Burial of glacier ice by deltaic deposition, Bylot Island, Arctic Canada

Size: px

Start display at page:

Download "Burial of glacier ice by deltaic deposition, Bylot Island, Arctic Canada"

Calvin Garrett
5 years ago
Views:

1 Permafrost, Phillips, Springman & Arenson (eds) 2003 Swets & Zeitlinger, Lisse, ISBN Burial of glacier ice by deltaic deposition, Bylot Island, Arctic Canada B.J. Moorman Earth Sciences Program, University of Calgary, Calgary, Alberta, Canada F.A. Michel Department of Earth Sciences, Carleton University, Ottawa, Ontario, Canada ABSTRACT: Burial of glacier ice and a lateral moraine by deltaic deposition is documented for a small lake that is periodically dammed against the side of a glacier on southern Bylot Island, Arctic Canada. Two very different streams flow into the lake. One of the streams carries a very heavy load of sand resulting in the development of a multi-level deltaic complex that extended over and buried the lateral moraine and adjacent glacier ice. Radiocarbon dating of an organic-rich horizon within one of the youngest deltas indicates that this ongoing process has preserved ice for likely hundreds of years. The ice-poor nature of the sand blanketing the ice enhances the potential for long-term preservation, except where the area is undercut by fluvial erosion. 1 INTRODUCTION Massive ice takes several forms; pingos, ice wedges, ice lenses, and tabular massive ice. The development of many types of subsurface ice is fairly well understood; however, the origin and formational processes of tabular massive ice bodies are still matters of some debate. Tabular massive ice bodies occur throughout the Canadian Arctic and are distinctive in that they tend to have much larger lateral dimensions than thickness. Some tabular massive ice bodies have been interpreted to be of segregated origin (Mackay 1971, Moorman et al. 1998), while others are thought to be buried glacier ice (Michel 1985, Fujino 1986, Vaikmae et al. 1993, Robinson et al. 1992). Tabular massive ice bodies can have complex histories and some bodies may contain ice representing more than one formational type. A complicating factor is that some of the processes of ice formation at the base of a glacier (e.g. regelation) are similar to ice segregation processes occurring in permafrost (Souchez & Lorrain 1991). Landforms resulting from the burial of glacier ice are readily observable in many proglacial and periglacial settings (e.g. kettle lakes). However, the process of ice burial has not been thoroughly examined. This paper presents the findings of our investigations at a site, where the burial of glacial ice by deltaic deposition in a small lake was documented. 2 STUDY SITE The study site is in the southern portion of Bylot Island in Arctic Canada (79 30 N, W). The area is well within the zone of continuous permafrost (Heginbottom 1995), and an equilibrium permafrost thickness of up to 400 m has been estimated from shallow ground temperature measurements (Moorman & Michel 2000). The glacial history of the island is complex; however, the extent of modern glaciers is thought to have been relatively stable over the last several tens of thousands of years, with a minor glacial maximum occurring within the last 100 years (Klassen 1993). Some of the glaciers on southern Bylot Island are currently retreating while other glaciers show little or no sign of retreat. Near the snout of Glacier C93 (Fig. 1), a tongue of ice flows northwest from the main glacier following a major fault line and the junction between the Archean-Proterozoic crystalline basement rock and the Cretaceous-Tertiary sedimentary platform. The study site is located approximately half way along this tongue, where an indentation in the valley side has enabled a small lake to be dammed against the side of the glacier, into which a delta grew and partially covered the glacier. 3 METHODS Aerial photographs were used to study the historical record of lake levels and delta growth. The stratigraphic relationships within the delta were examined in exposures at stream cut-banks. The subsurface geometry of the delta architecture was imaged with groundpenetrating radar (GPR). Ice physical properties were also recorded to differentiate the origins of various ice units. 4 RESULTS AND DISCUSSION 4.1 Sedimentation history Aerial photographs and repeated site investigations from 1993 to 1999 revealed considerable temporal 777

Figure 1. A false colour 1989 Landsat image of the study area on Bylot Island, Arctic, Canada.

Aerial photograph from 1948 showing the ice dammed lake at just below the D1 stage. The ice-marginal stream enters the eastern basin at the top right.

Figure 3. Aerial photograph from 1982 showing the residual ponds after lake drainage.

2 Figure 1. A false colour 1989 Landsat image of the study area on Bylot Island, Arctic, Canada. Note that the eastern basin still has a residual lake (black), while the western basin is almost completely drained (grey). Figure 2. Aerial photograph from 1948 showing the ice dammed lake at just below the D1 stage. The ice-marginal stream enters the eastern basin at the top right. The terrestrial stream enters the western basin from the lower portion of the photograph. The outlet stream is to the left. The light coloured water reveals the high suspended sediment content. Figure 3. Aerial photograph from 1982 showing the residual ponds after lake drainage. The eastern basin is still ice covered as it is fed by cold glacier runoff, while the western basin is fed from the warmer terrestrial stream and is already ice free. variability in lake levels (Figs 2, 3). The lake into which the delta formed consists of two main basins. The basin to the east is fed by an ice-marginal stream running along the south side of the glacier, while a terrestrial stream flows over the ice-free sedimentary terrain to the south of the glacier into the western basin. The general history of the site, as outlined in Moorman and Michel (2000), consists of four major stages: (1) initial glacial retreat resulting in the development of the ice-cored lateral moraine and allowing the initial lake to form, into which delta complex D1 grew (Fig. 4), (2) expansion of the glacier that blocked the outlet and caused the water levels to rise resulting in the deposition of delta complex D2, (3) glacial retreat enabling partial drainage of the lake, leading to incision of the D1 and D2 complexes and the formation of lower delta complex D3, and currently 778

3 Figure 5. Stratigraphic section through an unburied portion of the lateral moraine to the east of the delta within the western basin. The thickness of the units containing foreset beds range from less than 0.5 m in D2 to over 4 m in D3, indicating that lake stands were highly variable during formation of the delta. The differences in sand volume and lateral extent of the three deltaic complexes indicate that the duration of lake stands resulting in the deposition of each of the delta complexes was variable as well (Fig. 4). This is supported by the many strand lines on the valley sides. Direct measurements were also made of water levels dropping by 5 m in 7 days in the late summer of Ice burial Figure 4. Map of the western basin showing the distribution of deltaic sands as of (4) continuing retreat of the glacier resulting in the further lowering of the base level and down cutting through the D3 delta complex. The ice-marginal stream carries a wide range of sediment, dominated by silt and clay. A bedrock ridge (center in Fig. 2) provides a stable base level for the eastern basin resulting in relatively constant lake levels in the basin. The base level for the western basin is controlled to a greater extent by the glacier position as the outlet stream is ice marginal (at left in Fig. 2). The terrestrial stream entering the western basin originates in the poorly consolidated sandstone to the south and is laden with sand and almost completely devoid of silt and clay. The silt and clay carried by the ice marginal stream entering the lake from the east blanket the entire lake bed. When water levels are high, there is considerable mixing and the suspended sediment from the ice marginal stream is dispersed throughout both basins (Fig. 2). When lake levels are low, the basins are isolated and the ice marginal stream does not contribute to sediment deposition in the western basin (Fig. 3). Ice burial at this site has occurred in three ways; containment within the lateral moraine as the glacier retreated, burial beneath deltaic sediments as the delta expanded over the lateral moraine and then onto the exposed glacier ice, and by the redeposition of sediment derived from retrogressive thaw slumps onto exposed ice. Compared to other glaciers in the region, Glacier C93 is currently retreating slowly and has very insignificant lateral moraines. Where visible, the crest of the lateral moraine rises less than 2 m above the surrounding terrain. Exposures and GPR profiles revealed that a 4 10 m thick ice core is present beneath approximately 1 m of ice-poor till and a lacustrine mud cap (Fig. 5). As a result of the delta extending out onto the glacier, up to 20 m of ice has been preserved beneath the deltaic sand. As the delta first built out over the lateral moraine, it encased the ice-cored moraine as well as directly covering the glacier ice. Deltaic sediment at lower elevations (e.g. the lower reaches of the D3 complex) are typically well sorted fine to medium size sand (Fig. 6). Deltaic deposits at higher elevations (e.g. D1) are frequently interbedded 779

4 connected and the only sediment entering the western basin is the sand from the terrestrial stream. The interbedded nature of the sediment higher in the delta suggest that there was considerable fluctuation in lake levels in the past as there is today. The ice buried at the study site varies greatly in appearance and physical properties. Ice encased by the moraine ranged from being laden with very poorly sorted glacial debris, to having a fine grained sediment content of less than 5%. This is typical of the basal ice observed in the region. White bubble-rich, sediment-free ice, similar to the exposed ice of the neighbouring glacier, was discovered at higher elevations and positions closer to the centre-line of the glacier. The ice exposed immediately beneath the D1 and D3 deltaic sands was clear, bubble and sediment free, roughly 1-cm diameter equidimensional crystals with no preferred C-axis orientation. Figure 6. Stratigraphic section through delta complex D3, showing multiple layers of sand with appreciable amounts of organics contained with two thin layers. Figure 7. Stratigraphic section through the delta complex D1 showing considerable lacustrine deposits accumulating before deltaic sedimentation began. with thin stringers of lacustrine mud (Fig. 7). This is the result of the two basins being hydrologically connected during high stands, and the fine grained sediment from the ice-marginal stream entering the eastern basin being dispersed into the western basin. The layers of lacustrine mud ranged 1 20 mm in thickness. In comparison to the grey till found in the moraine, the lacustrine mud was brown in colour. When the lake levels are low, the two basins are not 4.3 Ice segregation Segregated ice was present within the interbedded sand and mud units. The shallow lacustrine setting, where frost susceptible sediments are interbedded with high permeability sediments offered an ideal environment for lake bottom ice segregation, similar to that reported by Burn (1990). The ice content ranged from isolated lenses less than 1 cm in size to tabular layers of ice up to 1 m thick, with laterally continuous ice layers less than 5 cm thick being the norm. The heave demonstrated by the thicker layers of segregated ice represent extraordinary heave rates considering the unstable thermal conditions associated with dramatically changing surface hydrological conditions. Overall, segregated ice makes up a very small portion of the ice covered by the deltaic sands. It is estimated to be less than 0.1% of the total by volume. Segregated ice was not present in the exposed moraine or surficial lacustrine sediment cap. The upper ice unit in Figure 6 was interpreted to be segregated ice, while the lower ice unit is likely glacier ice. Below 4 m, the exposure was covered with colluvium. 4.4 Temporal context Down cutting within the younger D3 complex exposed a section containing two organic-bearing silty-fine sand units overlying clear massive ice (Fig. 6). The lowermost silty fine sand unit rested directly over the massive ice and was subsequently covered by 45 cm of medium to coarse deltaic sand. The upper organic-rich silty sand unit, radiocarbon dated at 1300 / 60 years B.P. (WAT# 6335), underlies a thin (approximately 5 cm) continuous horizon of coarse equigranular ice. The remainder of the sequence above the ice layer consisted of deltaic sands. 780

Figure 8. A schematic cross section through the delta in the western lake basin, with the location of the sections shown. Note that the legend is the same as in Figure 4.

5 Figure 8. A schematic cross section through the delta in the western lake basin, with the location of the sections shown. Note that the legend is the same as in Figure 4. The radiocarbon date of 1300 years on organics within this D3 complex provides a time frame for the formation of this delta and the older D1 and D2 complexes. The organic material used for dating clearly did not grow in situ, but was carried into the lake from the surrounding catchment area. Since it is likely that the organics represent material accumulated over a period of time, the 1300 year age should be considered as a maximum age for deposition within the D3 sequence. Since the D3 complex is the youngest sedimentary deposit at the site, it is likely that no extended high level lake has occupied the valley since. The 1948 aerial photograph in Figure 2 confirms that lakes do periodically flood the study site; however, the lack of associated sediments indicates that they are of short duration. As noted earlier, rapid fluctuations in lake level were observed during our investigations. 4.5 Ice stability Preservation of segregated ice and the ice buried beneath the lateral moraine and delta complexes at the side of Glacier C93 depends greatly on the erosional potential of the stream and the hydrological base level created by the glacier ice dam (or lack thereof). In general, the deltaic sand remains stable when thawed thus providing a durable cover for the buried ice. Unlike in the deltaic sands, when the ice encased by fine-grain sediments (e.g. segregated ice and moraine ice core) melts, a liquid slurry forms. As a result, retrogressive thaw slumps are common in areas of exposed moraine. The thaw slumps are characterized by sloping headwalls 2 10 m in height with near vertical faces near the surface of the ice-poor glacial or lacustrine sediment capping the massive ice. In the current configuration, retrogressive thawing of the ice-rich lateral moraine continues unabated, although debris is accumulating at the toe of the slumps. On the other hand, the ice-poor nature of the deltaic sands make the areas covered by sand more stable, and they are not subject to erosion except when undercut by a stream. Thus the major threat to preservation of buried glacier ice within the deltaic complexes is by down cutting and exposure due to stream migration. 5 CONCLUSIONS Sandy deltaic sediments that were deposited within temporary lakes dammed along the ice margin have covered glacier ice and the adjacent lateral moraine. The average lake stand has since dropped and the stream is depositing its sediment load at lower levels. Burial of the glacier ice by deltaic sand in this permafrost environment has resulted in its preservation for hundreds of years. The ice-poor nature of the deltaic sands makes them less susceptible to thermal erosion and thus enhances the probability of long-term glacier ice preservation. The greatest threat to maintaining a stable ice-cored deltaic complex is down cutting by the stream during periods of low lake level, and the erosion of buried ice-cored moraines. ACKNOWLEDGEMENTS The authors would like to thank the Polar Continental Shelf Project for its logistical support, and NSERC for 781

6 its financial support to F.A. Michel. The field assistance from Mark Elver, Deb Kliza, Lynn Moorman, Amy Lyttle, and Carolyn Deacock is greatly appreciated. Renard Emmanual and an anonymous reviewer are acknowledged for their very useful comments in improving the manuscript. REFERENCES Burn, C.R Frost heave in lake-bottom sediments, Mackenzie Delta, Northwest Territories. Permafrost- Canada: Proceedings of the Fifth Canadian Permafrost Conference, Collection Nordicana No. 54, Centre d études nordiques, Université Laval, Quebec, Heginbottom, J.A Canada Permafrost. In National Atlas of Canada, 5th edition. Ottawa, Natural Resources Canada. Fujino, K. (ed.) Characteristics of the Massive Ground Ice Body in the Western Canadian Arctic Related to Paleoclimatology Sapporo, The Institute of Low Temperature Science, Hokkaido University. Klassen, R.A Quaternary Geology and Glacial History of Bylot Island, Northwest Territories. Geological Survey of Canada Memoir 429, Ottawa, Department of Energy, Mines and Resources Canada. Mackay, J.R The origin of massive icy beds in permafrost, western Arctic coast, Canada. Canadian Journal of Earth Sciences 8: Michel, F.A Nature and History of Ground Ice in Yukon. Contract report, Ottawa, Energy, Mines and Resources, Canada. Moorman, B.J. & Michel, F.A The burial of ice in the periglacial environment on Bylot Island, Arctic Canada. Permafrost and Periglacial Processes 11: Moorman, B.J., Michel, F.A. & Wilson, A The development of tabular massive ground ice at Peninsula Point, N.W.T., Canada. Proceedings, Permafrost, Seventh International Conference, June 23 27, Lewkowicz, A.G. & Allard, M. (eds). Collection Nordicana No. 57, Centre d études nordiques, Université Laval, Quebec, Robinson, S.D., Moorman, B.J., Judge, A.S., Dallimore, S.R. & Shimeld, J.W The application of radar stratigraphic techniques to the investigation of massive ground ice at Yaya Lake, Northwest Territories. Muscox 39: Souchez, R.A. & Lorrain, R.D Ice Composition and Glacier Dynamics. Berlin, Springer-Verlag. Vaikmae, R., Michel, F.A. & Solomatin, V.I Morphology, stratigraphy and oxygen-isotope composition of fossil glacier ice at Ledyanaya Gora, Northwest Siberia, Russia. Boreas 22:

THE DEVELOPMENT OF TABULAR MASSIVE GROUND ICE AT PENINSULA POINT, N.W.T., CANADA

THE DEVELOPMENT OF TABULAR MASSIVE GROUND ICE AT PENINSULA POINT, N.W.T., CANADA Brian J. Moorman 1, Frederick A. Michel 2, Alex T. Wilson 3 1. Earth Science Program University of Calgary 2500 University