Open system, or more correctly, hydraulic pingos, are genetically poorly understood. A continuing problem concerns their need for a perennial groundwater supply (intra- or sub-permafrost). This has to be maintained despite the existence of continuous permafrost in many areas where they are located. Recent work on Disko Island has suggested a new type of hydraulic pingo developing only in a "marsh environment". It is argued that the marsh setting is not relevant to the formation of these features and that they are simply hydraulic pingos.

Keywords

Hydraulic pingos, permafrost, Greenland, Disko.

Stephen D Gurney: Department of Geography, The University of Reading, PO Box 227, Whiteknights, RG6 6AB. United Kingdom. Peter Worsley: Postgraduate Research Institute for Sedimentology, The University of Reading.

This discussion intends to highlight problems with the hypothesis proposed by Christiansen (1995) concerning the pingos of Mellemfjord. Christiansen's work appears to be part of a larger ongoing study as outlined by Humlum et al., (1995), therefore, discussion at this stage may prove particularly useful. It should be noted that the pingos in this region were first discussed by Donner (1978).

Pingos are one of the few periglacial landforms which are true indicators of permafrost conditions and take their name from the Inuit for conical hill which was first introduced by Porsild (1938). The current global classification scheme recognises two types of pingo: the hydrostatic pingo (closed system) and the hydraulic pingo (open system). Fritz Müller (1959) originally proposed this genetically based classification and Mackay (1979) extended its use. Indeed Mackay's long term monitoring work concerning the hydrostatic pingos of the Mackenzie Delta and Tuktoyaktuk peninsula area, N.W.T., Canada, has done most to characterise pingo processes (e.g. Mackay, 1973; 1977; 1987; 1994). Despite this large body of work, the understanding of hydraulic pingo genesis has advanced little since the early 1980's with important contributions from Liestöl (1977) and Hamilton & Obi (1982). The basic processes involved with hydraulic pingos involve the generation and movement of groundwater under artesian pressure which forms a perennial sub-surface supply to the pingo ice-cores. The type-area for hydraulic pingos is east Greenland particularly northern Jameson Land and the Karup Valley area of Traill Island (e.g. Müller, 1959; Cruikshank & Colhoun, 1965; Lasca, 1969; O'Brien, 1971; Allen et al., 1976; Worsley & Gurney, 1996). The other areas where hydraulic pingos occur include Alaska and Spitsbergen (e.g. Liestöl, 1977; Hamilton & Obi, 1982).

The region of central Alaska in question represents an area of discontinuous permafrost and is therefore one in which recharge and movement of groundwater does not represent conceptual difficulties. Spitsbergen and east Greenland, on the other hand, are areas of continuous permafrost and hence present problems for the accepted genetic hypothesis. In order for a continual supply of groundwater to be achieved over considerable periods (and hydraulic pingos require considerable periods in which to grow) zones of groundwater recharge and groundwater pathways must be maintained. It may be argued that the groundwater is sub-permafrost in nature and this avoids the necessity to maintain conduits within the permafrost, yet the question remains as to how the groundwater supply is replenished at altitude since this is required to generate the artesian pressures required. In the context of Spitsbergen, recharge of groundwater beneath glaciers has been suggested by Liestöl (1977), however, at many of these localities glaciers are either not present or are small in size. That is, warm-based glaciers with sufficient thickness to introduce a reasonable volume of water into the subsurface. Traill Island which has classic examples of hydraulic pingos forms an extremely useful case study of these problems (see Worsley and Gurney, 1996).

Discussion

Turning to the genesis proposed by Christiansen (1995) for
the Mellemfjord pingos, a continuous permafrost setting
and a typical valley bottom location is described for what

are termed "open system pingos", the genesis described, however, is more closely related to that of hydrostatic (closed system) pingos. That is, it involves surface frost penetration of newly exposed sediments following marine regression. This freezing eliminates any taliks and expels pore water from their sediments which, in turn, supplies the pingo ice-cores. This is not the genesis of hydraulic pingos. Either these features are the result of the elimination of taliks formerly sustained by marine waters (following aggradation of the area), and represent a type of hydrostatic system, or they are related to an hydraulic (artesian) groundwater supply and are therefore hydraulic pingos. Whichever is the case, Christiansen appears to have incorrectly classified these features.

It is here postulated that these features represent true hydraulic pingos sustained by perennial intra- or subpermafrost groundwater as are the pingos of east Greenland. Their location at the valley mouth is merely a confusing factor and is not significant in their origin. Hydraulic pingos could easily progress through a growth and decay cycle within the 2,700 years which Christiansen states is the period that this lowermost section of the valley has been exposed following marine regression. The concentration of pingos in this area does not represent a problem either, since on Traill Island, east Greenland complexes of hydraulic pingos numbering up to 10 have been documented (e.g. the Zurich group: Worsley & Gurney, 1996).

Of obvious importance, to the comprehension of pingos in Mellemfjord is the work of Humlum et al., (1995) which documents landscape evolution in the Mellemfjord area. This work includes a geomorphic map of the valley in question (Figure 3, page 31) which locates the 5 pingos documented by Christiansen (1995) as well as another pingo, 5 km up valley from the mouth by the sea. This feature appears to be located just above the valley floor. Whilst the pingos at the valley mouth are discussed in Humlum et al., (1995) the pingo further from the coast is not, probably because it is not interpreted as belonging to the "marsh initiated group". This solitary pingo is important since it is likely to represent an hydraulic pingo and provides evidence for groundwater upwelling at a locality which cannot be related to the elimination of taliks following marine regression. Furthermore, Christiansen's map (1995, Figure 2, page 43) shows an icing (naledi) which again provides independent evidence of perennial groundwater upwelling unrelated to the marine regression. Thus, if groundwater systems are operative throughout the area their absence at the valley mouth requires explanation.

It is salient to note that other workers have attempted to subdivide hydraulic pingos, for example, Yoshikawa and Harada (1995). It is, perhaps, not surprising since hydraulic pingos are so poorly understood that meaningful generalities are difficult to substantiate in all cases. This accounts for why particular groups are often credited as forming a separate species of the general genus. Nevertheless if this is to be attempted it must be made certain that the genesis proposed conforms firstly to true pingo genesis as follows:

(i) The mounds are true pingos in that they are perennial, exist within a permafrost environment and rely on a pressure system (as opposed to cryosuction) to deliver water to their core where it is frozen;

(ii) The mounds belong to one of the two sub-groups i.e.
the origin of the pressure system is either hydraulic
(artesian) or hydrostatic.

Where the above criteria have been proved to determine hydraulic pingos, only then can a sub-division of this group be made, however, any such sub-division must conform to the criteria set for hydraulic pingos. It is, of course, possible that some pingos may be identified which cannot be readily assigned to either hydraulic or hydrostatic groups due to the paucity of data regarding the groundwater movement within permafrost beneath the features. Here the features may be termed mixed or polygenetic in origin (see Gurney and Worsley, in press).

Conclusions

It would appear that whilst the work of Christiansen (1995) and Humlum et al., (1995) documents the existence of at least 5 more pingos to add to the growing list for west Greenland (e.g. Yoshikawa, 1991) no good evidence for a new type of hydraulic species have been presented. Whilst the location of the majority of these features adjacent to the valley mouth does much to constrain the maximum age of these features it does not indicate that these features are unique to this environment. If they are, then their processes of formation owe more to hydrostatic pingo genesis than to hydraulic pingo genesis because they involve the elimination of taliks which progressively forces water out of the sediments and into the pingo core. Therefore, it is

concluded that these features merely represent hydraulic
pingos in an unusual setting.

References

Allen, C.R., O'Brien, R.M.G. & Sheppard, S.M.F. (1976): The chemical and isotopic characteristics of some northeast Greenland surface and pingo waters. Arctic and Alpine Research, 8: 297-317.

Christiansen, H.H. (1995): Observations of open system pingos
in a marsh environment, Mellemfjord, Disko, central west
Greenland. Geografisk Tidsskrift, 95: 42-48.

Cruikshank, J.G. & Colhoun, E.A. (1965): Observations on
pingos and other landforms in Schuchertdal, northeast
Greenland. Geografiska Annaler, 47A: 224-236.

Donner, J. (1978): Holocene history of the west coast of Disko,
central west Greenland. Geografiska Annaler, 60A: 63-72.

Gurney, S.D. & Worsley, P. (in press): Genetically complex and
morphologically diverse pingos of the Fish Lake area, south
west Banks Island, NWT, Canada. Geografiska Annaler.

Hamilton, T.D. & Obi, CM. (1982): Pingos in the Brooks
Range, northern Alaska, USA. Arctic and Alpine Research,
14: 13-20.

Humlum, 0., Jacobsen, 8.H., Christiansen, H.H., Rasch, M., Nielsen, N., Hansen, B.U. & Hasholt, B. (1995): Holocene landscape evolution in the Mellemfjord area, Disko Island, central west Greenland: Preliminary results. Geografisk Tidsskrift, 95:28-41.

Lasca, N.P. (1969): The surficial geology Skeldal, Mesters Vig,
Northeast Greenland. Meddelelser om Grønland, 176: 3,
56pp.

Liestöl, O. (1977): Pingos, springsandsand permafrost in
Spitsbergen. Norsk Polarinstitutt Årbok, 1975: 7-29.

Mackay, J.R.. (1973): The growth of pingos, western arctic coast
Canada. Canadian Journal of Earth Sciences, 10: 979-1004.

Mackay, J.R. (1977): Pulsating pingos, Tuktoyaktuk Peninsula,
Northwest Territories. Canadian Journal of Earth Sciences,
14: 209-222.

Mackay, J.R. (1979): Pingos of the Tuktoyaktuk Peninsula area,
Northwest Territories. GÜographie physique et Quaternaire,
33: 3-61.

Mackay, J.R. (1987): Some mechanical aspects of pingo growth
and failure, western arctic coast Canada. Canadian Journal of
Earth Sciences, 24: 1108-1119.

Mackay, J.R. (1994): Pingos and pingo ice of the western arctic
coast, Canada. Terra, 106:1-

Müller, F. (1959): Beobachtungen über pingos. Meddelelser om
Grønland, 153: 3, 127pp.

O'Brien, R. (1971): Observations on pingos and permafrost
hydrology in Schuchertdal, north east Greenland. Meddelelser
om Grønland, 195: 20pp.

Porsild, A.E. (1938): Earth mounds in unglaciated arctic, north
western America. Geographical Review, 28: 46-58.

Worsley, P. & Gurney, S.D. (1996): Geomorphological and hydrogeological significance of the Holocene pingos in the Karup Valley area, Traill Island, northern east Greenland. Journal of Quaternary Science, 11: 249-262.

Yoshikawa, K. (1991): Age of growth of two pingos, Sarqaq
Dalen, west central Greenland. Permafrost and periglacial
processes, 2: 245-252.

Yoshikawa, K. & Harada, K. (1995): Observations on nearshore
pingo growth, Adventdalen, Spitsbergen. Permafrost and
periglacial processes, 6: 361-372.