The two sources of natural variability of the climate system discussed in the previous chapter, volcanic eruptions and air-sea interactions, affect the global surface temperature by at most a few tenths of a degree Celsius over a period of a year or two. In Section 1.3, we noted that the globally averaged temperature has increased by roughly .5[o]C over the past 130 years. Has the earth ever experienced bigger changes in temperature? The answer is unequivocally yes as summarized in Fig. 5.1.
Figure 5.1. Global temperature variations during: (top) past million years, (middle) past ten thousand years, and (bottom) past thousand years. The dashed line across each panel of the figure indicates the present globally averaged temperature of 15[o]C[*].
The conditions at the present time are found at the right edge of each of the panels. The panels are derived from paleoclimate (past climate) indicators that will be discussed in the next sub-section. Note that the temperature scale common to the upper two panels is narrower than that at the bottom panel. We see that the natural temperature variations experienced by the earth-atmosphere climate system become larger and larger as we view the climate system on longer and longer time scales.
The upper panel shows an estimate of the globally averaged surface temperature over the past million years. Notice that over this span of time, the temperature of the earth has rarely been higher than the present value. The particularly cold periods are glacial episodes, more commonly referred to as ice ages, in which a significant fraction of the land area poleward of 50[o]N is covered by ice while the intervening warmer periods are interglacial episodes in which the glacial coverage of the Northern Hemisphere wanes. The ice ages occur at irregular intervals spaced about 100,000 years apart. The last ice age began abruptly around 120,000 years ago and was preceded by one of the warmest interglacial periods of the past million years. This interglacial period is referred to as the Eemian interglacial. The last ice age reached its peak about 18,000 years ago and then ended abruptly 12,000 years ago We have been in an interglacial period for the last 10,000 years. This period is referred to as the Holocene. It is only a matter of time (at least from the viewpoint of paleoclimatology) before we enter another ice age[*].
The middle panel of Fig. 5.1 summarizes the changes in temperature observed over the past 11,000 years. The conditions at the extreme left edge of the figure reflect the final stages of the last ice age. Roughly 4,000 to 6,000 years ago, the earth-atmosphere climate system was characterized by warmer conditions than those at the present time. This period is referred to as the Holocene maximum and it represents the warmest period observed to date during the present interglacial.
The bottom panel examines the variations in temperature during the past thousand years. During the middle ages (1000AD-1300AD) the globally averaged temperature was a few tenths of a degree warmer than that observed now. After that point, the globally averaged temperature cooled below that found at the present time. This period is often referred to as the Little Ice Age, since glaciers in many regions of the world expanded signficantly during this period. The warmer conditions observed since the middle part of the 19th century may reflect natural variations in the climate system on time scales of a few hundred years.
What relevance does the behavior of the globally averaged temperature over thousand-year to million-year time scales have for the global warming problem? Most importantly, it shows that the earth-atmosphere climate system has undergone large temperature swings within a fairly well-defined temperature range. During the past million years, the earth's surface temperature has remained within 5[o]C of the present temperature.
Many authors of articles in the general literature have suggested that increased carbon dioxide emissions may lead to a runaway greenhouse effect that might result in the earth's atmosphere resembling that found around Venus. Venus has surface temperatures in excess of 460[o]C. Since Venus is closer to the sun than the earth is, it receives roughly twice as much solar radiation at the top of it's atmosphere. However, more significantly, the atmosphere of Venus is composed nearly entirely of carbon dioxide (96.5%) with only a few percent of nitrogen and other gases. Based on our simple reasoning in Chapter 2, we can think of the Venusian atmosphere as being totally opaque to infrared radiation and its effective radiating level to space would be found at the top of the atmosphere. That is why the surface temperature on Venus is so high. Based on the small observed climate variations shown in Fig. 5.1 and the small anthropogenic emission rates of carbon dioxide into the atmosphere, it would be highly unlikely that the earth's atmosphere will ever resemble the one found on Venus. This, of course, does not mean that our climate will be unaffected by rising concentrations of greenhouse gases.
In Section 1.1, we showed the observed variation in carbon dioxide over the past
Figure 5.2. Concentration of carbon dioxide in ppm over: the past 160,000 years (upper panel) and the past 250 years (lower panel). The dashed line in the upper panel indicates the present concentration of carbon dioxide. The heavy shading in the upper panel indicates that a range of values are possible.
40 years. We now show in Fig 5.2 estimates of its concentration over the past 250 and the past 160,000 years. These long term records of carbon dioxide concentration were determined from air trapped in the ice of Antarctica. Notice that the amount of carbon dioxide in the atmosphere during the past 160,000 years was substantially below that found at the present time. However, during the present interglacial and the Eemian interglacial (around 120,000 years ago), the concentrations of carbon dioxide are higher than those found during the intervening ice age. In other words, the concentration of carbon dioxide in the atmosphere and the globally averaged temperature of the earth are related to one another. What is difficult to determine from such paleoclimate records is whether the periods of high concentrations of carbon dioxide cause the high temperatures or vice versa. The bottom panel of Fig. 5.2 shows that, prior to the industrial age, the concentration of carbon dioxide was on the order of 280 ppm and it has risen sharply since then.
Fragmentary weather records are available for some regions of the globe for the past millenia or two. Records of major weather catastrophes (floods, droughts, etc.) in China extend back several thousand years. However, these records are of little use for determining the globally averaged temperature of the earth's surface.
A variety of proxy methods have been developed to determine long-term climate variations in temperature. These methods fall into two main cateogories:
* faunal and floral techniques. These methods rely on the naturally occurring recording systems inherent in the growth and decay of life on the planet. The annual growth of trees and the rate of growth of coral reefs in the oceans are examples of these naturally occurring recording devices.
* Sedimentological and stratigraphic techniques. These methods rely on analyses of the structure and chemical composition of soils and ice to infer climate information. The concentration of different forms of oxygen in the sediments at the bottom of the ocean or in the polar ice caps is the basis for one method for determining climate variations.
There is no paleoclimate method that provides all of the information required to determine the state of the global climate system. The panels in Fig. 5.1 are composites based on a variety of paleoclimate records, all of which are subject to considerable uncertainty. Each indicator reflects a complex, integrated response to a number of climate variables. For example, the rate of growth of trees as measured by the width of the annual growth (referred to as the width of the tree ring) responds not only to the temperature that the tree is exposed to but the amount of moisture available as well. Obviously, trees grow as a function of many factors unrelated to climate, such as changes in the chemical composition of the soil (for example, increasing acidity or akilinity). In addition, we must assume quite a bit to expect that the rate of growth of corals in the tropical oceans or the composition of ice in Antarctica is representative of the entire global climate system.
Nonetheless, these proxy paleoclimate indicators are the only means available to document the history of the earth-atmosphere climate system. There is another approach, however, to understand how the climate system might have worked in previous epochs. This entails using a climate model of the atmosphere such as the type described in Section 3.3. Instead of using the present conditions from which to begin the model simulations (such as the current locations of the polar ice caps and coastlines), the conditions appropriate for a previous period are used instead. These types of simulations provide detailed information on how past climates might have been constructed. However, there is limited paleoclimate data to verify whether they are accurately simulating the atmosphere of that era.
We will defer for the moment a complete explanation for why ice ages occur. Accept for now the following scenario for how ice ages come about:
Figure 5.3. A schematic of the ice-albedo feedback.
The above schematic shows a feedback loop by which increases in the coverage of snow and ice can lead to a colder climate that eventually leads to still greater coverage of snow and ice. This particular feedback is referred to as the ice-albedo feedback. It refers to the effect that additional snow and ice have on the albedo of the earth-atmosphere climate system. Since snow and ice can reflect appreciable amounts of solar radiation back to space, an increase of the area covered by snow and/or an increase in the amount of time during the year that the snow coverage lasts can reduce the amount of solar radiation absorbed globally by the earth's surface. We can use the same feedback loop to explain how ice ages might end. If snow coverage falls, then the albedo will decrease and the amount of solar radiation absorbed will increase. Then, the globally averaged temperature will increase, which results in less snow cover.
At the present time, ice sheets cover Antarctica and Greenland. The spatial coverage of these ice sheets in relation to the area of the globe is not particularly large. They are deep, though, on average several thousand meters thick. From the hydrologic cycle diagram in Section 3.2 it can be seen that 2% of the earth's water is locked up as ice in the ice sheets and glaciers of the world.
The advance and retreat of glaciers in the mountains of Europe have been well documented over the past millenia. During the Middle Ages, the Norse settled the coastal fringes of Greenland, which from its name was a somewhat more habitable locale at that time. However, the onset of the Little Ice Age with colder temperatures and advancing glaciers contributed eventually to the demise of these colonies. During the Little Ice Age, most of the glaciers of Europe expanded gradually down the mountain valleys, since the rate of accumulation of snow exceeded the seasonal melt during the Little Ice Age.
At the height of the last Ice Age around 18,000 years ago, the Antarctic ice cap was deeper than it is now. More importantly, ice covered nearly all of the land areas poleward of 50[o]N as shown in Fig. 5.4.
Figure 5.4. Ice sheets covered most of the land areas poleward of 50[o]N at the height of the last Ice Age.
There were also glaciers further south in the Himalaya Mountains, for example, and in the Uinta Mountains of Utah. Not only was the coverage larger 18,000 years ago, but the ice sheets were very thick. Most of Canada lay under 3,000 m of ice while it was nearly as thick over Europe and Russia.
Where did all of this ice come from? Obviously, this water had to come from the largest reservoir of water: the oceans. In fact, the average depth of the ocean during the last Ice Age was roughly 100 m lower than it is at the present time. During an ice age, water is evaporated from the oceans, carried to high latitudes and deposited as snow over the ice sheets.
The regional climates during the last Ice Age were quite different as well. It rained much more in Utah in those days. So much so that the area from the Wasatch Front westward into Nevada and southward past the Utah town of Delta was all under Lake Bonneville. The shores of this lake at various stages during the last Ice Age (and even previous ones) are easily seen along the Wasatch Front.
An ice age begins with the activation of the ice-albedo feedback in high latitudes of the Northern Hemisphere by a trigger whose nature we examine later. As noted earlier in the chapter, the last ice age began around 120,000 years ago and ended abruptly around 12,000 years ago. There have been some 20 ice ages over the past 2.8 million years. During this lengthy interval the areal extent of the ice and snow in high latitudes of the Southern Hemisphere varied very little. This is because that area is limited by the ocean that surrounds the Antarctic continent. This water surface will not support the slowly and persistently advancing sheet of ice and snow that long term action of the ice-albedo feedback requires. The situation is otherwise in the North Hemisphere, where Siberia and Canada provide a vast land area on which snow and ice can accumulate. Thus, the scene of operation of the ice-albedo feedback during an ice age is the high latitude region of the Northern Hemisphere. Of course,the chill that results from this increased reflection of solar radiation is felt globally.
The relative abundance of different forms of the oxygen atom is one means to infer the state of the climate system over the past million years. Before we explain how this technique works, we must first introduce some basic concepts from chemistry.
Figure 5.5. Two electrons orbit around a nucleus composed of protons and neutrons. This schematic is not drawn to scale; the electrons are much smaller in size compared to the nucleus.
.
Here, the superscript indicates the number of neutrons and protons combined in
the nucleus. We could get away with writing the above as O, except that there
are three different forms of oxygen in the world as shown below:
.
These different forms of oxygen are called oxygen isotopes. The
first one is the most common form of oxygen while the other two are rarely
present in amounts in excess of a few tenths of a percent. The differences
between the three isotopes of oxygen are the number of neutrons found in the
nucleus; the
isotope has two additional neutrons and so is 12% heavier than the
isotope.
A molecule is an aggregate of 2 or more atoms. The molecule carbon dioxide is made up of one carbon atom and two oxygen atoms. So, we can use our chemical shorthand of CO2 to denote a carbon dioxide molecule. Note that with this notation, we don't say which isotope of oxygen is present (nor, for that matter which isotope of carbon).
We will write chemical reactions later that describe how different atoms and molecules react with one another. Let's use as an example the process of photosynthesis by which plants consume carbon dioxide (CO2) and water (H2O) and produce molecular oxygen (O2) and a carbohydrate molecule that is a fundamental part of glucose (sugar):
Imagine this reaction in terms of the transfer of atoms: here the carbon atom leaves the carbon dioxide molecule and attaches itself to the water molecule. This creates as a by-product the molecular form of oxygen.
Another chemical reaction that we will have occasion to cite later on is the one by which many marine plants and animals make shells. This is
Ca[++] + 2HCO3[-] --> CaCO3 + H2O + CO2
The calcium ion (Ca[++]) is a calcium atom that has lost two of its electrons and hence has a net positive charge of two units. The bicarbonate ion (HCO3[-]) consists of a hydrogen atom, a carbon atom and three oxygen atoms, with one electron missing from the combination. There is no net gain or loss of charge when, as in the above reaction, two of these combine with one calcium ion. The reaction yields calcium carbonate (CaCO3), which is the material of the shell, plus water and carbon dioxide. After the creature that has made the shell dies, the shell may dissolve back into its original chemical constituents, in which case the reaction above proceeds from right to left instead of from left to right as we have shown it here.
Better indicators of the climate variations associated with the ice ages can be
obtained by examination of thin diameter cores of sediment taken from the
bottom of the ocean. Ocean cores are taken by using specialized equipment that
is able to reach through the depths of the ocean and drill tens of meters deep
through the accumulated sediment on the ocean floor. Once the sample of
sediment or ice is retrieved, then it is subjected to analysis that measures
the relative abundance of the oxygen isotope [18]O to the
isotope [16]O. As the waters of the ocean evaporate during
an ice age, the water molecule that contains the lighter [16]O isotope evaporates more readily than the water molecule
that contains the [18]O isotope. The water left behind in
the oceans of the ice ages therefore has a greater abundance of [18]O compared to that found during interglacial periods. The
shells of micro-organisms found in the oceans, both plant and animal life,
contain more of the [18]O isotope during ice ages than they
do during interglacials because the oxygen is extracted from the water in which
they live. We shall learn more about the chemistry of shell building in
Chapter 6. When these micro-organisms die and sink to the bottom of the ocean,
they leave in the bottom sediment a record of the relative abundance of the two
different isotopes of oxygen in their shells.
Figure 5.6. Variation in the concentration of [18]O in shells as a function of depth (in m) in a sediment core in the Pacific Ocean. The present time is on the right edge(0 depth) and the left edge (16 m) corresponds to roughly a million years ago.
The results of a determination of the abundance of [18]O relative to [16]O in a core taken from the Pacific Ocean is shown in Fig. 5.6. The most recent part of this record can be calibrated in terms of a record of sea level that goes back to 130,000 years before the present. The top half of Fig. 5.7 shows a magnified copy of the first part of the record in Fig. 5.6. A smoother version of this is then reproduced in the bottom figure and the scale of relative abundance of [18]O is converted to sea level (relative to the present). This is possible because of the independent determinations of sea level represented by the crosses. These in turn have come from radioactive dating applied to old coral reefs that once fringed the islands of Barbados and New Guinea. Although not shown in the figure, more recent work on Barbados corals has determined the maximum sea level drop to have been -121 +/- 5 m at 18,000 years ago.
We thus have in Fig. 5.7 a fairly detailed chronology of sea level (and hence amount of water locked up in ice and snow on land) over the past ice age. One advantage of the [18]O record over the record of the sea level determinations used to calibrate it is that it is continuous and therefore captures the fluctuations such as those seen in Fig. 5.7 that occur within an ice age. A second advantage of the [18]O record is that it goes back much further in time than the sea level determinations. This enables us to identify other glacial and interglacial periods and (although it is not very obvious from the visual image in Fig. 5.6) to establish that ice ages have occurred repeatedly at intervals spaced roughly 100,000 year apart.
Need to put in figure
Figure 5.7. The top half contains a magnified view of the first part of the oxygen isotope record that was shown in Fig. 5.6. A smoothed version of this curve is seen in the bottom half of the figure. The calibration in terms of sea level has been done from dating of old coral reefs whose tops were close to sea level on the island of Barbados (B) and the island of New Guinea (NG).
The cycles of glacial and interglacial periods over the past million years have spawned many theories to explain their occurrence. Milutin Milankovitch (a Serbian scientist) built on earlier work and proposed in the 1920's that these cycles were due to subtle variations in the solar energy incident on the earth that arise from changes in the orbital characteristics of the earth about the sun. Until the past decade or so, this theory was largely dismissed in the scientific community. However, the systematic documentation of the temporal characteristics of ice ages obtained from oxygen isotope analysis of the sediments in deep sea cores have rendered this theory a plausible explanation for the behavior of our climate system over these long time periods.
The Milankovitch theory was treated with limited regard for many years because it was assumed by most scientists that the intensity of solar radiation received at the top of the atmosphere was constant. That is why in Chapter 2 we referred to the amount of energy incident at the top of the atmosphere as the solar constant and determined its value to be 1368 W m[-2]. However, measurements from satellite instruments have been able to show that even within the last 100 years, the solar constant has varied, although the changes are small relative to those on longer time scales. These satellite observations have been used by scientists to extrapolate backwards the variations in the solar constant over the past 100 years as shown in Fig. 5.8.
Notice that the solar constant has varied by as much as 1 W m[-2
]over the past hundred years, with the most pronounced variations occuring
in the past 40 years. There also appear to be semi-regular cycles in the
solar constant with a period of roughly 11 years. A similar semi-regular 11
year cycle has also been observed in the number of sunspots (dark blotches) on
the surface of the sun with the maximum number of sunspots occuring at the
times of maximum solar constant. Many people have attempted to relate
variations in the number of sunspots to changes in weather and climate during
the 20th Century with limited success. The Little Ice Age at its peak during
the 1600's has been linked for many years to the occurrence of a reduced number
of sunspots. It may be possible to infer from the fewer number of sunspots
that the amount of solar radiation incident on the top of the earth's
atmosphere was reduced at that time.
Figure 5.8. An estimate of the variations in the solar constant over the past 100 years. The values along the left margin refer to the solar constant in W m[-2].
Milankovitch developed his theory regarding the occurrence of ice ages at a time when there was no definitive evidence in the geological record for the timing of the ice ages and no corroborative evidence of fluctuations in the amount of solar energy incident on the top of the atmosphere. He postulated that diminished amounts of solar radiation incident on the high latitudes of the Northern Hemisphere in summer contributed to the occurrence of the ice ages. These periods of reduced solar energy he attributed to subtle changes in the characteristics of the earth's orbit around the sun that we will discuss in a moment. Although it was not possible in Milankovitch's day, it is now possible to calculate the earth's orbit for millions of years in advance. Equally well it can be predicted backwards into the past (that is, reconstructed) and in this way we can say exactly what were the variations in the solar constant in the past. These orbital calculations have been carried back 5 million years. This may seem odd, since we discussed in Section 3.3 that it is impossible to predict the weather more than 10 days in advance. The orbital characteristics of the earth constitute a much simpler system than the fluid motion of the atmosphere.
a.
b.
Figure 5.9. (a) The earth tilts at the present time at an angle of 23.5[o] relative to a line drawn perpendicular to the plane on which the earth revolves around the sun. (b) A schematic of the earth in which it has no tilt.
The essence of the Milankovitch theory is that ice ages come and go in response to the slow variations in the earth's orbit around the sun and the slow variations in the orientation of the earth's rotational axis that together act to slightly vary the intensity of solar radiation received at the earth. What is operating in the ice-albedo feedback during the development of an ice age is an increase in the area covered by ice and snow in high latitudes of the Northern Hemisphere. Milankovitch theory specifically ascribes this to relatively cool summers in that part of the world accompanying incident solar radiation that is slightly weaker than normal. The idea is that, when the earth's orbital characteristics are such as to produce a long epoch of relatively cool summers in high latitudes of the Northern Hemisphere, the snow which falls there in the winter doesn't melt completely. In this way, a snow covered polar cap creeps ever further southward, increasing the area covered and thereby maintaining the ice-albedo feedback loop of Fig. 5.3.
Milankovitch recognized that the earth's orbit around the sun varied in three different manners:
* tilt,
* eccentricity, and
* precession.
We will discuss briefly these ways in which the earth's orbit changes. It is difficult to visualize these changes in the earth's orbit, especially because all three are taking place at the same time. To simplify our discussion, we will investigate each type of orbital variation separately, as if the other two are held constant. Only at the end will we examine the cumulative effect of all three changes.
Let's review first the orientation of the earth with respect to the plane of it's revolution around the sun as shown in Fig. 5.9.
The tilt of the earth provides us with our pronounced seasons in the temperate latitudes of the Northern and Southern Hemispheres. The sketch drawn in Fig. 5.9a applies to the winter season of the Northern Hemisphere. Notice that since the sun is located far off the right edge of the figure, the high latitudes of the Northern Hemisphere will receive relatively little solar radiation while those of the Southern Hemisphere are exposed to larger amounts of solar energy.
If the earth had no tilt (and if the orbit of the earth about the sun were a circle), then the amount of solar radiation received in the polar latitudes of the Northern and Southern Hemispheres would be the same (see Fig. 5.9b). Would that mean that the weather and climate over the United States would be milder, since there would not be a seasonal variation in solar energy? Actually, the earth would probably have more violent weather, since the atmosphere (and oceans) must continue to transport the excess energy in the tropical latitudes poleward. There would be a greater contrast in temperature between the equator and poles, and the atmospheric heat engine would have to work harder to maintain the earth-atmopshere climate system in radiative equilibrium.
While it is impossible for the earth to ever reach a point where it has no tilt, Milankovitch knew that the tilt varies between 22[o] and 24.5[o] over a period of 41,000 years: that is, in a span of 20,500 years, the tilt changes from 22[o] to 24.5[o] and returns to its original value in another 20,500 years. He also knew that the eccentricity of the earth's orbit varies over a period of approximately 100,000 years. At the present time, the earth's orbit can be described as nearly circular, with a slight eccentricity (that is, elliptical component) as in Fig. 5.10.
Figure 5.10. The present orbit of the earth about the sun is nearly circular (heavy line) compared to that 50,000 years ago (thin line).
The eccentricity arises as a result of the combined gravitational attraction of the planets as they rotate about the sun at present. At the present time, the earth is closest to the sun during January (when the earth is near the left edge of Fig. 5.10) and furthest away during July (when the earth is near the right edge). The time when the earth is at its closest point to the sun is called perihelion and when it is furthest away it is called aphelion. The differences in solar radiation at the top of the atmosphere between these two times of the year is about 7% at present. So, the polar latitudes of the Northern Hemisphere receive less solar radiation during summer than do corresponding latitudes of the Southern Hemisphere.
The eccentricity of the earth's orbit varies over a period of a 100,000 years. Fifty thousand years ago, the earth had a much more elliptical orbit and this will be the case again some 50,000 years hence. As the earth's orbit becomes more elliptical, the amount of solar radiation received at aphelion is reduced. If Northern Hemisphere summer occurs at the time of aphelion as it does at present, then summers will be markedly cooler when the earth's orbit is highly eccentric than when it is nearly circular.
a.
b.
Figure 5.11. (a) At the present time, the axis of the earth points to the north star. (b) The axis will point in the "opposite" direction 11,500 years hence.
The present coincidence of aphelion with Northern Hemisphere summer is a temporary one, however. The time of aphelion is drifting slowly backward toward spring. The cause of this drift is the final aspect of the earth's orbit that we will consider, precession. This orbital characteristic is literally the wobble of the earth about its axis of rotation. You have seen precession whenever a top or gyroscope is spun. When a top is spinning fast, there is no precession but, as it slows down, the direction in which the axis of rotation points traces out a large circle with respect to the vertical.
Right now the earth's axis points towards the pole star each day during the year (that is, at each point of the earth's orbit about the sun) as shown in Fig. 5.11a. This axis of rotation will very slowly precess about a line perpendicular to the plane of the earth's orbit around the sun. The period of precession is 23,000 years. In 11,500 years, the earth will point away from the pole star as shown in Fig. 5.11b. This in turn will mean that the time of perihelion will have advanced to July, while aphelion will have regressed to Janaury. Then, our summers in the Northern Hemisphere will tend to be warmer than they are at the present time and the summers in the Southern Hemisphere will tend to be cooler.
We can summarize these concepts by stating that an ice age is likely when:
* the tilt of the earth's axis is more than that at present,
* the eccentricity of the earth's orbit is more than that at present, and
* the time of aphelion is in Northern Hemisphere summer.
To help visualize the net result of these three factors, consider Fig. 5.12.
Here, three records of solar radiation are arbitrarily constructed with the
same amplitude and with periodicities proportional to those observed for the
tilt, eccentricity, and precession of the earth's orbit. Notice that the sum of
these three records of solar radiation, each of which individually have
well-defined periodic behavior, exhibits rather irregular changes in radiation
with time, in which there are occasional periods of high solar radiation and
other periods with low radiation.
Figure 5.12. Three records of solar radiation are shown schematically as a function of time. The bottom solid curve depicts a record with a period of 100,000 years; the heavy solid line depicts a record with a period of 41,000 years; and the dashed curve indicates a record with a period of 23,000 years. The top figure indicates the sum of these three records.
Figure 5.12 is simply a mathematical construct created by us to help visualize how tilt, eccentricity, and precession combine with one another. The actual variations of solar radiation calculated from the orbital characteristics of the earth are shown in Fig. 5.13. The correspondence between this figure and records from sediment cores is not particularly impressive but that is because the three contributing effects in reality have different amplitudes.
This leads us to mention the most important respect in which Milankovitch theory by itself fails to explain the observed behavior of the ice age cycles. In the observed sediment cores, the 100,000 year cycle is the dominant one. This fact is reflected in our statement that ice ages are spaced about 100,000 years apart. However, the two shorter cycles of 23,000 and 41,000 years (precession and tilt, respectively) dominate the solar radiation estimate (Fig. 5.13).
One possible explanation for the discrepancies between the solar radiation predicted from orbital variations and the actual record of ice ages is that the nonlinear interactions in the earth-atmosphere climate system lead to chaotic behavior on these very long time scales (see Section 3.3 for a discussion of chaos). On the other hand, there are potential feedback processes in the climate system that may help to explain the discrepancy. Some current possibilities are the dynamics of the ice sheet itself and the process of the sinking of the land as the ice sheet grows.
We can summarize much of this chapter by stating that the earth-atmosphere
climate system has undergone fairly large variations in the globally averaged
temperature field. As we investigate the climate system on longer and longer
time scales, we see that the amplitude of the variations in temperature become
larger and larger, yet remain within fairly well-defined ranges. We can gain
some comfort from this in the sense that our climate system appears quite
stable. Nonetheless, we must be mindful of the fact that it has been during an
interglacial period that our civilization has evolved. Even insignificant
climate fluctuations on the time scales investigated in this chapter may have
profound effects on the world economy and our personal lives.
Figure 5.13.Solar radiation (in W m[-2]) incident at the top of the atmosphere during Northern Hemisphere summer estimated from calculations of the orbital characteristics of the earth.
1. How does the magnitude of the climate variations induced by volcanic activity and air-sea interactions compare to that induced by variations in the orbital characteristics of the earth?
2. What causes an icea age? How often do they occur? How long do they last? When did the last ice age reach it peak and whe did it end?
3. Why might the Eemian interglacial period that took place 120,000 years ago and the Holocene optimum period 5,600 years ago be considered proxies for future periods with higher carbon dioxide concentrations than at present?
4. How and why do analyses of deep sea cores provide us with information on the temporal characteristics of ice ages?
5. Why does the depth of the ocean decrease during ice ages?
6. What is the ice-albedo feedback loop?
7. Why is the concentration of
relative to that of
greater in sediment cores during glacial periods?
8. Invent and describe a feedback process by which an ice age might come to an end.
[*] Prior to the mid-1970's there was more popular concern over the effects of the impending ice age than global warming. In fact, one author went as far as having the Empire State Building embedded in a glacier on a book's cover. These concerns have surfaced again in the media after the cold and snowy 1993-94 winter in the eastern United States.
[*]Electrons and protons determine the charge of the atom; the neutrons have no charge. You may view charge as the ability of electrons to move from the orbit of one atom to another.