding scale, one is drawn to the arguments of C.E. Thorn (2003, p. 413) where, in a discussion on space, he states “Separation of time and space is a heuristic necessity in geomorphology : in reality time infects space and space infects time and complete separation of the two is impossible.” He goes on to say (p. 414) “Consequently the absence of a form is predicated upon the absence of only one, but perhaps more, necessary factor(s), while its presence is predicated upon the presence of all necessary factors. This makes the significance of the presence and absence of individual landforms asymmetrical” ; for ‘landform’ one can read ‘character/type of weathering’. Thus, with recourse to the dichotomy of scale in weathering (macro/micro), one is also left with the temporal attribute of the measurements - where do they fit within the temporal framework and thus what do they tell us ? In other words, just as in deducing the totality (in time) of landform evolution from observations taken today ; in taking measurements at point x in time can we put those data in the context of what preceded and thus perhaps what actually precipitated the very conditions we see at point x ? A classic example of this might be that of blockfields (‘felsenmeer’), features commonly associated with frost action (see French, 1996), but which recent studies (e.g. Boelhouwers and Sumner, 2003 ; Whalley et al., 2004) have shown to be the product of (at least in part) pre-Quaternary weathering processes limited in terms of frost action but with substantial chemical weathering to produce the associated clays. That the blockfields are observed today in cold environments, and may be subject to some degree of frost action at this time, has led to the possible misconception of the requirement for cold-based processes in their origin and formation.
12Thus, to see a certain form of weathering occurring now (e.g. flaking, granular disintegration, block formation, etc.) does not, of necessity, imply either continuity of type of weathering or of process beyond an observation as to what is currently happening. Equally, when we see large blocks being produced by weathering (say, as a result of annual freeze events) does this imply that no granular weathering has occurred or that in our preoccupation with the former, and larger, we fail to consider the latter (and smaller) ? Indeed, in some ways this is the very argument recently put forward by A.M. Curry and C.J. Morris (2004, p. 103) where they suggest that there is “... a need for reassessing small-scale weathering in temperate and cold environments...” as, at this time, this may “...represent the major mode of current rockwall retreat...” Hidden within this argument is a practical issue - it is clearly ‘easier’ to monitor and measure loss of large blocks than it is the fine material. Not only are the large blocks much more obvious, but they suffer less from complications of calculating material weathered free from a rock face versus fine material developed in situ by on-going weathering of previously released larger material. A practical scale question for the production of blocks, from exploitation of bedding and/or jointing, must surely be the scale of the initial weathering. Joints and bedding planes are initially at the micro-scale and become enlarged by weathering. Thus, despite the end product being a large block, the initial weathering is micro-scale in character - it cannot be otherwise as the initial size of the joints prohibits any other scale level. By the time a block is identified as ‘broken free’ - possibly by macro-gelivation - so this is but the exploitation of the now enlarged joint. Indeed, as P.D. Sumner (pers. comm., 2005) has suggested, the movement of a large block by frost wedging may not actually be weathering but rather simply ‘transport’ - as the block was no longer bonded along its boundaries as a result of the preceding micro-scale weathering. Thus, true frost wedging and macro-scale gelivation may be quite rare, block production being, in reality, the product of micro-scale processes.
13It can be argued (Hall, 2003) that perceptions of weathering are largely a function of our data. If, at the turn of the last century, we had measured thermal conditions at one-minute intervals in and on rocks then it is very likely that our perception of the weathering regime(s) for cold environments may have been very different to that which we have today, based largely as it is on low-frequency thermal monitoring (often at 6-hour or more intervals). For example, consideration of rock temperature data (from northern Canada) for one day at 1-minute intervals as compared to that, for the same day, taken at 6-hour intervals shows just how our perspective would change (fig. 1). It can be seen that, for that day, the six-hour data show a ‘classic’ freeze-thaw cycle with thermal transition from -4 °C to -3.5 °C then up to +2.5 °C and finally back to -4.5 °C ; a freeze-thaw cycle exactly as generally presented in most texts and exemplified by much data. However, the one-minute data present quite a different picture, and the question must be asked : “If our original data had been at this frequency, would our perceptions of freeze-thaw weathering have been the same as those generated from low-frequency data ?”. During the same day for which the six-hour data present a ‘classic’ freeze-thaw cycle, the one-minute data indicate 16 crossings of the zero isotherm ! Further, the more detailed data also show a higher (by 8.5 °C) daytime temperature and a lower (by 2 °C) freeze temperature. Quite a different picture with quite different weathering implications. Although the general pattern of the six-hour data is replicated, notably the night-time freeze from 1800 hrs through to 0800 hrs and a day-time warming from1000 hrs through to 1700 hrs, there is a significant variability of thermal conditions within both periods. Most notable, apart from the substantially warmer day-time high, is that there are so many crossings of the zero degrees isotherm ; mostly during the day-time ‘warm’ period. No significance is given to the zero degrees threshold other than to exemplify the variability ; any other threshold would give the same sort of frequency difference (some such as -2 °C might even give a higher number of crossings as well as wider temporal distribution during the 24 hours). Thus, by this change of temporal scale a very different picture of the thermal environment the rock is subject to emerges. There is also a spatial component insofar as the depth of penetration of such short-term thermal changes is highly limited, and this brings weathering back to consideration of the micro- and nano-scale.
Fig. 1 – One-minute rock temperature data for a 24-hour period, Prince George, Northern Canada.
Fig. 1 – Données thermiques avec une fréquence d’une minute pour une période de 24 h, Prince George, Nord canadien.
The thick line shows one-minute rock temperature data for a 24-hour period, while straight line indicates what would have be recorded with a six-hour measuring interval. In terms of crossing of thresholds : using 0 °C for simplicity, the six hour data show two (2) crossings while the high-frequency data indicate at least 16. Use of any threshold would show comparable degrees of variability between the two data sets
Le trait gras montre les enregistrements à fréquence d’une minute sur une période de 24 heures, tandis que les segments linéaires indiquent ce qui aurait été enregistré avec un intervalle de 6 heures. En terme de franchissement de seuils et en utilisant 0 °C par mesure de simplicité, les données à 6 heures révèlent deux franchissements tandis que les données à haute fréquence en montrent au moins seize. L’utilisation de n’importe quel autre seuil traduit des degrés de variabilités comparables entre les deux séries de données.
14In detail, the higher frequency data also indicate large ΔT/Δt changes, in the range of 5 to 11 °C min-1 ; values well within the scale that has been suggested as able to cause thermal shock (e.g. Richter and Simmons, 1974 ; Yatsu, 1988) and certainly to cause thermal fatigue due to the multiple replications. The scale of impact for these ΔT/Δt events is that of the micro- / nano- and certainly not at the macro-scale, but it may play a significant role in granular disintegration or flaking - the very levels that ‘micro-gelifraction’ is said to be operating. The significance is that, ΔT/Δt events are not thermally constrained to any particular location on the temperature scale, and certainly do not require sub-zero temperatures, although falling temperatures are (roughly) twice as effective as stress producers (Marovelli et al., 1966) as are rising temperatures. Further, many of these rate of change of temperature events are bi-directional (rising and falling temperatures within the one-minute) and thus may be particularly important for causing thermal stress at the grain scale. So, in terms of scale, these are thermal fluctuations that, both temporally and spatially, exert an impact only at the micro-scale.
Fig. 2 - Exotherms on three successive days.
Fig. 2 – Exothermes sur trois jours consécutifs.
The transient rise in temperature as latent heat is released during the freezing of water. Note that the temperature at which freeze is initiated varies for each event. Thick, horizontal line shows 0 °C. Measurements taken at experimental site in Prince George, northern British Columbia (Canada).
Le pic soudain de températures est due à la libération de chaleur latente au moment du gel de l’eau. Remarquez que la température à laquelle le gel se produit varie pour chaque événement. La ligne horizontale épaisse indique le 0 °C. Mesures effectuées dans la station de Prince George au nord de la Colombie britannique (Canada).
15The high-frequency data also provide important evidence of two other weathering attributes. First, they are able to resolve the passage of latent heat released as interstitial water freezes (Hall, 2004), the ‘exotherm’ (fig. 2) ; this giving validation of freeze-thaw events. This is an extremely important observation, not least because it leads to the second important weathering attribute - the evidence for possible chemical weathering for a large part, if the not the whole, winter period (Hall, 2004). The monitoring of the exotherm is significant in that it clearly demonstrates that freezing of water did actually occur. In the absence of moisture data (by such techniques as used by O. Sass, 2004, 2005), the exotherm provides unequivocal evidence that water was actually there and that, within the temperatures experienced, the water actually froze ; it does not, however, evidence that any weathering actually occurred. Within this discussion of scale, the key is that, from the available data (Hall, 1999, 2004 ; Hall and André, 2001, 2003), it (the exotherm) is showing the occurrence of freeze events within the outer few centimetres of the rock and that these may occur (in this spatial zone) for a greater part of the winter. Again, in respect to weathering, the scale of operation is pointing to that of granular disintegration and/or flaking. Also in terms of scale, the temperature of the exotherm start (fig. 3) provides information regarding the temperature at which the water begins to freeze (Hall, in press) and this has shown that the oft-cited (e.g. Matsuoka, 2001) thresholds may not always work and thus simple ‘counting’ based on crossings of these thresholds may not be viable irrespective of whether water is present or not (an attribute for which no data are available when counting threshold crossings).
Fig. 3 - Exotherms on four successive days with freeze initiation at different temperatures.
Fig. 3 – Exothermes sur quatre jours consécutifs avec le déclenchement du gel à différentes températures.
Note high daytime rock temperatures (17 to 21 °C). Measurements taken at experimental site in Prince George, northern British Columbia (Canada).
Remarquez les températures diurnes élevées (17 à 21 °C). Mesures effectuées dans la station de Prince George, nord de la Colombie britannique (Canada).
16Finally, given that the occurrence of an exotherm must demonstrate that water was present (to freeze) then consideration of the rock temperature for the preceding and/or succeeding day (fig. 4) can indicate the possibility for chemical weathering (Hall, 2004). This is quite a significant observation at several levels. At the fundamental level, the importance is in respect of the (spatially constrained – see Hall, 2004) potential for chemical weathering to take place throughout the winter in spite of sub-zero air temperatures. With respect to weathering, it is often the average annual air temperature that is used (Ollier, 1984) and this has indicated that the winter is so cold that chemical weathering cannot occur, not the least because any available water would be frozen. However, data such as those presented here indicate that rock surface temperatures may go almost as high, during the day on the sun-facing aspect, as is recorded in summer and that the presence of unfrozen water is indicated by the subsequent nighttime freeze. Thus, chemical weathering could be quite active during such days. Returning to scale, the significance of this chemical weathering is that the potential resides within the outer shell of the rock rather than in the rock as a whole - the micro-scale rather than the macro-scale. Thus, even in extremely cold environments so long as there is a solar radiation input (i.e. the effectiveness will decrease temporally with increase in latitude) then there can be chemical weathering in the outer shell (<1 mm to c. 5 cm) of rock despite the greater part of that rock remaining frozen : micro- and nano- scale chemical weathering.
Fig. 4 – High daytime temperatures on days with water present within the rock.
Fig. 4 – Températures diurnes élevées avec présence d’eau dans la roche.
Presence of water within the rock is indicated by the subsequent nightime freeze (exotherms) which offers the possibility of chemical weathering in daytime (positive temperatures 5 to 26 °C) in the middle of winter (November) at this latitude. Thick, horizontal line shows 0 °C. Measurements taken at experimental site in Prince George, northern British Columbia (Canada).
La présence de l’eau dans la roche est indiquée par le gel nocturne (exothermes), qui offre la possibilité d’une météorisation chimique durant le jour (températures positives, de 5 à 26 °C), même au milieu de l’hiver (novembre) à cette latitude. La ligne horizontale épaisse indique le 0 °C. Mesures effectuées dans la station de Prince George au nord de la Colombie britannique (Canada).
17Thus, with the recent increasing recognition of the possible significance of role of micro-scale weathering in cold regions (Curry and Morris, 2004) coupled with granular level data (Hall and André, 2003) and the beginning of nano-scale weathering studies (Hochella, 2002) so there needs to be a reconsideration of scale in cold region weathering studies and landform development. Not the least there needs to be a reappraisal of the micro-/macro- levels and quite what these levels mean, whether they are the only levels (clearly not), how the various levels are connected, how the different scales operate both discretely and synergistically, and how these attributes vary temporally and spatially. In some ways this approach generates a paradigm shift (indeed, is there yet a paradigm in cold region weathering ?) insofar as it looks more at small-scale weathering, rather than the more often cited macro-scale attributes ; there is also an extension of chemical weathering potential (for the outer skin of rock) to the winter period. Thus, there is the need to reconsider weathering in cold regions in terms of scale, as well as the use of sensors (size, response time and resolution) and record intervals appropriate to the scale(s) of weathering under consideration. The influence of our data on our perceptions of weathering (Hall, 2003) is such that the question need be repeated, that “had our original data been of high-frequency would not our somewhat simplistic visions of cold region weathering been radically different ?”. If the answer is in the affirmative, then there is a need for better data and a reconsideration of theory, and perhaps reinterpretation, of weathering. If the answer is negative, then the scale attributes briefly discussed here appear to have been recognised by other disciplines and so geomorphologists are left with the question as to why they consider them not apply - and a clear answer needs to be articulated.
18Apropos the arguments presented here, R. Inkpen (2005, p. 23) identifies the problem of scale as a central issue in terms of advancing in our understanding : “The perennial problem of scale remained as a sticking point for the integration of processes found at different scales and acted as a brake on a purely reductionist view of the scientific endeavour in physical geography.” As R. Inkpen continues (2004, p. 130) “The scale of operation of a variable is defined by its relation to other variables, not by a fixed framework.” It is these very issues that have been questioned here with respect to our use of “micro-gelivation” and “macro-gelivation” : the assumed discreteness and independence of these scales as well as the associated process connotations. From a personal viewpoint, it seems an extremely simplistic and naive view of weathering, at odds with the inherent complexities of both the processes and their obvious linkages. In many ways, this situation is a function of our data but we seem not to have moved forward with the advances in technology that offers new data collection opportunities and hence new insights - or validations of older assumptions. Appropriate data at the macro-scale are not as available as may first be thought, while data at the micro-scale are rare (and at the nano-scale even rarer) and yet “An appropriate representation at one scale, derived from a specific method, will not form an appropriate representation at another scale.” (Inkpen, 2005, p. 131). Thus, the call is made for better and more appropriate (to the scale under consideration) data acquisition and then, and only then, a consideration of the spatial and temporal attributes of weathering. In defense of this approach, the comment by D. McCarroll (1997, p. 1) would seem somewhat germane : “We do not improve our theories or models by admiring them, or by proclaiming how well they seem to fit our observations. The only way to improve them, and therefore to make progress, is actively to seek conflict between our models and the real world.”