01. The Mystery of the Ice Age
CHAPTER 1
The Mystery of the Ice Age
Deposits resembling glacial debris cover the surface of northern North America, northern Europe, northwest Asia, and many mountainous areas of the world where glaciers do not exist today. In the tropics these features are found on the highest mountains, about one thousand meters below existing glaciers. This debris contains rocks of all sizes, chaotically mixed in a finer-grained matrix. End moraines similar to those associated with present-day mountain glaciers are abundant. Streamlined lens-shaped mounds, called drumlins, exist in large numbers. Large areas of North Dakota, Montana, and Saskatchewan are covered by parallel grooves with intervening ridges (Flint, 1971, pp. 104-106). These are best seen from the air. It is difficult to account for drumlins and fluted ground by any means other than glaciers.
Hard rock surfaces are polished, scratched, and grooved. Rock protuberances have one side smoothed and with parallel scratches, while the opposing side has a plucked or sheared surface. The large polished and scratched surfaces are called striated pavements, while the protuberances are called whaleback forms or roches moutonnées. These landforms indicate the passage of a deformable mass that was able to scratch and cut hard rock. There are many other landforms which also indicate glaciation (Sugden and John, 1976). The abundance of the various evidences for glaciation indicates that the mid and high latitudes of the Northern Hemisphere were once covered by large ice sheets.
Requirements for an Ice Age
How much climate change is required to account for these ice sheets? In other words, what are the requirements for an ice age? Scientists have debated these requirements for years, and have reached a reasonable consensus.
Some scientists once thought that colder winters were the main requirement for glaciation. However, winters are already cold enough for glaciation over most areas that were covered by the ancient ice sheets. In fact, winters are now too cold in many northern localities. Consider Siberia: The temperatures there average far below zero Fahrenheit in winter, but no glaciers exist. In most high latitude areas, cooler winter temperatures would not produce glaciation. In order to produce an ice sheet, winter snow must survive the summer and continue to accumulate year by year. Therefore, the crucial season for glaciation is summer. Summers must be drastically colder than today in order for the snow to survive. This is one of the reasons the snow doesn’t pile up in Siberia-summers are too warm.
Probably a more important ingredient is greater snowfall-enough snow must have fallen the previous winter to survive until the next winter. If the snowfall is light, the snow would melt, even if the summer were much cooler. Consequently, sufficient snow must also accompany the lower summer temperatures. Therefore, the requirements for an ice age are a combination of cooler summers and greater snowfall than today (Fletcher, 1968, p. 93).
Inadequacy of a Uniformitarian Ice Age
Now that we have stated the requirements for an ice age, how much summer cooling and annual snowfall are required? Actually, the amount is not exactly known. Only a modest summer cooling and snowfall increase have been suggested by most paleoclimatologists. A few popular science writers have even made an ice age seem so easy that the next one is due relatively soon (Calder, 1974)! A summer cooling over northern Canada of only four to six degrees Centigrade, together with the current precipitation, had been assumed to be adequate (Williams, 1979, p. 445). However, this threshold was never rigorously tested. While the weak astronomical theory of the ice age has been supposedly “confirmed,” recent research indicates that much more climate change was needed to glacierize northern North America than had been previously thought (Loewe, 1971; Williams, 1979). A 6°C summer temperature drop is not enough. Loewe (1971) likely was the first scientist to point out the need for a greater climate change. He summarizes: “The origin of the North American ice sheet raises some difficult questions” (Loewe, 1971, p. 332). Although Loewe specifically was referring to the North American ice sheet, similar difficulties would be encountered for other ancient ice sheets, such as the Scandinavian ice sheet in northern Europe and the ice cap that covered the Alps. The focus of this chapter, however, will be primarily on the Laurentide ice sheet of northeast and northcentral North America. At least two centers of ice sheet growth in northern Canada are assumed by most workers: 1) Keewatin, northwest of Hudson Bay; and 2) Labrador-Ungava, east of Hudson Bay (Figure 1.1). Winters in these areas are now very cold, in fact too cold for significant snowfall. Summers, on the other hand, are relatively warm (except for coastal locations modified by cool water). Currently, the average June-to-September temperature is about 10°C (Loewe, 1971).
Keewatin and Labrador-Ungava Plateau with the median date for last snow cover of one inch or more for 20 springs in eastern Canada (Redrawn from Potter, 1965).
According to estimates by geophysicists, Hudson Bay likely was non-existent between ice ages (if there was more than one) due to isostatic rebound (Loewe, 1971, p. 333). Isostatic rebound is the uplift of the earth’s crust after the ice melts. When an ice sheet develops, the weight of the ice depresses the crust. Hudson Bay is currently rising, and since it is shallow, it may become dry land before the next supposed ice age. Since Hudson Bay causes a very pronounced regional summer cooling, summer temperatures in the area would have been significantly warmer before ice sheet formation in the standard uniformitarian model.
Current precipitation for the two Canadian ice centers is much different. Labrador-Ungava is wetter, with a yearly average precipitation of about 74 centimeters (29 inches) of water. Keewatin is very dry, a polar desert, with a yearly average of about 20 centimeters (eight inches). But less than half this precipitation actually falls as snow, most falling as summer rain (Loewe, 1971, pp. 339, 340). As a result, very little snow accumulates in Keewatin, and, “at present the summer temperatures are so high that the snow easily disappears” (Loewe, 1971, p. 339). Although snowfall is heavier in Labrador-Ungava, winter snow melts over both areas at about the same time, usually by June 15, except for the extreme north, which is not far behind. Figure 1.1 shows the average date of last snow cover of one inch or more for 20 spring seasons in eastern Canada (Potter, 1965, p. 39). A drop in summer temperature of 6°C would of course cause the winter snow to melt more slowly. This drop also would tend to increase the proportion of annual precipitation that falls as snow. However, the additional snowfall would be smaller than expected, because the upper air temperatures (which determine the snow level) would likely not cool as much as postulated for the ground surface, due to the atmospheric circulation (Williams, 1979, p. 448). The above tendency to increase the annual snowfall would be more than offset by another factor: The cooler the air, the less moisture it can hold. This factor is a serious problem for uniformitarian ice age theories that depend, more or less, on presently observed processes. Figure 1.2 graphs the relationship between air temperature and the water vapor carrying capacity at saturation (Byers, 1959, p. 161). At warm temperatures, the carrying capacity changes rapidly with temperature, but at temperatures below freezing, it changes very gradually. The relationship between air temperature and the water-vapor-carrying capacity is observed between the warm and cold seasons-summer storms drop much more precipitation than winter storms.
Graph of water vapor capacity at saturation (100% relative humidty) versus temperature. Note the 60% drop in capacity as temperature cools from 10°C to -2°C.
If a 6°C summer temperature drop is not enough to bring on glaciation, how much cooling is needed? A computer model that calculates the energy balance over a snow cover in northeast Canada has recently estimated the necessary spring and summer cooling (Williams, 1979). The model is realistic. It has a variable albedo (solar radiation reflectivity) for different snow characteristics and makes fairly realistic solar and infrared radiation estimates by taking into account the variations in cloud cover. The proportion of precipitation falling as snow is generously allowed to increase in direct proportion to spring and summer cooling. This computer model closely predicts the observed seasonal changes of the Decade Glacier, on Baffin Island.
One of the purposes of this computer model was to test the astronomical theory of the ice age. In doing so, the summer sunshine was decreased to the presumed solar radiation minimum at 116,000 years ago. Williams began the computer run with the average April-to-August temperature 6°C below normal, and then decreased this value by increments of 2°C until he reached 12°C below normal. Figure 1.3 shows the area permanently snow covered in Canada, after a reduction in average spring and summer temperature of 10 and 12°C. Williams (1979, p. 443) concludes, “...much more climatic change is required for extensive glacerization of either Keewatin or Labrador-Ungava than has been suggested, equivalent to a 10 to 12°C summer temperature decrease....” His model even overestimates perennial snowcover because of two simplifying assumptions.
Boundaries of permanent snow cover in northeast Canada for a 10°C and 12°C spring and summer temperature decrease (Redrawn from Williams, 1979) The scientific basis for this conclusion is that the melting of snow is controlled more by solar radiation, which is abundant at higher latitudes in summer, than by air temperature (Paterson, 1981, p. 313). Researchers have been focusing too much on the latter. At a temperature drop of 12°C, Figure 1.2 shows that the air would hold 60% less water vapor at saturation. This is a large decrease in moisture, which was not taken into account by Williams. Not only that, the above temperature decrease only accounts for a permanent snow cover in northeast Canada. To produce an “ice age,” the snow must accumulate year by year, change to ice, and advance down to 37°N latitude in the central United States. More summer cooling than 12°C is likely required. As a result of this temperature criterion, an ice age is extremely difficult to account for, especially when only present processes are allowed.
Possible Solutions to the Difficulty
Some scientists are aware of the magnitude of the difficulty, but most are not. Several possible solutions have been proposed:
One solution is a climate change caused by a modest spring and summer temperature drop that enhances moisture transport to northeast Canada (Barry, 1966; Crowley, 1984). The amount of winter snow would be increased, with a greater chance that the snow cover would remain through the summer. Perhaps this atmospheric circulation of more moist air could be initiated by a temperature drop of only 3°C. (This modest cooling was suggested at a time researchers believed the threshold temperature drop for glaciation was 6°C.) The main moisture source for this proposal was the North Atlantic.
Support for the above scenario was provided by Lamb and Woodruffe (1970, pp. 36, 37), who estimated a 150-300% increase in precipitation for the early ice age circulation. Ruddiman and McIntyre (1979) even calculated that sea surface temperatures in the North Atlantic were 1 to 2°C above normal for about the first half of ice sheet growth, based on geological evidence from deep-sea cores. Warmer sea surface temperatures would evaporate more water vapor and create a larger land-ocean temperature contrast during winter. This would cause the storm track to lie more parallel to the east coast (Ruddiman and McIntyre, 1979). More southerly and easterly winds from storms would then transport more moisture from the North Atlantic to the growing Laurentide ice sheet (Figure 1.4).
Presumed storm tracks caused by a 1-2C warmer surface termperature in the North Atlantic. Stippled area glaciated (redrawn from Ruddiman and McIntyre, 1979).
Although this proposal seems plausible, a closer examination reveals serious scientific evidence against it. Referring to the more moist Labrador-Ungava, Loewe (1971, p. 338) states: On the other hand, it is not easy to see how a substantial rise of total, or a shift to winter, precipitation can be reconciled with the smaller capacity of the cooler air to hold water vapor. It is also doubtful whether a simultaneous change in general circulation would be able to provide the necessary snowfall.
Lamb and Woodruffe (1970) based their precipitation estimates on extreme months in the current climate that came closest to the “assumed” pattern at glacial onset. The atmosphere is constantly changing, and it is doubtful if an extreme pattern would last very long, not to mention the thousands of years envisioned for the building of an ice sheet. Barry et al. (1971, p. 417) questioned Lamb and Woodruffe’s use of extreme months (which in actuality caused only twice the normal precipitation and not up to three times the normal): “It is doubtful to what extent an extreme circulation pattern may persist for a full season or even more so for a long time interval.” So the exaggerated conclusion of Lamb and Woodruffe is clearly not justified scientifically. The second claim of warmer sea surface temperatures for the first half of a uniformitarian ice age is also contrary to modern observations. How only a 1 to 2°C warmer sea-surface temperature could significantly increase atmospheric moisture is difficult to understand. Actually, observations indicate that sea surface temperatures in the North Atlantic would cool rapidly, due to a colder climate in eastern Canada. In an extensive analysis of sea-surface temperatures for 120 years, Folland and Kates (1984) correlated below-normal, sea-surface temperatures for the Northern Hemisphere with below-normal air temperature, mainly from land. Sea-surface temperature only lagged air temperature by about 15 years, so the cooling response of the ocean is rapid. The relationship is most pronounced in the North Atlantic, where the data are most complete.
Below-normal temperatures in eastern Canada are caused by a little cooler atmospheric circulation (van Loon and Williams, 1976 a, b; Williams and van Loon, 1976). Cooler air blowing off the land over the adjacent North Atlantic Ocean cools the ocean surface. Barry et al. (1975, p. 980) state: “...this evidence suggests that it may be difficult to sustain high sea surface temperatures during the initial phase of a glacial period.”
Ruddiman and McIntyre (1979) claim the Labrador Sea would be ice-free for the first half of glacial buildup. However, modern observations indicate that this would not happen any more than would the proposed warmer sea-surface temperature discussed above. Sea ice on Hudson Bay, the Davis Strait, and the Labrador Sea is much more extensive when air temperatures are cooler than average (Herman and Johnson, 1978; Johnson, 1980; Catchpole and Faurer, 1983). Ledley (1984, p. 596) states that sea ice is highly correlated to changes in air temperature, with the following implication: “The extent of sea ice controls the availability of moisture for snowfall. As the sea ice extends farther south during an ice age it caps off the oceans and thus reduces the available moisture.” In other words, cooler temperatures will cause more sea ice in the moisture-source regions of Labrador, Ungava, and Keewatin, resulting in further drying. The response is almost immediate (Monastersky, 1987), not thousands of years, as suggested by Ruddiman and McIntyre. Sea ice also reinforces the atmospheric cooling, because of its much higher solar reflectivity than water and its barrier to the escape of the ocean’s heat and moisture. A second possible solution to the difficulty of a uniformitarian ice age is to propose one extreme year of high snowfall in Canada that was caused by a brief change in the general circulation of the atmosphere. Hopefully, this anomaly, combined with a solar radiation minimum (as proposed in the astronomical theory of the ice age), could cause the snow to persist through the summer and start an ice age. Williams (1978) showed that above-normal snowfall in September caused a modest temperature decrease and probably a snow increase in October, in northern Canada. However, in the years he analyzed, it was difficult to tell whether the circulation caused the cooler, wetter Octobers instead of the heavy snowfall the previous month. Besides, only in autumn, the season with the highest snowfall in northern Canada, do such anomalies occur. Early fall is still relatively warm and the air contains more moisture than in winter. The storm tracks are still relatively far north. But once winter arrives, the climatic drying sets in and the storm track is shifted southward, due to very cold temperatures. And even if winters were warmer in northern Canada, winter snowfall could not significantly increase. Figure 1.2 shows that at the winter temperatures of northern and eastern Canada, the air could not hold much more water vapor at warmer temperatures. The largest changes in water-vapor-carrying capacity are at the warmer temperatures during summer, not at the cold temperatures characteristic of Canadian winters. The deeper autumn snow depth cited in Williams’ 1978 investigation was not correlated with cooler summer temperatures. Although late summer and early autumn snowfall was heavy, the annual snowfall during the years of his analysis was only a little above average. For instance, in 1972, heavy, late-summer snowfall resulted in a snowpack only 20% above normal by the end of December (Williams, 1975, p. 289). Assuming this trend for the remainder of the cold season, a 20% increase in snow depth is not significant for glaciation.
Williams (1979) modified his modeling of summer cooling over northern Canada by doubling the cold season snowfall. Even with a 10 to 12°C summer cooling, doubling the snowfall produced only a moderate increase in the perennial summer snow cover. Figure 1.3 has been drawn, taking into account twice the normal snowfall. Williams (1979, p. 443) concludes: “...increased winter snow accumulation (the maximum observed at each station) does not greatly increase the area of perennial snow cover, nor does the possible effect of unrecovered glacioisostatic rebound....”
Just for the sake of argument, let us suppose that some large atmospheric circulation anomaly, by chance, dropped enough snow to last an entire summer, in northeast Canada. Would this start an ice age? No, it wouldn’t even be close, because of many other variables that come into play: First, the snow cover and the cooler temperatures accompanying the snow cover would generate less warm-season snow, due to the drier air (Figure 1.2). Therefore, by the time the next spring rolled around, the snow cover very likely would be below normal and would easily melt during the second summer. Even if the snow depth was above normal, Williams (1979) showed that this would not be significant, unless, of course, the snow was five or more times greater.
Hand in hand with the drying tendency from the cooler air, the atmospheric circulation would become unfavorable for an increased buildup of snow. An extensive snow cover has a tendency to cause an upper-level, low pressure system or an upper trough, and a lower-level, high pressure system or anticyclone. This pattern is common over northeast Canada, in winter, but a cooler summer, with a snow cover, would continue this pattern into the summer. Williams (1979, p. 444) states:
Because incident solar radiation is mostly reflected from a snow surface, the air above an extensive snow cover is colder, and atmospheric pressure decreases more with altitude in the colder air....This tends to create an upper-air “cold trough” above an extensive snow cover...
Ruddiman and McIntyre (1979, p. 173) also recognize the problem with expanding an ice sheet (which would be similar to that of expanding an area of snow cover): But the growth of these extensive bodies of ice also implies an expansion of the polar anticyclone normally positioned over ice cover in high latitudes of the Northern Hemisphere. This expansion of dry cold air would reinforce the normal high-Arctic aridity and slow or stop the rapid growth of ice sheets unless opposed by other parts of the climatic system.... The “other parts of the climatic system” are the 1 to 2°C warmer temperatures for the North Atlantic Ocean, which they believe occurred during the first half of glaciation. The significance of an upper trough and a low-level anticyclone is that the storm tracks would be suppressed further south and east, resulting in less precipitation for the area. This is observed during the seasonal change from summer to winter in today’s climate. As the temperatures in the north cool from late summer to early winter, the average storm track is progressively displaced southward, with the higher latitudes becoming colder and drier. Tarling’s (1978, p. 14) statement concerning ice-age theories also applies to the above hypothetic case of increasing an area of year-round snow cover:
Apart from the difficulty in isolating different interactive causes, the evidence is always complicated by the strong climatic influences exerted by the ice sheets themselves, as these locally increase the Earth’s albedo and create their own atmospheric-pressure zones, with resultant equatorial displacement of pre-existing climatic belts.
Therefore, the atmosphere and ocean would respond to a summer snow cover, in northeastern Canada, with a tendency to cause drier conditions. And with less precipitation, the snow would easily melt, the next summer, in our hypothetical example. A third possible solution out of the difficulty of a uniformitarian ice age is to postulate an increase in cloudiness. Increased clouds can cause cooler temperatures, especially in summer. This topic will be discussed further in Chapter 3. However, summers are presently very cloudy in northeast Canada, particularly in Labrador-Ungava (Williams, 1979, p. 454). So, invoking increased cloudiness does not help the problem. In summary, the proposed solutions cannot provide the sustained cooling and heavy snow to glaciate northeastern North America under essentially uniformitarian conditions. Modem research shows that much more summer cooling than previously thought is a prerequisite. Even doubling the normal snowfall is not sufficient. A Multitude of Theories A large number of theories have been exposited to account for the ice age. As of 1968, over 60 theories had been proposed (Eriksson, 1968, p. 68), but all of these theories have serious difficulties. Charlesworth (1957, p. 1532), who has extensively researched the ice age, stated, over 30 years ago: “Pleistocene phenomena have produced an absolute riot of theories ranging ‘from the remotely possible to the mutually contradictory and the palpably inadequate.’” The Pleistocene period of geological time is generally the time of the ice age starting about 2,000,000 years ago, and ending about 10,000 years ago. A time span of 10,000 years ago to the present is called the Holocene period, and both eras, combined, are called the Quaternary period. (References to conventional geological time are used for communicative purposes, only, and are not to be construed as indicating belief in the evolutionary/uniformitarian time scale.) Brian John (1979, p. 57), years later, reminiscing on Charlesworth’s comment, says the problem has not improved: “Things have become even more confusing since then....”
Ice-age theories are usually classed as either extra-terrestrial or terrestrial. Several of those, which at one time or another have been popular, will be briefly discussed (Imbrie and Imbrie, 1979, pp. 61-68; Tarling, 1978. pp. 14-18).
One obvious extra-terrestrial possibility is a decrease in solar output. Since the sun empowers the climate, researchers have been trying to correlate small fluctuations of the solar “constant” with climatic variables. For instance, atmospheric scientists have developed climate simulation models that are tuned to the present climate. They then decrease or increase solar radiation one to several percent and examine how the model specified changes in climate. This modeling has yielded mixed results, at best. Some mechanism for relatively large changes in solar radiation must be found for the theory to be viable. Furthermore, the solar variations during the past must be known. The theory suffers from a common problem-it can never be proved or disproved, scientifically.
Another extra-terrestrial theory, the galactic dust cloud theory, states that ice ages were caused when the earth passed through cosmic dust, blocking some of the solar radiation. This theory suffers from the inability of astronomers to map areas of cosmic dust and to predict when the earth would have passed through these clouds.
There are a large number of terrestrial theories. One of the most popular is a decrease in the atmospheric gas, Col 2:1-23. This gas is transparent to solar radiation but strongly absorbs terrestrial, or infrared radiation at certain wavelengths. Although, the concentration of Col 2:1-23 in the atmosphere is very small, a decrease in Col 2:1-23 would cause a temperature drop, the opposite of the greenhouse effect. However, some scientists cannot see why or how this would happen, or why Col 2:1-23 would be lower during ice ages.
Some theories have proposed cooling from volcanic dust. Although volcanic dust and aerosols will cause cooler temperatures, the volcanic activity would have to be much larger than today (Bray, 1976). Each ice age is believed to have lasted 100,000 years; therefore the frequency of very large volcanic eruptions would have to be high and continue for a long time. Since there is no evidence of substantial volcanic activity throughout the 2,000,000 years of the Pleistocene period, the volcanic dust theory is not taken seriously.
Two well-known scientists, several decades ago, proposed that an ice-free Arctic Ocean would greatly increase the moisture over higher-latitude continental areas. A permanent snow and ice cover over land would be formed, increasing the solar reflectivity, or albedo, of the surface (Donn and Ewing, 1968). When temperatures cooled far enough the Arctic Ocean would freeze, and the continental ice sheets would dissipate due to a lack of sufficient atmospheric moisture to maintain them. Subsequently, the climate warms and the ice on the Arctic Ocean melts, and the process begins anew. However, researchers believe that the Arctic Sea icecap has been in place for the past several million years. Thus, this theory is rejected. At least Donn and Ewing recognized the importance for the ice sheets of a moisture source, which is lacking in practically all other theories. An ingenious theory for glaciation proposes that the West Antarctic ice sheet, which is grounded well below sea level, periodically surges into the ocean due to basal decoupling. This is even a concern today among a few scientists (Denton and Hughes, 1981). This surge would result in increased solar reflectivity over a greater area of the Southern Hemisphere, and may initiate an ice age. There are a number of problems with this theory, one of which is how a small increase in the average reflectivity in the Southern Hemisphere would trigger an ice age in the Northern Hemisphere. Researchers also find no evidence for the resulting catastrophic rise in sea level.
Despite the crude modeling effort of Donn and Shaw (1977), continental drift cannot be invoked to initiate Quaternary glaciation, because the continents have been supposedly in nearly their current configuration since well before the start of the ice age. Then why didn’t the ice ages begin much sooner, in the geological time scale? An obvious solution to the problem has been to propose that mountain building in the late Cenozoic initiated icecaps that coalesced to form large ice sheets. Further climatic cooling would cause lower plateau areas to become ice covered (Flint, 1971, pp. 808, 809). The problem with this theory is that many mountainous areas are not now draped in snow and ice, and the ancient ice sheet in northeast Canada developed at low altitudes. Even during the ice age when local mountain icecaps developed, for instance, on the Alps and the Tibetan Plateau, the ice did not expand outside the mountainous areas. Other putative mountain-building episodes in the past did not cause ice ages. Even with the high mountains of today, ice ages are extremely difficult to account for (Williams, 1979).
Mountain building has recently been called upon to aid other ice-age mechanisms. Kerr (1989a) reports the modeling effort of William Ruddiman and John Kutzbach, in which the Rocky and Himalayan Mountains are progressively raised, and the resultant climate is compared to the present. When the mountain reach nearly their present height, the astronomical theory (discussed in the next section) is then able to initiate glacial/interglacial oscillations. The mountains deflect the usual west-to-east tropospheric wind currents. Northwesterly winds cause colder air to flow further south in eastern North America. On the other hand, when the upper winds switch to the southwest, warm air is transported farther north. Unfortunately, the model, so far, is able to generate only a very modest 2°C average summer cooling, due to the total uplift of the mountains.
Recent uplift of the mountains is also suggested as a mechanism to decrease Col 2:1-23 in the atmosphere, and thus boost the astronomical theory (Horgan, 1988). The mechanism is rather circuitous, and very hypothetical. Higher topography results in greater precipitation and erosion rates. As a result, fresh rocks are more easily exposed. Increased weathering of the exposed rocks produces greater quantities of positively charged ions of sodium, potassium, magnesium, calcium, and others, that are carried to the oceans by streams. The oceans become more alkaline which decreases the amount of carbon dioxide in the water. Since the ocean-surface layer and the atmosphere are in near Col 2:1-23 equilibrium, the atmosphere loses Col 2:1-23. As will be shown in Chapter 3, even if the above mechanism can significantly decrease atmospheric Col 2:1-23, the resulting temperature drop would be very small. The stochastic ice-age theory has been recently proposed (Hasselmann, 1976). This idea has become popular with some scientists, and is backed up by sophisticated mathematical arguments. It states that since there are random variations in climate on short time scales, there should be large fluctuations inherent in the climatic system over long periods of time. This theory is difficult to test, and essentially relies on random chance.
Revival of the Astronomical Theory The last theory that will be discussed is the astronomical theory of the ice ages, commonly called the Milankovitch theory, or mechanism. Although this theory does not state how the series of ice ages began, it offers a solution to glacial/interglacial fluctuations. Its popularity has grown immensely during the past 20 years, and there is confidence that the mystery of the ice age has been solved by it (Imbrie and Imbrie, 1979). Therefore, this theory will be discussed in more detail. The astronomical theory is based upon slight changes in the intensity of sunlight reaching the earth, which are caused by periodic differences in the earth’s orbit around the sun. The gravitational pull of the moon and planets causes three orbital variations: 1) slight changes in the eccentricity of earth’s orbit, 2) small variations in the tilt of the earth’s axis with the plane of the ecliptic, and 3) the precession of the equinoxes. Only the first variation will be discussed, since it is considered the main cause of glacial/interglacial oscillations. The earth’s orbit is not a perfect circle, but is slightly elliptic. The orbit changes from nearly circular to slightly elliptic, and back to nearly circular about every 100,000 years. A measure of this change is the eccentricity of the orbit. The eccentricity varies from zero for a perfect circle, to one for an orbit completely flattened to a line. Figure 1.5 (exaggerated to show the slight difference) illustrates the present earth’s eccentricity of 0.017, which results in the earth being 3,000,000 miles closer to the sun in Northern Hemisphere winter, than in summer. Less sunshine at higher latitudes in summer supposedly causes cooler temperatures that trigger an ice sheet. Since the variations are periodic, ice ages, separated by interglacials at regular intervals of 100,000 years, are postulated.
Present eccentricity of Earth’s orbit (flattened to illustrate the phenomenon). Seasons are in reference to the Northern Hemisphere. The theory was first proposed in the late 1800s and helped persuade scientists to believe in multiple ice ages, as opposed to just one. According to the theory at that time, the last ice age ended about 70,000 years ago. Scientific evidence was marshaled to prove this date. The astronomical theory was not well developed until the 1920s and 1930s, when Milutin Milankovitch worked out many of the details. According to the revised astronomical theory, the ice age peaked about 18,000 years ago, and data now “prove” this date. The theory was later discarded, for good reasons, during the 1950s and 1960s.
Before the Milankovitch theory once again became popular, in the 1970s, West (1968, p. 213) stated how the theory could be tested: “If there was a correlation between a Milankovitch-type curve and the geological evidence for climatic change, then the point might be decided.” In the 1970s, the astronomical theory was revived, due to the influence and persistence of several eminent scientists. The theory was “proved,” by matching the earth’s orbital variations with slight differences in the oxygen isotopic composition of small planktonic shells that have settled on the bottom of the ocean (Hays, Imbrie, and Shackleton, 1976). As many as 20 or 30 ice ages, separated by complete melting, are now assumed to have developed in succession, over the past several million years (Kennett, 1982, p. 747).
Many serious problems, which have been overlooked in the enthusiasm for this theory, are discussed elsewhere (Oard, 1984 a, b, 1985). A few brief comments are in order. The changes in summer sunshine at higher latitudes, postulated by the theory are actually small-too small to cause the dramatic changes needed for an ice age. The “proof” is only a statistical correlation with geological data. The theory does not tell how each ice sheet actually developed. Furthermore, heating at higher latitudes depends only partially on sunshine. Northward transport of heat by the atmosphere and oceans is also important, but is mostly neglected by proponents of the theory. This transport would lessen the cooling at higher latitudes, caused by reduced sunshine. Meteorologists have known the weaknesses of the theory for a long time, and these weaknesses contributed to its earlier downfall. Famous astronomer, Fred Hoyle (1981, p. 77), expresses his sentiments with the following words:
If I were to assert that a glacial condition could be induced in a room liberally supplied during winter with charged night-storage heaters simply by taking an ice cube into the room, the proposition would be no more unlikely than the Milankovitch theory. The night-storage heaters are the other processes that supply heat to higher latitudes in winter, and the ice cube represents the slight cooling of the astronomical theory.
Data from the ocean bottom supposedly show that ice ages repeat every 100,000 years. This period matches only the eccentricity variation of the earth’s orbit (Figure 1.5). However, this particular orbital variation is the smallest of the three variations by far, changing the solar radiation on earth, at most, 0.17% (Fong, 1982, p. 4). Scientists are greatly perplexed, and are seeking a secondary mechanism to boost the weak orbital variations. The ocean bottom data should be examined more critically. Many poorly known processes can influence the oxygen isotopes in plankton shells (Berger and Gardner, 1975). For instance, the temperature of the water in the past, when the shell formed, must be known within one or two degrees. The planktonic animals often live in the surface layer of the ocean, which exhibits seasonal changes of ten degrees or more at mid and high latitude. Adding to the confusion, the planktonic animals change depths at times. Since the ocean cools with depth, especially at lower latitudes just below the surface, a large unknown source of error is introduced. For meaningful oxygen isotope measurements from the ocean bottom, the sediment must lie undisturbed since deposition. However, shells lying on the bottom of the ocean are commonly subject to erosion by ocean currents, mixing by abundant bottom-feeding worms, and dissolution of the calcium carbonate. Dissolution can even change the oxygen-isotope ratio by dissolving thinner shells that are isotopically lighter (Erez, 1979; Bonneau et al., 1980). Thus, the astronomical theory does not escape the serious scientific difficulties encountered by all other theories, in explaining the ice age.
What about Climate Simulations?
Although a uniformitarian ice age seems meteorologically impossible, and proposed solutions to the problem are inadequate, atmospheric climate simulations have recently shown that the small changes in solar radiation proposed by the astronomical theory supposedly do cause ice ages. Since climate simulations have been analyzed elsewhere (Oard, 1984a), they will only be summarized here. Interestingly, the small changes in solar radiation, presumed to have generated ice ages, have been considered adequate only after the astronomical theory was “proven” by correlations with deep-sea cores. The desired results are actually the consequence of radiation-sensitive initial conditions, such as ice sheets already in place, uncertain values for input variables, and by inexact “parameterizations” in the models. A parameterization is simply the statistical representation of a poorly understood variable, in terms of another, better-known variable (Schneider, 1984, pp. 853, 854). For instance, the complex process of energy loss to space by infrared radiation is typically parameterized in terms of surface temperature, by comparing present radiation with seasonal and latitudinal variations in temperature.
One variable that is favorable for the development of an ice age in these models, is the albedo of snow, which is too high. It is usually treated as a constant with a value of about 0.7 (Suarez and Held, 1979, p. 4829). This is a good value for dry, winter snow, but is much too high for melting snow, and is especially high for exposed, glacial ice. The albedo of melting snow drops to 0.4 in about two weeks (U. S. Army Corp of Engineers, 1956, Plate 5-2, Figure 4). The albedo of ice ranges from 0.2 to 0.4 (Paterson, 1981, p. 305). In addition, there are positive feedback mechanisms that aid the melting of snow or ice. These will be discussed further in Chapter 6. They can lower the albedo below the 0.2-to-0.4 range.
Ice-age models have generally used unreasonably high values of precipitation. A common value seems to be 1.2 meters/year for northeastern North America (Birchfield and Others, 1981, p. 130; Hyde and Peltier, 1985, p. 2179). This precipitation rate is almost twice the average value for the Labrador-Ungava plateau, and is six times too high for Keewatin. With such extremely high values for snowfall and snow albedo, it should not come as a surprise that these models predict an ice age due to small changes in radiation that are correlated to the Milankovitch oscillations. One model even predicts that we should be in an ice age at present (Suarez and Held, 1976, 1979)!
There is a very basic problem with all of these climate models. Although causal mechanisms may be tested, even the most sophisticated general circulation model is still too crude for anything but qualitative results. And these results may be severely distorted by some of the considerations noted above. In other words, we still do not have sufficient knowledge to justify confidence in climatic reconstructions. Bryson (1985, p. 275) states: “In general, physical models of the atmosphere-hydrosphere-cryosphere system are not yet sufficiently advanced, nor is our knowledge of the required inputs, to allow for climatic reconstruction, per se.” A New Paradigm Needed On one hand, extensive glacial deposits cover the surface of mid and high latitude continents, providing undeniable evidence for extensive past glaciation. On the other hand, atmospheric science and related disciplines strongly suggest that an ice age, which depends upon present processes, (uniformitarianism) is nearly impossible. The only other possible solution is with a catastrophic mechanism. Such a mechanism is, by definition, dramatic, and out of the range of normal experience, but many scientists are now convinced that a catastrophic mechanism has much scientific support. The model presented in this monograph is based on the historicity of the Bible, especially in its account of Creation and the Genesis Flood. Both scientific and religious implications are involved. Most men of science, 150 to 200 years ago, accepted the first 11 chapters of Genesis as historically valid. Furthermore, these chapters have never been proved wrong. They have been arbitrarily rejected by a newer generation of scientists who preferred the theory of evolution and the uniformitarian principle (Gould, 1987). If the reader has difficulty accepting this starting point, at least evaluate the following presentation with an open mind. If one does not know both sides of an issue, he really cannot say he is truly educated on that issue. All possibilities should be examined carefully. The solution proposed is tantamount to a paradigm shift in glaciology and the historical sciences. A paradigm is, essentially, a supermodel, or a set of foundational principles, which determine what is science in a particular field (Kuhn, 1970). It also determines the way research is conducted, and what problems will be tackled. Paradigms have advantages for scientific research, but they rarely motivate scientists to examine the basis of the paradigm. Usually, data that don’t agree with a prevailing paradigm are overlooked, ignored, or forced to fit. Thomas Kuhn (1970, p. 24) writes:
Mopping-up operations are what engage most scientists throughout their careers. They constitute what I am here calling normal science. Closely examined, whether historically or in the contemporary laboratory, that enterprise seems an attempt to force nature into the performed and relatively inflexible box that the paradigm supplies. No part of the aim of normal science is to call forth new sorts of phenomena; indeed those that will not fit the box are often not seen at all.
Paradigms are engrained into scientists early in their careers, while they are students. Young scientists tend to accept theories, not on the evidence, but on the authority of the teacher or text, since they usually do not know the alternatives or have the competence to make an independent choice (Kuhn, 1970, p. 80). A wrong paradigm, or, more commonly, one based on half truth, can dominate for a considerable time, especially if the paradigm supports the desires of prominent individuals. A large number of significant anomalies must build up to invoke a crises sufficient to make scientists recognize new foundational principles. Thomas Kuhn gives examples from the physical sciences, his expertise, to demonstrate this. There are just as many examples in the biological sciences.
What about the facts? Aren’t they scientifically solid? Yes, if facts are observed, and “all” the observations are adequately described. This last point may not be appreciated. Unfortunately, it is common practice, in science, not to publish negative or inconclusive studies (Peterson, 1989). The facts on record may just be a biased sample. We must be careful, even when using “observational” data. The problem of understanding the ice age is not the “facts,” but the interpretation of those facts. There is rarely just one way to view a given body of data. “Philosophers of science have repeatedly demonstrated that more than one theoretical construction can always be placed upon a given collection of data” (Kuhn, 1970, p. 76). So there may be other frameworks within which to place a set of data. Before new ideas can be entertained and/or accepted, a revision of basic assumptions or a shift to a new paradigm may be needed. This shift is basically a different way of viewing the data. Kuhn (1970, p. 85) writes:
One perceptive historian, viewing a classic case of a science’s reorientation by paradigm change, recently described it as “picking up the other end of the stick,” a process that involves “handling the same bundle of data as before, but placing them in a new system of relations with one another by giving them a different framework.”
New paradigms practically always advance scientific knowledge, and the change comes about in a revolutionary manner. Such a change is needed in order to solve the mystery of the ice age. The uniformitarian paradigm in glaciology and historical sciences has been long dominant. Paradigms in historical science are much more difficult to change than those in the observational sciences, because it is almost impossible to objectively test the former. Because of the many anomalies that have cropped up, some scientists have already opted for a neo-catastrophism that allows local catastrophes. The data from geologic investigations are making strict uniformitarianism more and more difficult to believe (Ager, 1973). The theory of evolution-the basis for most of mainstream historical science is coming under-attack, even from non-Christian scientists and intellectuals (Himmelfarb, 1962; Macbeth, 1971; Zuckerman, 1971; Grassé, 1977; Fix, 1984; Denton, 1985). In view of this trend, we will take another look at the global Genesis Flood and see if it provides a basis for a catastrophic ice age. This will be a paradigm shift, but making use of previously available data. Parker (1980, p. 186), in a review of the popular book on the astronomical theory Ice Ages: Solving the Mystery states: The earlier chapters make mention of the conflict of religious belief and scientific theory. The authors would have done well to state that the conflict is one of basic principles, not of conflicting observations. The basis of all scientific research into the distant past is the principle of ‘uniformitarianism’, i.e. that the laws of nature have always been the same as they are now. Research could not proceed without such an assumption, and the results should be taken as true in so far as that assumption holds. Belief in God’s creative and other activities in the past is not intellectual suicide but the choice of a different set of basic principles. The solution to the ice age mystery requires a multidisciplinary endeavor. The author’s expertise is in atmospheric science, and that will generally be the focus of this book. However, information from related fields will be brought together to develop a reasonable synthesis.