05. Glacial Maximum
CHAPTER 5
Glacial Maximum
Glacial maximum occurred when the largest volume of ice and snow covered the land (see Figure 6.3). At this time, some ice sheets would be melting, while others would be growing, but the net volume of ice would begin decreasing. Figure 5.1 presents the postulated areas of glaciation at maximum ice volume and the position of the storm tracks at that time. The time to reach glacial maximum, in this post-Flood model, probably seems like mere speculation. It is speculation, in that the many variables for the post-Flood climate cannot be precisely known. But ice age maximum depends on two principal controlling variables-the coolness of the summers and the annual snowfall. When one, or both, of these conditions ameliorated, the ice sheet would have begun melting. An estimate of the length of time from the Flood to glacial maximum, is, basically, an estimate of how long the controlling conditions favored an increase in ice volume.
Distribution of snow and ice and storm tracks at maximum glaciation. Notation the same as in Figure 3.6. Circular areas within the North American ice sheet represent postulated ice domes. Little sea ice has formed as yet.
Time to Reach Glacial Maximum
Volcanic eruptions during the ice age eventually dropped off, so that significantly more sunshine penetrated to the surface of the earth. But, as long as the other cooling mechanisms-higher cloudiness, highly reflective snow, and low atmospheric carbon dioxide remained effective, the volume of ice would continue to increase. The first two cooling mechanisms depend upon the available moisture at mid and high latitudes, while the third is of small consequence, as explained in Chapter 3. High atmospheric moisture would be maintained as long as the ocean surface was sufficiently warm. Once the moisture source dwindled enough, cloudiness and snowfall would decrease, and the solar radiation would become more effective in melting snow and ice. Therefore, the temperature of the ocean surface at mid and high latitude is the controlling variable for the length of glacial buildup. The gradual cooling of the ocean surface would be controlled by the temperature of the water below the surface, since the temperature of the surface at mid and high latitudes is determined by the three-dimensional ocean circulation. At some threshold average temperature of the ocean, the warm surface would finally cool to a point at which the volume of global ice would begin to decrease. Consequently, the time for glacial maximum can be approximated by estimating the time to cool from an initial uniformly warm ocean, to an appropriate threshold temperature.
Now that we have established the key variable for determining the time until glacial maximum, how can we obtain a quantitative value for it? It is impossible to determine a precise value, because there are too many variables that are poorly known. Even the present values of these variables are only roughly known. But a “ballpark” figure can be found from reasonable estimates of post-Flood climatology and the heat balance equation for the ocean. Since the variables for the post-Flood climate cannot be precisely calculated, maximum and minimum estimates will mostly be used, thus bracketing the time to reach glacial maximum. As it turns out, both extremes involve very short time spans, compared to the standard ice-age chronology. For an ocean that is cooling, the heat balance equation for the whole ocean, from the Flood to glacial maximum, is (Budyko, 1978, p. 86):
FR - FE - FC= -Q/T(5.1) where FR is the average surface radiation balance per unit of time between absorbed solar radiation and net outgoing infrared radiation; FE and FC are the average evaporative and conductive cooling, respectively, per unit of time; and Q is the total amount of heat that was lost by the ocean, from the Flood until the time of glacial maximum, T. The heat balance for the ocean is represented in Figure 5.2. Geothermal heat added at the bottom of the ocean is small, and will be neglected. The sign of the terms in this, and other balance equations, can cause a problem. In this chapter and in Appendix 1, variables that add heat to the system are positive, and those that subtract heat, are negative. For example, in Equation 5.1, the net radiation balance, FR, adds heat to the ocean, and is positive; this is the only variable that heats the ocean. Evaporation and conduction at the surface, FE and FC, subtract heat from the ocean, and are negative. The net effect of the three terms on the left is to cool the post-Flood ocean. Hence, the term on the right is negative. Solving equation 5.1 for T is complicated, and the details are shown in Appendix 1. Only a brief summary of key points will be given in this section.
T = -Q/(-FRE + FA + FO)(5.2) where FRE is the average radiation balance between absorbed solar radiation and the outgoing infrared radiation per unit time at the top of the atmosphere, and FA and FO are the average higher latitude heat transport by the atmosphere and ocean, respectively, per unit time (see Figure A1.1 and A1.2 in Appendix 1). To find Q, the initial ocean temperature at the end of the Flood was assumed to be 30°C, and the threshold temperature at glacial maximum was estimated at 10°C. A 20°C average cooling of the ocean temperature represents a total heat loss of 3.0 x 1025 calories. The solar radiation absorbed by the earth and atmosphere depends upon the average amount of volcanism, cloudiness, and snow-covered area. These mostly reflect solar radiation, but they also absorb and re-radiate infrared radiation back to space. Since these effects are unknown, it was assumed that the amount of solar radiation lost to space would range from a minimum of 10% to a maximum of 75% of the present average. Infrared radiation loss at the top of the atmosphere was assumed greater in the post-Flood climate than at present, for latitudes higher than 60°. The reason for this assumption is the greater warmth at high latitudes compared with present circumstances, and the fact that infrared radiation loss is roughly proportional to the surface temperature (Budyko, 1978, pp. 93, 94). The values of FA and FO in the present atmosphere are not well known, individually, but their sum can be estimated indirectly. The corresponding values for the post-Flood climate would, very likely, be less than those for modern conditions. Maximum and minimum estimates of these variables are given in Appendix 1. The minimum estimate for the higher-latitude heat transport was set at zero-an extreme approximation. Nevertheless, the minimum value gives us an idea of the sensitivity of equation 5.2 to variable heat-transport values.
Inserting all the variables, with their extreme ranges, into equation 5.2, gives a time to reach glacial maximum of 174 years for a 75% loss of solar radiation with no heat transport to higher latitudes. With a 10% loss of solar radiation and a maximum heat transport, the corresponding time Isaiah 1,765 years. These extremes are unrealistic, but they demonstrate that no matter what estimates are used for the variables in the heat balance equation for the ocean, only a short time is required to cool the ocean. The best estimate is probably 500 years, based on a 25% depletion of solar radiation and a 12.5% decrease in the current values of the atmospheric and oceanic heat transports. An estimate of 30°C for the initial ocean temperature may be too high. If this initial average temperature of the ocean after the Flood is lowered to 25°C or to 20°C, the estimated time for glacial maximum is even shorter. These calculations make it apparent that the post-Flood ice age was a rapid development. The higher values for solar radiation loss (a 50% and a 75% decrease) give us insight into the rate of glaciation. As a consequence of initially high volcanism, such values were probably characteristic of the beginning of the ice age. Table A1.4, in Appendix 1, shows that at these values of solar radiation loss, glaciation would be the most rapid. In other words, during periods of strong volcanism and reflection of solar radiation back to space, the cooling, over land, would be more intense. This, in turn, would cause colder, drier air to blow out over the warm ocean. As a result, ocean cooling would be more rapid, and the amount of moisture evaporated into the air would be higher. Consequently, ice sheets would grow rapidly, with high volcanism. Conversely, they would develop more slowly, or even melt back at the margins, during volcanic lulls. Higher solar radiation loss at the beginning of the ice age is likely the reason glaciation extended into the central midwest of the United States early in the ice age, and then melted back when the solar radiation absorption at the earth’s surface increased, before glacial maximum. Variable volcanism probably was responsible for ice margin oscillations, which would deposit multiple till layers, with non-glacial sediments sandwiched between (see Chapter 7).
Average Post-Flood Ice Depth To find the average ice depth after 500 years of post-Flood glaciation, the available moisture must be estimated. Since the geography of the Northern and Southern Hemispheres is so different, separate estimates of the moisture for the higher latitudes of each hemisphere must be made. In addition, an estimate must be made for the proportion of this moisture that precipitates on the ice sheets. The two sources of moisture for ice-sheet growth are represented in Figure A1.2, in Appendix 1-The first is oceanic evaporation at mid and high latitudes, FE; the second is atmospheric transport from lower latitudes, FA. In Figure A1.2, these sources are represented by the latent heat carried by the water when in vapor form. FE, the latent heat from evaporation at mid and high latitude, was much greater in the early post-Flood climate, than it is at present. As has been explained before, evaporation at mid and high latitudes was the principal source of moisture for the ice sheets. FE and the latent heat portion of FA (FA also includes some energy due to temperature difference) are averages in calories per year from the Flood to ice-age maximum. The estimate of the total latent heat transfer is obtained on multiplication by 500 years. Latent heat is transformed into the mass of moisture by dividing the total latent heat transfer by about 600 calories/gm-the latent heat of condensation. The details of the calculations and the probable amounts of snow falling over the ice sheets are given in Appendix 2. The calculation begins with the solution of FE from the heat-balance equation for the ocean (equation A1.8, in Appendix 1). Estimating from the present values of FR and FO, a range of probable values during the post-Flood period is plugged into equation A1.8, together with an average cooling for the ocean of 20°C in 500 years. Ranges in the values of latent heat transport to higher latitudes are found by assuming that the present transport in each hemisphere is the maximum post-Flood value. The minimum post-Flood value is assumed to have been 50%, and the best average post-Flood value is assumed to have been 75% of the present average. To determine the total amount of moisture available over a 500 year period for the regions north of 40°N latitude and south of 60°S latitude, the areas of the ocean from which the two major sources of moisture originated must be estimated. In making this calculation, it has been assumed that the size and configuration of the oceans were the same during the ice age as they are today.
Precipitation does not fall evenly over a large area. Areas close to major or minor storm tracks are especially favored. Also, the colder portion of a storm usually receives more precipitation (see Chapter 3). The ice sheets and non-glaciated land close to a storm track would receive much more precipitation than the mid-ocean areas. Based on these general meteorological considerations, estimates were made for the precipitation that fell on the ice sheets in both hemispheres. For a minimum estimate, a uniform distribution of precipitation over land and ocean was chosen. For the maximum, precipitation over land three times as great as over the ocean was selected. A median-estimated precipitation, twice as great over land as over the ocean, is probably the best post-Flood value. These estimates must be recognized as having a high degree of uncertainty.
Considering the extremes of all the variables discussed in this section, the average ice depth over the Northern Hemisphere was found to range from a minimum of 515 meters to a maximum of 906 meters. A depth of about 700 meters is considered the best estimate. The average ice age accumulation rate for the land north of 40°N was estimated to be 1.4 meters/year, which is at least three times the present average (Trewartha and Horn, 1980, back cover). For Antarctica, the estimated ice depth varied from a minimum of 726 meters to a maximum of 1,673 meters. The best estimate is about 1,200 meters, with an average annual accumulation of 2.4 meters/year, which is about an order of magnitude greater than the modern average. The spread in the minimum and maximum ice depths for each hemisphere is not too high, considering the uncertainties in the estimates. The assumptions used in the estimates of available moisture and of the precipitation distribution between land and ocean are not crucial to the main conclusion-that the ice sheets were relatively thin, on the average.
Ancient Ice Sheet Thickness The uniformitarian estimates of ice thickness are significantly larger than the average depths calculated here. They are about 150% larger for the Northern Hemisphere, and 50% larger for Antarctica. Is there any objective basis for the uniformitarian estimates, and are the values estimated for the post-Flood ice age hopelessly in error? We shall examine the methods for estimating ice depth, to check whether the uniformitarian values have a firm basis, and whether there is evidence to support thinner values.
Four main methods have been used to estimate past ice-sheet thickness: 1) analogy and theory, 2) the height of nunataks and of lateral features, like moraines, 3) the maximum lowering of sea level, and 4) the amount of isostatic rebound. This section will focus mainly on the Laurentide ice sheet, which is the largest of the past ice sheets. We assume the same, or similar, arguments hold for the other ancient ice sheets in Europe and northwest North America. The above methods utilize a limited amount of hard evidence, and circular reasoning is commonly employed. Paleoclimatologists Ericson and Wollin (1967, p. 136) admit that past ice-sheet thicknesses are really guesses: “the estimates vary, because one can only guess how thick the ice sheets were....” The first, and perhaps the most widely used method, is analogy and theory. Uniformitarian scientists have an overabundance of time for ice sheets to develop, and little factual data. In reference to the Laurentide ice sheet, Bloom (1971, p. 367) states: “Unfortunately, few facts about its thickness are known.... In the absence of direct measurements about the thickness of the Laurentide ice sheet, we must turn to analogy and theory.” Andrews (1982, p. 12) corroborates Bloom:
There have been several reconstructions of various Pleistocene ice sheets based essentially on glaciological theory. These have relied implicitly or explicitly on the analog premise that the appearance of the former ice sheets was not unlike that of the Greenland or Antarctic ice sheets today. This premise may not be valid.
Consequently, most ice age investigators assume the ice sheets grew to the size of the Antarctic and Greenland ice sheets. Using the Antarctic ice sheet for analogy, investigators theorize the depth, basal shear stress, movement, surface slope, and other properties for the Laurentide ice sheet. This is, essentially, how Denton and Hughes (1981) arrived at their reconstruction of the past ice sheets at maximum glaciation (Andrews, 1982). In the Denton and Hughes reconstruction, the Laurentide ice sheet had one huge center over Hudson Bay, from which ice flowed outward in all directions, all the way to the margins. From an atmospheric science viewpoint, this seems theoretically impossible (see Chapter 1).
Although field evidence is used to support the Denton and Hughes reconstruction, a large body of field evidence from the Arctic contradicts the single ice-dome model. The Denton and Hughes reconstruction “... is radically different from the interpretation of most field workers...” (Andrews and Funder, 1985, p. 2). Andrews and Miller (1985, p. 361) write: Of course we can get into an argument about the interpretation of “firm stratigraphic evidence” because in our reading of this paper it is interesting how much of the field evidence from arctic arms is discounted in favor of a hypothesis developed largely on the basis of a reasonable model as to how the Antarctica Ice Sheet functions.
What is the field evidence against the single-dome model? The field evidence consists of the direction of striae, drumlins, flutes, and roches moutonnées, and the provenance of distinctive erratics (Andrews, 1982). These data mediate against the former existence of one large ice center, but implies two or more smaller centers, or ice domes, within the ice sheet. The first dome was over Labrador-Ungava, and the second over Keewatin. Other postulated domes are located south of Hudson Bay, over the Baffin Island-Foxe Basin area, and over the Queen Elizabeth Islands (Figure 5.1). However, this evidence has been dismissed as the result of late glacial thinning of the ice sheet, after the single dome melted down to multiple domes (Andrews, 1982, p. 10).
Although much of this evidence does represent late glacial ice movement, one component, especially, represents long-term glacial motion. From an analysis of distinctive glacial erratics on the surface, and, most significantly, below the surface in Keewatin, Shilts et al. (1979) state that the Keewatin ice divide has always existed, and that ice never flowed westward from a large ice center in Hudson Bay into Keewatin (see Figure 1.3 for locations). In addition, Hillaire-Marcel et al. (1980) and Shilts (1980) contend that the ice-flow direction, on the east side of Hudson Bay, was always from the east. This evidence is strongly against the single-domed theory. The above is not the only evidence favoring a multi-domed Laurentide ice sheet. In models based on an Antarctic ice sheet analog, the ice sheet in eastern Canada would have been buttressed on the continental shelf, and would have partially drained by fast-moving ice streams, through Hudson Strait, the Gulf of St. Lawrence, and other large geographic troughs. Much evidence can be brought against this hypothesis. Andrews (1982, p. 16) states that the ice sheet was drained by a series of fjords, and did not terminate on the continental shelf, but farther inland. (This statement was made in the context of the last ice age in the multiple glaciation system, but, since practically all the glacial debris is attributed to the last ice age, Andrews’ statement likely applies to the entire multiple ice-age scenario. Besides, evidence will be presented, in Chapter 7 which strongly indicates that there was only one ice age.) Moreover, Hudson Strait does not provide evidence for parallel ice flow in the past, but, rather, for convergent flow from ice domes on Baffin Island, to the north, and on Labrador, to the south (Andrews et al., 1985). Hudson Strait and other large troughs, in the Canadian north, show very little evidence of deep, glacial erosion, as expected, from the single-domed theory, because these valleys are floored by “soft” pre-Quaternary sediments (Andrews, 1982, p. 25). In addition, large sections of Baffin Island and portions of the Ungava Peninsula exhibit little signs of glacial erosion (Andrews et al., 1985). Slight glacial erosion is a common observation, over most of Canada (Flint, 1971, p. 115; Eyles, 1983, p. 4). All this evidence favors the multi-domed, thinner, ice-sheet model, for eastern Canada. To add further support to the multi-domed model, Andrews et al. (1983) have published evidence that Hudson Bay was deglaciated, at times, during the last ice age. Although dated by amino acid ratios, the basis of their claim comes from water-laid sediments containing marine shells, and interlayered with glacial till (Andrews et al., 1984). This result, if correct, fully contradicts the single-domed-ice-sheet reconstruction model. The evidence for the multi-domed model implies that the Laurentide ice sheet was substantially thinner than required by the single-domed model (Andrews, 1982, p. 1). Occhietti (1983, p. 13) states: “These results change the concept of the Laurentide ice sheet radically. They imply, notably, a much smaller ice volume, and complex margins.” Field data from the interior area of the Laurentide ice sheet do support the thin post-Flood ice sheet model. Not only that, these data also indicate that the “reconstructions” of global ice volume, based on oxygen isotopes from deep-sea cores, have serious problems (Andrews, 1982, pp. 1, 2; Occhietti, 1983, pp. 14-16). The second method for deducing the thickness of ancient ice sheets is the height of nunataks, which are mountains or hills that protruded above the ice, and the height of lateral features, like moraines left from previous marginal ice lobes. As with the first method, problems and misinterpretations abound. Conclusions are often based on an assumed thick ice sheet, which, naturally, should have overridden the mountains of New England, and elsewhere. Recent research on most marginal areas, where lateral features and nunataks are found, mostly indicates that the periphery of the Laurentide ice sheet was thin. The periphery is defined as a rather narrow strip, about 300 or 400 kilometers wide, around the edge of the ice sheet.
It has been known for some time that the southwest margin of the Laurentide ice sheet was thin. This conclusion is based on nunataks and driftless areas in Montana, Alberta, and Saskatchewan (Lemke et al., 1965; Mathews, 1974, p. 39). The north-south ice sheet profile, at the margin, was about one-fifth as steep as would be expected, based on analogy with the Antarctic ice sheet. Ice flow indicators also show that the ice sheet was strongly influenced by topography (Clayton et al., 1985, p. 235). The marginal evidence does not apply only to deglaciation, when the ice sheets were thinning, it also covers the entire period of glaciation. In the north-central United States, the margin of the ice sheet repeatedly surged. These surges have left behind lateral moraines and other glacial features along previous ice lobes. The gentle slope of these lateral features indicates that the ice sheet must have been notably thin (Mathews, 1974; Clayton et al., 1985; Beget, 1986). The thin surges occurred during late glacial time, since they are on the surface, and mostly would have erased previous surface features. This surge pattern could have existed throughout the ice age; there is no evidence to the contrary. The driftless area, in Wisconsin, suggests that the ice sheet was thin throughout glaciation. Not only was the southwestern and south-central periphery thin, but recently evidence has been obtained which indicates that the northwest margin also was thin (Beget, 1987). The only marginal area showing evidence for a thick ice sheet is the southeast periphery. Several mountain ranges, or peaks, over 1,500 meters elevation, exist in this area. Most glaciologists believe the Laurentide ice sheet swept down from the north and completely covered these mountains. Evidence from glacial-flow indicators and distinctive erratics high up on the mountains are cited as proof. On the other hand, without dismissing the evidence of an earlier inundation of the mountains by an ice sheet, other investigators claim late glacial icecaps-not a glacial advance from the northwest-covered the mountains of New England and southeast Quebec. It is difficult to tell whether glacial features were produced by local icecaps, or by advance of the Laurentide ice sheet (Wagner, 1970, p. 2467). The evidence favoring the various hypotheses seems equivocal. For instance, Waitt and Davis (1988, p. 513) claim that far-traveled erratics, from the north, have been found within 45 meters of the summit of 1,605-meter-high Mount Katahdin, in central Maine. But Caldwell et al. (1985, p. 55), in referring to the Laurentide ice sheet, which is presumed to have covered the mountain, state: “No deposits were left by the thinning ice sheet on Mt. Katahdin until the ice surface lowered to the 760-m elevation.... Deposits related to the 760-m level occur in many places on the mountain.... “Furthermore, erratics are scarce. Caldwell et al. (1985) believe in local mountain glaciation on Mount Katahdin, after general deglaciation. They base their conclusion on the sharp features of cirques that would have been planed down by an overriding ice sheet.
Some of the presumed glacial features, in the mountains, actually can be due to debris flows and avalanches-probably a more common situation than most workers are willing to admit (Gerath et al., 1985, p. 27; Waitt and Davis, 1988). Debris-flows can mimic most glacial features. Even grooves and striations on stones and bedrock can be carved by debris flows (Schermerhorn, 1974, pp. 679-682). For example, the striations and grooves in a cirque in the northern White Mountains of New Hampshire, previously cited as proof of southward ice flow from the Laurentide ice sheet, have been determined to be formed by recent creep (Bradley, 1981, p. 323). Creep is the very slow, downslope movement of rocks and soil under the force of gravity.
Evidence that the Laurentide ice sheet may have been thin along the southeast periphery, is accumulating. Caldwell and Hansen (1986) conclude that, at maximum glaciation, the ice was only about 800 meters above sea level on Mt. Katahdin and other mountains in the same area. The nongranitic rocks on Mt. Katahdin, which have been assumed to be erratic, are claimed, by Caldwell and Hanson (1986), to be left-over, weathered country rock that the granite pluton intruded. Similar to Caldwell and Hanson (1986) and Caldwell et al. (1985), Wagner (1970, 1971), claims abundant evidence for low-level, valley glaciation in the Green Mountains of northern Vermont. Although he is not clear whether he believes the Green Mountains were previously overrun by the Laurentide ice sheet, he clearly rejects the local icecap concept on the higher mountains. The evidence he cites may be deposits left over from a thin Laurentide ice sheet that filled the valleys of the Green Mountains. To add to the confusion, illustrating the difficulty of interpreting glacial evidence in New England, Waitt and Davis (1988, pp. 501-513) dispute Wagner’s evidence for even low-altitude glaciation. The possibility of ice-advances, from local mountain icecaps, or from a thin general ice sheet in New England, is additionally supported by the abundant evidence for northward flow of ice into the St. Lawrence Valley (Lamarche, 1971; Borns et al., 1985). This evidence indicates that there was an ice divide in northern Maine. Some of the early geologists noticed the indicators of northward glacial movement, but, for many years, this evidence was overlooked, showing the strong psychological influence of models. Detailed surveys in the 1960s and 1970s even reported no indication of northward ice flow. Chauvin et al. (1985, p. 112) report: By the early 1960s the idea of Appalachian glaciers in Quebec was almost completely abandoned. Extensive mapping work in southern Quebec carried out by the Geological Survey of Canada firmly establishes that the last glacial flow in southern Quebec was toward the southeast and does not provide any evidence supporting the existence of Appalachian glaciers.... The now-abundant evidence for northward glacial movement is attributed to late glacial motion, which was opposite to the general trend. However, the late glacial motion could indicate the general trend, and give support to the thin, ice-sheet model.
It is supposed that a thinner ice sheet is more likely, because it is also supported by the evidence from other areas of the Laurentide ice sheet. Why should only the southeast periphery be thick? This area probably would have a relatively thick ice sheet in the post-Flood ice age, since it was close to a main storm track most of the time, and, also, close to the moisture source of the warm ocean. Regardless of whether the southeast periphery was thick or thin, sufficient evidence from the remainder of the Laurentide ice sheet supports the contention that this ice sheet was significantly thinner than specified by most uniformitarian estimates. The thin peripheral ice with a low slope-as much as one-fifth the surface slope of the Antarctica ice sheet-presents theoretical problems for models of past glacial flow. The investigators, actively working on this problem, have developed several possible solutions. Boulton and Jones (1979), and Boulton (1986), propose that the ice sheets mostly flowed on soft, easily deformable, pre-Quaternary sediments. These sediments, and the till from these sediments, were water-saturated, in many areas. Consequently, most of the basal movement of the ice sheets is presumed to have been by till deformation. Support for this view comes from the rapid motion of ice stream B, on the edge of the East Antarctic ice sheet. This ice stream has a low slope, and flows over water-saturated till (Blankenship et al., 1986; Alley et al., i986).
Further evidence of fast ice movement comes from Northern Hemisphere ice cores that come from drilling, which probably has penetrated ice that was formed during the ice age. Ice-age ice apparently is different from later ice. Ice-age ice possesses smaller ice crystals, and more microparticles from wind-blown dust (Reeh, 1985; Fisher and Koerner, 1986). Because of these properties, ice-age ice is estimated to have flowed three to four times faster than recently formed ice. The third method of inferring ancient ice thickness, is the reduction of sea level at maximum glaciation. The water held in the ice sheets is water lost from the ocean. But past sea levels are not known, and are mainly estimated from theory and analogy. The sea-level lowering is usually estimated from presumed ice-sheet volume. Flint (1971, pp. 317, 318), while discussing the problem of estimating past sea levels from the volume of glacial ice on land, candidly admits: A greater potential error lies in the estimation of average thicknesses and volumes of glaciers, particularly ice sheets that no longer exist. Thus far the profiles of such glaciers have been reconstructed by analogy with those of existing ice sheets, which for one reason or another may not be truly analogous.
Circular reasoning is obvious, when relating sea level lowering to ice sheet thickness, since each has been used to support the other, and each is just as unknown. The raw, sea-level data are equivocal. Blackwelder et al. (1979) claim that the sea level data near the maximum of the last ice age, are faulty. New measurements on “non-movable” ancient sea-level indicators revealed that the difference between modern sea level, and sea level at glacial maximum, was only about half that given by previous estimates. This implies “... that substantially less ice was present from 17,000 to 10,000 years B. P. [before present]” (Blackwelder et al., 1979, p. 620). Consequently, the high uniformitarian estimates of thickness based on sea level data are not reliable, and there is support for a thinner ice sheet. Sea-level data will be discussed further in Chapter 8. The fourth and last method of estimating past ice-sheet thickness is by glacial isostasy, which is also discussed in more detail in Chapter 8. Glacial isostasy is the depression, or rebounding of the earth’s crust, due to the addition or subtraction, respectively, of an ice sheet. The amount of crustal rebound that has occurred over previously glaciated terrain, plus the rebound remaining, could be used to calculate an ice-sheet thickness. Some authors have postulated a rebound of over 800 meters at the center of the Scandinavian ice sheet, probably based on models of presumed ice-sheet thickness (Mörner, 1980a). However, the highest shore line observed in Scandinavia Isaiah 290 meters above sea level (Eronen, 1983, p. 188), and around Hudson Bay, Isaiah 315 meters above sea level (Fairbridge, 1983, p. 7). The average rebound for the whole area covered by the Laurentide and Scandinavian ice sheets, is much less. If the ratio of uplift-to-thickness of the ice is one-to-three for a 3:1-density ratio of rock-to-ice, the uplift of Scandinavia and the Hudson Bay region suggests that the ice sheets were thin. Of course, the above conclusion depends upon the amount of isostatic rebound remaining. Some scientists believe they have found an index of the unrecovered isostatic rebound from gravity anomalies. The surface of the earth is characterized by slight positive and negative anomalies in the average force of gravity. This is mainly due to differences in crustal densities, but, also, is partially a result of distortion from the weight of ancient glaciers. Negative gravity anomalies that have been measured in the areas occupied by the former ice sheets, strongly indicate that mantle rock is flowing horizontally, towards these areas. Tide-gauge readings show that the formerly glaciated area is currently rising. For instance, the northern Gulf of Bothnia, in the Baltic Sea, near the presumed center of the Scandinavian ice sheet, is rising at 1 cm/yr. This figure is for only a small area; the average for all of Scandinavia is much less.
Based on gravity anomalies, values of 100 to 200 meters of potential, remaining crustal rebound are commonly stated. Unfortunately, gravity anomalies are difficult to measure, and the uplift, when not averaged over a large area, is rather chaotic, especially over Scandinavia (Mörner, 1980b). It is not certain that the gravity anomalies are due to unrecovered glacial rebound, either totally or partially (Walcott, 1980, p. 6). There are areas of positive gravity anomaly which should be subsiding, but actually are rising (Walcott, 1973, p. 20). Although the gravity anomalies probably do reflect unrecovered isostatic rebound, especially in view of current observations and indications of past changes, the magnitude of glacial uplift cannot be specified. The negative gravity anomalies, and the uplift that has occurred in Scandinavia and the Hudson Bay region, can be accounted for by a thin ice-sheet model, just as well as by thick models.