Optional Enrichment Article - A little more about glaciers
Types of Glaciers
Glaciers occur in different forms and sizes, and you might occasionally find knowledgeable scientists who disagree about what to call a particular glacier. An ice sheet is huge - the size of a continent, or at least the world's largest island (Greenland) - and spreads in all directions. An ice cap or ice dome is a smaller version of an ice sheet, sitting on a mountain top or high plateau, and also spreading in all directions (or at least in several directions). Many glaciers flow down from some mountain peak and may be called mountain glaciers, or valley glaciers, or just glaciers. An outlet glacier is a fast-moving part near the edge of an ice sheet or ice cap, especially if it flows between rock walls; fast-flowing parts near the edges of an ice sheet or ice cap flowing between slower-flowing ice are called ice streams. And yes, there are cases where an ice sheet is drained by a fast flow with ice on one side and rock on the other. Classifications such as this help us talk about things but are not precise.
Flowing Solids and Hot Ice
Dr. Alley has spent months of his life living on the great ice sheets of Greenland and Antarctica. (And Dr. Anandakrishnan has spent a lot more time on the ice sheets than Dr. Alley has!) Eating and sleeping and working at -30º, it is hard to think of ice as being a hot material, but that is exactly what it is, as noted earlier in this module.
Recall that heat is the vibration of atoms or molecules in a material and that in most solids including ice, the atoms or molecules are arranged in regular, repeating patterns. Melting of ice occurs when the typical molecule vibrates fast and hard enough to break free from the bonds that tie it to its neighbors and escape from that regular arrangement. When a material is almost hot enough to melt, the atoms vibrate almost hard enough to break free from their neighbors and move around, so it is relatively easy with a little extra push to move a few molecules at a time past their neighbors. The gravitational stresses caused by the surface slope of a glacier supply that little extra push and the ice deforms. (This deformation is primarily by dislocation glide - something like moving a carpet by making a rumple on one side of the room and then slowly shoving that rumple to the other side of the room.) When a material is not nearly hot enough to melt, the molecules are not even close to vibrating hard enough to break free from their neighbors, a whole lot of extra push is required to move molecules, and moving even a few molecules at a time is very difficult. The material then deforms elastically, or it breaks, but it does not creep and deform permanently in the way that a glacier does.
Most people measure temperature on a scale that gives “nice” numbers (something between 0 and 100) for typical daytime temperatures, so talking about the temperature is easy for us. But, other temperature scales make more sense in physics. If you slow the vibrations of molecules by cooling them, you can imagine that there must be some temperature at which vibration stops because all the heat has been removed. We call that temperature “absolute zero” or just zero on an absolute temperature scale. (Yes, in a quantum world, the Heisenberg uncertainty principle means that the last tiny bit of vibration can’t be removed, but vibration is almost completely stopped at absolute zero.) If we set the zero on our temperature scale to this “absolute zero,” and then use degrees that have the same size as in the commonly used Celsius or Centigrade scale, we get the Kelvin scale. Ice melts at 273ºK and water boils at 373ºK; there are 100 degrees between melting and boiling in Kelvin, just as in Celsius. (The Rankine scale uses Fahrenheit-sized degrees and absolute zero as zero, with ice melting at 460ºR and water boiling at 640ºR, but almost nobody uses Rankine anymore, so you are welcome to forget you ever heard about it.)
As a general rule, little or no permanent deformation (creep) occurs when the temperature (in Kelvin or Rankine!) is less than about half the melting temperature, and creep occurs rather easily when temperatures exceed about three-quarters of the melting temperature. The coldest mean-annual temperature on Earth today is about eight-tenths of the melting temperature of ice (that is 217ºK, which is also -56ºC or -69ºF, in case you still like old-fashioned thermometers). Most ice is as close to melting as red-hot or even white-hot iron being worked by a blacksmith. This is why glaciers usually flow rather than break—although breaking is still possible where deformation is very fast and where the pressure is very low, producing crevasses. So, you may find wintertime ice to be uncomfortably cold, but as a material, it is still hot!
Glacier Erosion
Ice can be well below freezing, or at the freezing point. Anyone who has ever defrosted an old-style freezer knows that subfreezing ice built up on the walls of the freezer is VERY hard to remove, but warming it until it reaches the melting point allows the ice to suddenly move easily and slide off. If you are using a chisel or screwdriver to chip the ice away, the sudden motion when the contact with the freezer thaws may cause you to scratch or gouge the freezer. Glaciers that are frozen to rocks beneath them don’t slip over those rocks rapidly and don’t erode those rocks rapidly, but if enough geothermal heat or other heat is supplied to thaw the contact between glacier and rock, the ice slides and can erode rapidly.
Erosion by thawed-bed glaciers occurs mostly in one of three ways: plucking, abrasion, and subglacial streams. We'll describe them a little more here.
First, recall that ice is an unusual material—higher pressure lowers its melting point rather than raising it, opposite to most materials. Ice has a sort of tinker-toy or construction-set structure with a lot of empty space between the molecules, and squeezing ice tends to force molecules to move closer together, making denser water. Most materials have less space in the solid than ice does, and melting requires knocking molecules out of orderly arrangements in ways that take up more space, a change that is opposed by higher pressure.
If a glacier is sliding across a bump in its bed, ice will tend to melt on the up-glacier side of the bump where the pressure is higher. The meltwater will flow around the bump to the down-glacier side, where the lower pressure will allow the water to refreeze. The heat given up by the refreezing will be conducted back through the bump, to allow more melting. But, you may remember that melting and freezing can open cracks in a rock. So, a glacier sliding over its bed can work rocks loose, and then freeze those rocks onto its base, in a process known as plucking. (When water spreads over the bed of a glacier in the spring as melting on the surface starts to feed water downward, the friction with rock that holds the ice back becomes concentrated on smaller regions of the bed not lubricated by the water, and this stress concentration breaks rocks, helping to cause plucking.) And, sometimes basal ice picks up rocks, and those rocks get stuck for a while against the rock beneath and then break free in a little earthquake. The quake pulls ice away from the downstream sides of bumps, lowering the water pressure there, while high-pressure water persists in cracks and spaces in the thawed-bed rock beneath a glacier, allowing a sort of hydrofracking that breaks rocks. )
Once glacier ice contains rocks at the bottom, it is like sandpaper—it drags those rocks over other rocks, scratching and polishing and knocking loose smaller rocks. This process is called abrasion. If you examine rocks on the walls of Yosemite, many still retain a polished appearance with parallel scratches or striations, showing where abrasion was active. Bumps are smoothed and even polished on one side—the up-glacier side—but may be rough and jagged on the down-glacier side where rocks were plucked off of them.
The melting of glaciers can produce a lot of water. The toe of a fast-melting glacier may supply more water to streams than does a similar-sized region in the rainiest place on Earth. The glacier acts to collect snowfall from a big area and take the snow to melt in a much smaller area. Trees and grass do not grow on glaciers to use the melt water but they do grow on the ground to use rainfall. Glacier melt usually flows down holes in the glacier, called moulins, that often form at the bottoms of crevasses. (Some brave or foolhardy people like to go caving in moulins after they drain during the winter.) The moulins eventually reach the glacier bed, where they feed large, steep, fast-moving streams. These erode in the same ways as streams outside of glaciers. Glaciers with much meltwater usually cause erosion to be faster than in non-glaciated regions. Fluctuations in water pressure, as moulins fill with water during daytime melting and drain as melting slows at night, contribute to cracking rocks for plucking.
More on the History and Future of Ice Ages
As we will see later in the course, the climate has changed naturally. Some times far in the past were very hot—too hot for people to live in large parts of the Earth—with the heat primarily from naturally higher concentrations of atmospheric carbon dioxide. (Humans are raising carbon dioxide in the atmosphere now primarily by burning fossil fuels. We are raising carbon dioxide faster than almost all of the natural changes, and human decisions will control how much we raise carbon dioxide and thus how hot we make the climate). Times of very high temperatures had no ice even near the poles, and are sometimes called “hothouse climates”. (Melting all the ice on Earth would raise global sea levels a bit more than 200 feet (60 m).) Natural processes including the formation of fossil fuels have caused cooler times to occur as well, when ice existed near the poles; such times are sometimes called “icehouse climates”, even though most of the world did not have ice. During a few special times, ice spread to cover the whole Earth; these are called “Snowball Earth” events.
The ice has not been constant during icehouse climates, but has gotten bigger and smaller—the ice-age cycles discussed earlier. Recall that these have been paced by features of Earth’s orbit rearranging sunshine by location and season, with the effects made global by changing atmospheric carbon dioxide levels. We are still in an ice-house climate, with ice in Antarctica and Greenland. When this icehouse was established millions of years ago, the ice grew and shrank fairly rapidly, with spacings of 41,000 years between big-ice times especially common. Then, about 800,000 years ago, the behavior shifted (for reasons that are not fully understood, although we have some good hypotheses) to cooling and ice growth for roughly 90,000 years, followed by warming and ice melt for roughly 10,000 years, then repeating. The rate of cooling initially has been slow, so you may read about 10,000 years of warmth followed by cooling. Today, the northern hemisphere has been in the not-much-change/slight-cooling phase for almost 10,000 years already, and you might expect that we are ready to begin sliding into the next ice age. But, it isn't quite that simple.
The 100,000-year pacing of a 90,000-year-cooling/10,000-year-warming world is linked to the interaction of the different orbital cycles, but the 100,000-year cycle in the out-of-roundness of the orbit is important. The orbit goes from nearly round to more squashed and back in about 100,000 years, largely controlled by the tiny tug from the gravity of the planet Jupiter as we pass it in our orbits. And, there is a slower modulation of the out-of-roundness that takes about 400,000 years. More or less, the orbit goes from nearly round to a little squashed, to nearly round, to more squashed, to nearly round, to even more squashed, to nearly round, to not as squashed, to nearly round, to barely squashed, and then this whole thing repeats, with the nearly-rounds spaced roughly 100,000 years apart. We are in the barely-squashed part now, and the last time that the orbit was in the barely-squashed mode, the warm time of the ice-age cycle lasted 30,000 years rather than 10,000 years. Climate models have confirmed that this points to our natural future; actually with roughly 50,000 more years of warmth before the next ice age starts. However, human burning of fossil fuels has already released enough carbon dioxide to warm the climate more than 50,000 years into the future, likely stopping that next ice age. (If we were truly interested in stopping that next ice age, we would wait until the cooling was due and then release the carbon dioxide.)
Also, note that the 19,000-year cycle noted in the text is an oversimplification. There is instead a “quasi” periodicity ranging from 19,000 to 23,000 years, as we mentioned briefly, and this was calculated by Milankovitch and is observed in the data collected to test Milankovitch's calculations, beautifully confirming his predictions. The whole story is a little more complicated than we can fit into a short Enrichment section here, but the basics are clear—orbits pace the ice ages by moving sunshine around on the planet, and this causes environmental changes that shift carbon dioxide between the deep ocean and the atmosphere, globalizing the changes.
Central Pennsylvania and Glaciation
During at least one old glaciation (probably over 1 million years ago), ice flowing south from Canada dammed the West Branch of the Susquehanna River and formed a lake in the Lock Haven area of Pennsylvania. If that lake filled to the next lowest bedrock outlet (into the Juniata River along the Bald Eagle Valley at Dix), then the water would have lapped at the steps of Old Main on Penn State’s University Park campus. There is no evidence of such a large lake, and before the lake filled all the way, it probably drained through the failure of the ice dam, but we’re not sure. With ice so close, however, central Pennsylvania was cold during the ice ages.
Isotopic Ratios of Dead-Bug Shells
In the main text, you learned how the changes in ice volume control the isotopic composition of water in the ocean, and how we can reconstruct the ice-age cycle from the history of shell isotopic compositions in a sediment core because the shells record the water isotopic composition. As usual, things are a bit more complicated than that. Shell isotopic composition also is affected by temperature. When there is more ice on land, the ocean has heavier isotopic ratios in its water, and this gives heavier isotopic ratios in shells growing in the water, but colder temperatures also give heavier isotopic ratios in shells. (At high temperatures, both heavy and light atoms have plenty of energy to jump out of a shell; at low temperatures, the heavy ones tend to get stuck in shells while the light ones can jump out.) Because both colder water and ice favor isotopically heavier shells, measurement of shell isotopic composition cannot tell you the relative importance of temperature versus ice volume.
One way around this is to go to a place that is cold today; the water was above freezing during the ice age (shells were living in it…), so there the signal must be primarily one of ice volume. Other approaches include finding additional paleo-thermometers, such as estimating the temperature from the species living in a place and leaving their shells, or using changes in other “contaminant” ratios in shells that depend on temperature. Yet another way is that there is water in spaces in mud, and the water in some sediments is from the ice age, so just measure the isotopic composition of that water.
The result of this is that isotopic ratios did change because there was much more ice during the ice age than today and because most places were much colder during the ice age than today.