Plate 26.1 Location map of the M'Clintock Channel Ice Stream (a) and satellite imagery of the bedform imprint on Victoria Island (b). The late glacial imprint of the ice stream occupies present day M'Clintock Channel and infringes on western Prince of Wales Island and eastern Storkerson Peninsula (thin red lines). Landsat satellite imagery in (b) shows the margin of the late glacial ice stream imprint on Storkerson Peninsula. However, older flow patterns (thin black lines) indicate that the ice stream may have been much bigger during the Last Glacial Maximum, extending eastward and occupying Hadley Bay.

Plate 26.2 Map of the inferred surface current-driven iceberg drift directions from the Canadian Arctic Archipelago (solid arrows) and concurrent hypothesized drift of Russian pack ice (broken arrows) during glacial intervals (modified from Bischof & Darby, 1997). The expanded M'Clintock Channel Ice Stream is shown in red and it can be seen that icebergs issued from this region would enter Fram Strait relatively rapidly compared to present day conditions (Bischof & Darby, 1997). The box indicates the area shown in Plate 26.1a.

Plate 27.1 Ice velocity of Petermann Gletscher, northwestern Greenland measured from Radarsat-1 interferometric synthetic-aperture radar (InSAR) data. Grounding line inferred from double difference InSAR is white. Ice flow is to the north. Bounding box of calculation of bottom melt rates is dotted white. Inset shows velocity V (red, in myr-1, left scale), thickness H (blue, in m, left scale) and bottom melt rate B (black, in myr-1, right scale) calculated over the glacier width, versus the distance (in km) from the grounding line.

Runoff Mapping India
Plate 30.1 Bathymetry of the mid-Norwegian shelf showing cross-shelf troughs and intervening banks. (Modified from Ottesen et al., 2002.)

Plate 30.2 Multibeam swath bathymetric image of sediment drifts and intervening channels on the western Antarctic Peninsula continental margin. (Modified from Dowdeswell et al., 2004b.)

Antarctic Peninsula Bathymetric Map

Plate 32.1 Modelled modern-day mass balance fields on the Greenland Ice Sheet and in the western Arctic, using degree-day methodology and climate fields from the NCAR Community Climate System Model (CCSM), v.2.0, with climate fields provided by B. Otto-Bleisner (personal communication, 2003). (a) and (b) show the precipitation and temperature maps that go into the calculation of mass balance fields. (c), (d) and (e) plot annual accumulation, ablation, and mass balance, all in myr-1 water-equivalent. (f) shows ice sheet thickness (m). Model resolution is 1/6° latitude by 1/2° longitude.

Plate 32.1 Modelled modern-day mass balance fields on the Greenland Ice Sheet and in the western Arctic, using degree-day methodology and climate fields from the NCAR Community Climate System Model (CCSM), v.2.0, with climate fields provided by B. Otto-Bleisner (personal communication, 2003). (a) and (b) show the precipitation and temperature maps that go into the calculation of mass balance fields. (c), (d) and (e) plot annual accumulation, ablation, and mass balance, all in myr-1 water-equivalent. (f) shows ice sheet thickness (m). Model resolution is 1/6° latitude by 1/2° longitude.

Plate 32.1 Continued

Temperature difference (2200-present) fC)

Temperature difference (2200-present) fC)

0 1 2 3 4 5 6 7 8 9 Melt rate change, 2200-present (m/yr w.eq.)

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 Snow accumulation change (m/yr w.eq.)


-280 -200 -120 Change in ice thickness (m)

Plate 32.2 Modelled air temperature, mass balance and ice thickness fields on the Greenland Ice Sheet and in the western Arctic for 2200. All plots are shown as difference maps from the reference modern-day (2000) conditions shown in Plate 32.1. (a) Difference in air temperature, 2200 - present (°C). (b & c) Difference in snow/ice ablation and accumulation rates, 2200 - present (myr-1 water-equivalent). (d) Difference in ice sheet thickness, 2200 - present (m). Model resolution is 1/6° latitude by 1/2° longitude.

Plate 38.1 Four ice sheet-scale reconstructions using inversion protocols by (a) Boulton et al. (1985), (b) Dyke & Prest (1987a), (c), Boulton & Clark (1990a,b) and (d) Kleman et al. (1997). The emphasis in (a) and (b) is on post-LGM configuration changes, whereas in (c) and (d) it is on ice-sheet evolution through the last glacial cycle, with an emphasis on events pre-dating the LGM. (Panel (a) is reproduced with permission from the Geological Society, London. Panel (b) is modified from Dyke & Prest (1987a). Panel (c) is reprinted by permission from Nature, 346, 813-817 (1990), copyright 1990 Macmillan Publishers Ltd. Panel (d) is modified from Kleman et al. (1997).)

Plate 38.2 A visualization of the differences between four ice-sheet reconstructions, focusing on particular time-space data domains and the data types used in the inversion procedures. Coloured items mark the primary data domains: thick red line marks the deglacial landforms; green, blue and purple mark glacial 'events' reflected by till lineations pre-dating the final decay phase; red diamonds schematically illustrate radiocarbon dates (which always reflect ice-free conditions); orange colour represents a 'stretching' of the deglacial landform record for inferences about older non-deglacial events.

Plate 38.3 The genetic and inversion problems in glacial geomorphology are associated with fundamentally different suites of assumptions, scale and generalization considerations, as well as methodical issues.

Details from Glacial map of Canada (Presleial 1968)

Details from Glacial map of Canada (Presleial 1968)

Plate 38.4 Details from the Glacial Map of Canada (Prest et al., 1968). (a) The southwestern sector of Keewatin displays a 'classic' glacial landscape where abundant eskers parallel a single coherent system of till lineations. Fields of ribbed moraine occur in the proximal part of the till lineation swarm, and probably mark areas that changed from cold-based to warm-based (wet-bed) conditions (Hättestrand & Kleman, 1999). This type of landscape is thought to have formed in marginal wet-bed zones of substantial width during ice-sheet decay (Kleman et al., 1997). (b) Eskers cutting obliquely across the convergent head zone of the Dubawnt lineation swarm in Keewatin. The lineations were probably formed by a short-lived ice stream (Stokes & Clark, 1999). The eskers indicate that a major change in flow direction occurred between the ice-stream phase and the deglaciation stage. (c) An intersection zone in central Quebec-Labrador where two different glacial landscapes, with opposing flow directions, occur in close contact. The southwest-orientated landscape in the lower left-hand half of the map displays a full suite of deglacial meltwater features aligned parallel to the lineation system, leading us to regard it as being formed during the last deglaciation. The NNE-orientated Ungava Bay swarm in the northeastern half, in contrast, almost entirely lacks eskers, leading us to believe that the lineations formed underneath central portions of the ice sheet during an earlier flow phase (Jansson et al., 2003). The apparent ice divide is probably entirely fictitious and instead denotes only the up-glacier boundary of wet-bed conditions of the southwastward flow, during the last deglaciation. (d) An isolated patch of relict N-S and NNE-SSW orientated lineations southwest of the Dubawnt lineation swarm. The patch is probably an erosional remnant of a lineation system formed during a glacial event that pre-dated the LGM (Kleman et al., 2002). Hence, its present extent is governed by subsequent Dubawnt ice-stream erosion to the north and sheet-flow erosion to the south. Eskers from the last deglaciation cut the relict north-south trending lineations in the patch at almost right angles. (e) A landscape without eskers on the northwestern flank of the Keewatin sector. Because lineations and aligned striae yield few clues to their age or the duration of flow, these landscapes are difficult to treat in inversion models. If they are part of a cold-based deglaciation landscape (which, typically, lacks eskers), the distribution of glacial-lake shorelines, spillways and drainage channels may give the only solid guidance for decay reconstruction in such areas (Borgström, 1989; Jansson, 2003). Reproduced with the permission of the Minister of Public Works and Government Services Canada, 2004 and Courtesy of Natural Resources Canada, Geological Survey of Canada.

Plate 38.5 (a) Relict surfaces lacking glacial landforms, such as the Tjeuralako Plateau, northern Sweden, are interpreted to mark sustained frozen-bed conditions under one or more successive ice sheets. Cosmogenic dating (10Be) of exposed bedrock on this plateau yielded an exposure age of 45kyr, indicating inheritance from one or more previous ice-free intervals, and negligible erosion by the last ice sheet (Stroeven et al., 2006) (b) In Fennoscandia, periglacially formed surfaces, such as this striped boulder surface (A) at Tjuolma, The Ultevis plateau, Sweden, occur preferentially on uplands with clear erosional boundaries (Kleman, 1992) to younger glacial landscapes (B) comprising fluting and drumlinization from the last ice sheet. An erratic perched on surface (A) yielded an exposure age (10Be) of 7.4 kyr, whereas bedrock exposed on the same surface yielded exposure ages of 32.7 and 35.2 kyr (Fabel et al., 2002; Stroeven et al., 2006). (c) The Stadjan-Nipfjallet upland in the southern Scandinavian mountains comprises marginal moraines older than the the last ice sheet, and a >2-m-deep weathering mantle, indicating negligible erosion by the last ice sheet. (d) Relict surfaces and glacially eroded surfaces display an archipelago-like pattern west of Kiruna, northern Sweden. The flow pattern indicated by lineations is consistent with the pattern expected for thick overriding ice and polythermal bed conditions, but inconsistent with the flow pattern expected from a thin-ice scenario comprising nunataks and ice-tongues in valleys. Modified from Kleman et al. (1999). (e) Frozen-bed extent under the Fennoscandian Ice Sheet, as inferred for three time periods. Approximate LGM extent, largely based on the distribution of ribbed moraine, inferred to have formed during transition from frozen-bed to thawed-bed conditions, shown as light grey shading. Approximate extent during Younger Dryas, after onset of major ice streams in Finland, is shown as medium grey. Black mark zones with abundant pre-Late Weichselian glacial and non-glacial landforms, and stratigraphic and cosmogenic dating evidence for non-erosive frozen-bed conditions under the last ice sheet. (f) Hughes (1981b) hypothesized that terrestrial core areas would comprise a frozen-bed core, a patchy transition zone to mostly thawed-bed conditions, lenticular frozen-bed patches in ice-stream heads, and ice stream corridors with sharp thermal boundaries to intervening ice-stream ridges. The collective evidence from the Fennoscandian and Laurentide ice sheet areas (Dyke et al., 1992; Kleman et al., 1999) confirms all essential aspects of this hypothesis.

a/ Chronological domains


Dateable domain, absolute b/ The relative-age domain

Dateable domain, absolute

Undateable domain only relative ages accessible

Undateable domain only relative ages accessible c/ Possible subglacial dating tool: deep-reaching ice streams


Plate 38.6 (a) Two chronological domains are defined: the extramarginal domain, to which all currently available dating methods pertain (radiocarbon, OSL, cosmogenic and amino-acid racemization techniques), and the subglacial, which is not accessible by any current absolute dating methods. The chronology of the deglacial envelope is currently defined mainly by a scatter of radiocarbon and OSL dates of widely varying spatial density. Through ancillary data, e.g. pollen, a specific dating often can be related to climatic evolution, but only rarely can any direct link to ice-dynamics be established. (b) In the subglacial dating domain, only relative ages are readily available. The relative age of flow events is established through cross-cutting relationships (Clark, 1993). Through analysis of landform assemblages and relative ages, inferences about ice dynamics can be made, but only rarely can any information pertaining to climate be gained. (c) Ice-stream landscapes have a substantial but yet largely unrealized potential for absolute dating of subglacial events, and may prove to be the only realistic tool for that purpose.

Plate 38.7 Palaeo-ice streams in the northwestern Canadian Arctic. Prime data sources are Prest et al. (1968) and morphological mapping using Landsat MSS and TM imagery (Kleman, unpublished data): thin black lines show till lineations mapped from MSS imagery; blue lines show major ice streams of type 1; green lines show type 2 ice streams; red lines show ice streams of type 3; grey lines show unclassified ice streams. See text for description of ice-stream types. Grey shading shows areas displaying relict non-glacial morphology and areas inferred to have been the sites of frozen-bed interstream ridges in the last ice sheet.

a/ Landform zonation


Till lineations Ribbed moraine

Relict landscapes

Ice stream -

Frozen bed

Thawed bed b/ Map representation of swarms





Plate 38.8 (a) Domains of landform formation in a time-distance diagram. Eskers form close to the ice margin in a time-transgressive fashion. Ribbed moraine is inferred to form during transition from a frozen bed to a thawed bed (Kleman & Hattes-trand (1999). Glacial lineations form wherever the bed is thawed and subglacial sediments are available. (b) The three swarm types we recognize are event swarms, ice-stream swarms and the deglacial envelope. The latter is defined by eskers and other meltwater landforms and may or may not be associated with till lineations. Swarms are simplified and spatially delineated map representations of many individual landforms.

Was this article helpful?

0 0

Post a comment