Discussion

We hypothesize that the types of landforms and sediments within a particular region are the result of how the ice sheet responded to the fundamental influences of climate, bed geology and topography. These influences were filtered in a non-linear manner by the ice sheet. The combination of these factors produced different dynamic behaviours of the ice-sheet margin and different depositional environments in various regions of northern USA. Below we discuss the influence of each factor on glacier dynamics and the resulting landsystems in each region.

6.6.1 The Role of Climate in the Genesis of Landform-Sediment Landsystems

Climate is a first-order control on glacier behaviour. Besides determining mass balance and the extent of ice, climate also influences the temperature of ice as well as its bed near the margin. It also influences the amount and nature of water movement. During the LGM the climate along the southern LIS margin varied spatially (Kutzbach, 1987). In the western and northern Great

Lakes regions, and perhaps in northern New England, permafrost formed before, during, and even after the LGM (Pewe, 1983). It is likely that the ice-sheet margin advanced over this permafrost many times. It may have taken several thousand years for this overridden permafrost to melt, depending on the initial thickness of the permafrost and the thickness and temperature of the overlying ice (Cutler et al., 2000). In the southern Great Lakes region of Illinois, Indiana and Ohio, permafrost was probably discontinuous (Johnson, 1990). Abundant wood in tills in this region (Ekberg et al., 1993; Hansel and Johnson, 1999) suggests that ice advanced into a warmer region than that in Wisconsin, Michigan and Minnesota, where till contains no wood and sub-till organic material indicates tundra conditions. In New York and New England ice may have advanced over permafrost as relict ice-wedge casts and ice-wedge polygons are reported there also (Pewe, 1983). In southern New England the proximity to the Atlantic Ocean may have warmed this part of the margin compared with areas to the north or west (Gustavson and Boothroyd, 1987).

In Wisconsin and Minnesota this permafrost lasted until about 13,000 14C years BP in the south, disappearing northward about 3,000 years later (Clayton et al., 2001). Not only was permafrost overridden, it must have reformed at times on the deglaciated surface, between 17,000 and 13,000 14C years BP. Discontinuous permafrost may have also been present in some areas of landsystem A (Johnson, 1990), but we believe this permafrost was not sufficiently extensive to have had the same impact as it did in the area of landsystem B. Throughout most of landsystem A ice advanced into spruce forest.

The marked difference between landsystem B and landsystems A and C is the presence of drumlins, tunnel channels and high-relief hummocky end moraines. We believe that these features are consistent with an interpretation that ice advanced onto permafrost and eroded pre-existing landforms and sediments. We suggest that when ice was at its maximum position, a zone of subglacial permafrost up to 100 km wide was present. Behind this zone, a zone of partially frozen bed may have been present back to the up-ice edge of the drumlin zone. Subglacial water was produced at the bed in this zone and probably accumulated behind the wedge of frozen ground near the terminus because insufficient water was able to drain through the groundwater drainage system. The trapped water was eventually released through tunnel channels that intersected the end moraine (Attig et al., 1989; Clayton et al., 1999; Cutler et al., 2002). We attribute the mobility and deformation of sediments in the drumlin zone in part to the likely high pore pressure and the high water content of the subglacial sediments upstream from the frozen margin.

In the southern Great Lakes region, where landsystem A is dominant, we suggest that the glacier bed was thawed and that basal motion, through some combination of glacier sliding, ploughing and bed deformation, was the primary result of ice motion. We suggest that little erosion was accomplished in this region. Mickelson and others (1983) report radiocarbon dates on wood between 23,000 and 14,000 14C years BP in Illinois and Indiana. Much of the wood is spruce and this suggests that ice advanced into a spruce forest or spruce parkland throughout the late Wisconsin, with the exception of the period between 17,500 and 16,000 14C years BP, when there are no wood dates from Illinois. Permafrost features are not as abundant as they are to the north, although Johnson (1990) reports ice-wedge casts and other permafrost features on late Wisconsin deposits that may have formed during this interval after ice advance to the maximum position. We conclude that the absence of drumlins in this area is because of the prevailing thawed-bed conditions and not because of a difference in topography or till texture. Drumlins are present on Illinoian surfaces just outside the late Wisconsin margin in Illinois in till of similar texture to that of late Wisconsin till plains (Lineback et al., 1983), indicating that texture was not a significant influence on the presence or absence of drumlins.

6.6.2 The Role of Bed Geology and Topography in the Genesis of Landsystems

In addition to climatic controls, the behaviour of the ice sheet was influenced by bed geology and topography. The distribution of bedrock lithologies was important to ice dynamics and hence landsystem distribution for several reasons. First, differences in the lithology, and therefore rate of bedrock erosion, have influenced the regional topography (Fig. 6.19) and hence regional patterns of ice flow. The lobate nature of the southern LIS was determined by bedrock topography, with lobes channelled down each of the basins in the Great Lakes region. For example, these and other major basins must have been present prior to the last glaciation, and were deepened by an unknown amount during the late Wisconsin. The style of ice-marginal behaviour differed between areas where the ice was advancing up a regional slope and those where it advanced down a regional slope. In addition, large lake basins significantly altered the mass balance and ice-surface profiles (Cutler et al., 2001). In the high-relief areas of central and northern New England, the style of deglaciation and resulting deposits were greatly controlled by topography (Koteff and Pessl, 1981; Mulholland, 1982; Stone and Peper, 1982; Goldthwait and Mickelson, 1982).

Second, the rate of erosion of bed materials also influenced the amount of debris being carried by the ice, and topography determined if that material could be transported away from the terminus. Ice in northern Wisconsin and in southern New England retreated progressively up a surface that was gently sloping away from the ice margin. This produced detachment of stagnant ice from active ice in zones up to 20 km wide (Koteff and Pessl, 1981; Mulholland, 1982; Stone and Peper, 1982; Gustavson and Boothroyd, 1987; Ham, 1994; Johnson et al., 1995). As these stagnant-ice zones melted, they produced large hummocky moraines with icewalled lake plains (called 'high kames' in New England). In the Dakotas, Minnesota, Wisconsin, New York and northern New England, ice retreated into lake basins and lowlands. This produced ideal conditions for the formation of proglacial lakes that trapped fine-grained sediments. Subsequent readvances after the LGM incorporated large amounts of these glacilacustrine silts and clays (Acomb et al., 1982; Clark, 1994b) and deposited these as finegrained tills (Fig. 6.19).

Lithology also influenced the nature of subglacial fluvial deposits. For example, the texture of glacier deposits controls the present appearance of glacial landforms (Fig. 6.19). High relief and steep slopes tend to occur where till is sandy. Silt- and clay-rich tills tend to have low-relief landforms because of postglacial weathering and erosion (Clayton and Moran, 1974). Finally, the nature and abundance of glacifluvial deposits must in part be a function of the availability of coarse material. Eskers and outwash fans cannot occur unless there is sufficient coarse sediment to be deposited.

Finally, the hydraulic conductivity of the glacier bed controls subglacial drainage and therefore water pressure and basal motion (Arnold and Sharp, 1992; Clark and Walder, 1994). The effective stress at the bed is in part controlled by the ability of water to drain towards the ice margin. If flow is impeded, effective pressure declines and basal motion may be enhanced. Consequently, lithologically dependent hydraulic properties of bedrock and overlying sediments

Bedrock Type I I No data I ] Sedimentary ligneous or metamorphic ^

Bedrock Type I I No data I ] Sedimentary ligneous or metamorphic ^

Till Texture Type

I_| Not classified

I I Sandy till I 1 Silty till Clay till

Till Texture Type

I_| Not classified

I I Sandy till I 1 Silty till Clay till

Figure 6.19 A) Map showing bedrock types in the northern USA (map derived from the digital version of the geologic map of the US, Schruben et al., 1999). B) Map showing the dominant texture of till matrix in the northern USA (map derived from data of Soller and Packard, 1998).

exert a direct influence on spatial variations in ice dynamics. The distribution of landsystems found around the southern LIS reflects this influence. In the western region and the southern Great Lakes region where landsystems A and C dominate, the bed is composed of thick silty or clayey tills derived primarily from Palaeozoic sedimentary rocks. This bed would have been soft, smooth and easily deformed, with low hydraulic conductivity. Low basal shear stress would result in low-profile, unstable glacier lobes. Enhanced subglacial sliding, bed deformation or surging behaviour would be associated with this bed type. In the northern Great Lakes and New England where landsystem B dominates, the bed is composed of igneous or metamorphic rocks or sandstone. This bed would have been hard, rough, thinly covered by sediment, and had a higher hydraulic conductivity compared with soft-bedded areas. This may have caused ice lobes to be thicker and more stable than those advancing over a soft bed.

6.6.3 The Role of Glacier Dynamics in Influencing the Distribution of Landsystems

A final factor in producing various landsystems is the dynamic behaviour of the ice sheet during advance to its maximum extent, and during deglaciation (Fig. 6.20). Dynamic behaviour directly results from the climate and bed geology discussed above. The ice margin appears to have retreated progressively at rates of between 50 and 500 m/year along much of the southern margin of the LIS in the Great Lakes and New England regions (Andrews, 1973; Mulholland, 1982; Colgan, 1996; Ham and Attig, 2001). Evidence for this includes radiocarbon chronologies, small push moraines, numerous well-defined retreat moraines, and extensive and well-dated varve deposits in proglacial lakes (Ashley, 1975; Johnson et al., 1999). In the western prairie region it appears that ice surged into lowlands several times and then experienced regional stagnation (Clayton et al., 1985). Retreat rates in this region were higher than 700 m/year and could have been as high as 2000 m/year (Fig. 6.20). As these rates are much higher than advance-retreat rates of normal glaciers it has been suggested that they reflect regional stagnation following each surge of a lobe in the areas west of the Great Lakes (Clayton et al.,

Lobes in the western prairie region also advanced to their maximum advance position out of phase with the rest of the ice sheet margin (Fig. 6.20) (Clayton and Moran, 1982; Mickelson et al., 1983; Clayton et al., 1985). The Des Moines lobe reached its maximum at about 14,000 14C years BP when the rest of the margin was rapidly wasting (Kemmis et al., 1981; Hallberg and Kemmis,

1986). We believe that this is an indication that these lobes surged. The resulting landforms and sediments (landsystem C), consisting of widespread ice-thrust and ice-stagnation features, are compatible with surging as a major process in this region after the LGM.

Reconstructed ice-surface profiles also provide information about ice-lobe behaviour that ties in with our interpretations of landform-sediment landsystems. Most of the major lobes have been reconstructed from ice-marginal features and moraine elevations for their LGM positions and for some deglaciation phases (Mathews, 1974; Beget, 1986; Ridky and Bindschadler, 1990; Clark, 1992; Colgan, 1996; Colgan and Mickelson, 1997; Socha et al., 1999). These profiles show that glacier lobes that created landsystem B in the northern Great Lakes and New England regions had comparatively steep ice-surface profiles. Lobes that created landsystem A in the southern Great Lakes region had lower ice-surface profiles. Finally, the lowest ice-surface profiles have been reconstructed for lobes in the western region (landsystem C), and for lobes that readvanced out of lake basins during deglaciation.

These ice-surface profiles reflect a combination of lithologic and climatic influences, as well as, in the case of the lower-profile lobes, internal ice-dynamic behaviour related to surges. In contrast with steep ice-surface profiles that suggest high basal shear stress and progressively (steadily) advancing and retreating lobes, the extremely low profiles of the western region suggest that ice was very thin and stagnant following surges (Clayton et al., 1985).

10

10

12-

• Big Stone P (A)

12-

~ i ■ Algona P.

? —=----Altamont P.

Q_

14-

? ^— -Bemis P.

□L

14-

CG

H1 ?

CG

^

16-

16

O

o

■t

18

^

18

CO

CO

O

O

^

20

H2 ?

1(

20

e

e

Ol

22

CT

22

<

? Tazwell P.

<

24

Des Moines lobe

24

26

Distance (km)

1000

200 400 600

Distance (km)

200 400 600

Distance (km)

200 400 600

Distance (km)

10

12

□L

14

CG

16

o

18

CO

o

1(

20

e

CT

22

<

24

26

—- ■ Onaway P

- Port Huron P

H1

?

■ \ Nissouri P.

H2

?

Lake Erie lobe

Distance (km)

200 400 600

Distance (km)

200 400 600

Distance (km)

200 400 600

Distance (km)

10

12

n

14

CG

16

n

18

o

1(

20

e

22

<

24

26

= Two Rivers P

(D)

_ Port Huron P

H1

Crown Point P-

? ■

Woodstock P. Livingston P. Putnam P.

?

H2

Shelby P.

Lake Michigan lobe

j Marengo P.

200 400 600

Distance (km)

200 400 600

Distance (km)

200 400 600

Distance (km)

200 400 600

Distance (km)

Figure 6.20 Time-distance diagrams for lobes of the southern margin of the Laurentide Ice Sheet. A) Des Moines lobe in the western region. A major advance occurred in phase with the rest of the margin at about 21,000 l4C years BP, but this was not the maximum advance. The maximum advance occurred at about 14,000 l4C years BP when the rest of the margin had retreated 100-300 km behind the Last Glacial Maximum margin (Kemmis et al., 1981; Hallberg and Kemmis, 1986). B) Lake Superior lobe (Attig et al., 1985). C) Green Bay lobe (Colgan, 1996). D) Lake Michigan lobe (Hansel and Johnson, 1999). E) Lake Erie lobe (Clark, 1994). F) New England (Clark, 1994). Grey shading shows Heinrich events at about 14,500 (Hl) and 20,500 (H2) l4C years BP.

1000

1 000

1 000

1 000

1 000

Project Management Made Easy

Project Management Made Easy

What you need to know about… Project Management Made Easy! Project management consists of more than just a large building project and can encompass small projects as well. No matter what the size of your project, you need to have some sort of project management. How you manage your project has everything to do with its outcome.

Get My Free Ebook


Post a comment