6. Subglacial Lakes
In 1960, the Soviet pilot R. V. Robinson noticed extensive flat areas in the otherwise wrinkled surface of the Antarctic ice sheet. He called these areas lakes, probably never knowing how right he was. Robinson surely knew that these smooth areas were ice, frozen solid by the frigid temperatures. He would have known that these flat surfaces, though they looked similar to the frozen surface of lakes in more temperate climates, did not have liquid water a few feet below. But in a way Robinson never could have known, these regions of flat ice were indeed the covering of a lake, albeit a lake located miles beneath the surface.
Underneath Russian research station Vostok on the East Antarctic Ice Sheet, lays Lake Vostok, a lake unlike any other. No human has yet seen Lake Vostok, which is a pocket of water trapped between the Antarctic bedrock below and the massive Antarctic ice sheet above. It is a lake without an air/water interface. It has a rock/water interface at its bottom and an ice/water interface at its roof, but that roof is over two miles distant from the atmosphere, far from air and from light. Lake Vostok is enormous, its estimated volume close to that of Lake Ontario. And, what is especially interesting from a biological point of view, there is no direct interaction between the water in Lake Vostok and Earth’s atmosphere. The water in this lake has not been exposed to air for at least 100,000 years, and by some estimates not for a million or perhaps even ten million years. This means whatever life resided in that pocket of water as the Antarctic ice sheet grew is still present, having evolved independently of the bacteria and fish and mammals located far above. Although researchers speak only of comparing the bacteria from Lake Vostok with those found elsewhere on Earth, there is always the unspoken possibility: What else might live there? Could something larger than bacteria have learned to survive there over the course of a million years? Could evolution have proceeded in these subglacial lakes along a path different from that found in lakes on the surface far above?
Lake Vostok is by far the largest of the known subglacial lakes. It is 155 miles long and about 50 miles wide with a total surface area of 5400 square miles and a volume of 1200 cubic miles. The thick sheet of ice above Vostok is impressive, but Vostok itself is also impressive, having a depth of more than 3280 feet, deeper than several of the very deep lakes we’ve already discussed such as Crater Lake and Lake Nyos; it is deeper than all of the Great Lakes. Lake Vostok is so big that it ranks as the seventh largest lake in the world.
But while Lake Vostok is the largest subglacial lake on Earth, and the first discovered, it is by no means the only one. Excitement over its discovery spurred studies of subglacial lakes in general. Aided initially by seismic soundings, and then radio soundings, and finally by satellite scans, glaciologists have found many more subglacial lakes, estimates rising steadily from little more than a dozen in the 1970s to about 400 today. Much like their terrestrial counterparts, subglacial lakes are often connected by subglacial river systems, some lakes rising at the expense of others. These transfers of water from one lake to another can be significant enough to change the height of the ice sheet located miles above the water surface. We now know the base of the Antarctic ice sheet has something of a secret hydrological life, where meltwater is channeled along riverine systems to and from various lakes. Indeed, it is estimated that as much as half of the glacial base of Antarctica is wet, not frozen, covered either in lakes or in flows of meltwater traveling from one place to another.
Antarctica is a very cold place, and exactly how any liquid water can exist at all, let alone an enormous body of liquid water like a subglacial lake, is an interesting question in and of itself. The temperature above Vostok Station, the Russian outpost where much of the research on Lake Vostok takes place, is extremely cold with temperatures as low as −75 degrees F. Indeed, the lowest temperature ever measured on Earth was recorded at Vostok Station, on July of 1983: −128.6 degrees F. A temperature this low is something to ponder. If you were in a place where the temperature was that low, and you were somehow able to increase that temperature by a whopping 100 degrees F., you would still be just shy of 30 degrees below zero, requiring yet a second 100-degree boost in temperature just to get to room temperature. If you wanted to melt and then boil a piece of ice from the ice sheet at Vostok on that very cold day in 1983, you would have to increase its temperature by 341 degrees F. Vostok is clearly a place that does not favor the liquid phase of water.
It is disconcerting to realize such a harsh environment exists on our planet, a place so cold that the liquid in our own bodies would always be on the verge of solidification. Yet, in spite of this incredible cold, liquid water does exist in Antarctica, albeit far from the surface.
The reason this is possible involves several factors. First, when we speak of the incredible cold in Antarctica, we are referring to the air temperature. The air temperatures at the poles are cold due to the reduced energy received from the sun. Indeed the South Pole is dark 24 hours a day during its winter, receiving virtually no energy from the sun during this time. In the summer it is light all day long, although the altitude of the Sun barely skirts above the horizon, reaching its peak zenith at a measly 23.5 degrees above the horizon at the summer solstice, providing very little energy to the surface. In contrast, regions closer to the equator receive significant energy from the sun. In temperate regions, digging a hole in the ground generally exposes soil that is cooler than the air above it, certainly in the summer, something your dog may have taken advantage of to cool her belly on a hot summer day. But at the poles, drilling downward into the ice sheet actually results in an increase in temperature.
To better understand this, it bears remembering that the center of our planet is quite hot, and the heat generated in the core conducts outward to the surface. The amount of heat emanating from the Earth’s surface due to the hot core of the planet is called the “geothermal heat flux,” and its value varies with location but is typically around 50 milliwatts per square meter. This is not a very large flux of heat, and it has little effect on the temperature near the equator or temperate regions of the planet where the flux of heat from the Sun exceeds 1000 watts per square meter (one million milliwatts per square meter), over 20,000 times the geothermal heat flux. Indeed, if one sought to obtain the amount of energy needed to power an old-fashioned 100-watt lightbulb from this geothermal heat flux, a surface area of over 21,500 square feet would be needed, a patch of land 147 feet on a side. But, though small, this quantity of heat can melt ice at the bottom of the Antarctic ice sheet because of several factors. First, at Vostok, the ice sheet is approximately 2.5 miles thick. Much like the much thinner walls of an igloo, this blanket of ice serves effectively as an insulator, protecting the bedrock from the very cold air at the surface. So, although the geothermal heat flux is quite small, the energy it provides from the Earth’s inner core is not snatched away by the howling winds of an Antarctic winter but is held in place, much like the heat emanating from your body is kept in place by a good coat.
Secondly, as we learned in an earlier chapter, glacial ice sheets are virtually always in motion, continuously sliding downhill from higher regions of ice accumulation. This flow results in frictional heating as the ice slides over the bedrock or flows by deformation under the enormous pressure of the ice sheet. Such heating adds to the geothermal heat flux, providing another source of energy to melt ice. Finally, the immense pressure at the bottom of the ice sheet reduces the melting point of ice, decreasing it by 5.4 degrees F., from 32 degrees F. to 26.6 degrees F., allowing the ice to melt at a lower temperature than would otherwise be the case. Combined, these factors cause ice to melt in the region of Lake Vostok, as well as in the many other regions beneath the Antarctic ice sheet where subglacial lakes exist.
Given that water can melt in the region of Lake Vostok, one might wonder what keeps the process in check? Why isn’t Vostok even bigger, with more of the ice sheet melting to create more lake water? Or, what keeps it from shrinking, from cooling just enough to reverse the melting and cause the lake to solidify? The answer comes from an energy budget for the lake, a comparison of the amount of heat coming in compared to the amount going out. Heat comes in from geothermal heating and from frictional heating and perhaps from relatively warmer meltwater flowing in. Heat is lost if relatively warm meltwater flows out of Vostok*, and heat is lost to the colder ice that slowly flows across the lake surface, sucking heat away from the relatively warmer lake water. If the amount of heat entering the lake exceeds that lost, melting will occur. If the amount of heat leaving exceeds that going in, it will cool and perhaps freeze. The lake size will grow or shrink until equilibrium is attained, a point where the inflows and outflows of energy equal each other, at which point the lake size will be stable.
But the above treats all of Lake Vostok and the other subglacial lakes as if they are homogenous, each part of the entire lake under more or less the same conditions. But the situation is far more complex and interesting. Indeed, along the roof of Lake Vostok, freezing is actually occurring in some places, ice accreting in that locale, while melting occurs in other locations where ice is lost. This process leads to a circulation of water within the lake, hardly a torrential flow, but vigorous enough to suspend sediment from the lakebed into the water. This is interesting because it means that the water samples that have been obtained from the lake (more on which below) could reveal its history. The suspended bottom sediment may provide a window into the ancient life that has existed in this lake. The flow is due to a complicated chain of physical events, and it all depends on a small bit of ice/water physics, the simple but odd fact that the roofs of all subglacial lakes are tilted.
At its northern shore, the roof of Lake Vostok is about 1310 feet lower than at its southern shore. It is a strange fact, and one with profound consequences. It is strange because at the surface of the ice sheet above Vostok, there is an opposing and much smaller tilt of just 131 feet. The situation is revealed in Figure 5, which shows first the situation one might expect to exist in a lake covered with ice and, in the lower part of the figure, the situation that actually exists at Vostok as well as at other subglacial lakes (albeit with different thicknesses).
The fact that the surface of the ice is tilted one way and the roof of the lake is tilted the other is counterintuitive and probably best understood by focusing on the floor of the lake. Let us imagine this floor is flat as shown in Figure 5 (it is not, but this does not affect the mechanism we are trying to understand). Now, referring to the upper portion of the figure, let’s begin with the simple case where the ice sheet is flat at the surface. The figure shows two locations, A and B, both at the bottom of the lake. In the upper portion of Figure 5, the combined weights of the water and ice are the same at A and B because the thickness of the ice and water layers above each of these points is the same. Let us now imagine the amount of ice on the surface changes so that there is 131 feet more ice above point A than above point B. In other words, imagine the upper ice surface is tilted so that it is 131 feet higher on the left than the right. This would make the weight at A higher than at B, therefore the water pressure at A would be higher than B. Such a situation could not exist in equilibrium, and the system would adjust itself so that equilibrium is reestablished. Here, the ice/water interface tilts until the pressures at A and B equilibrate. Since, in our thought experiment, the weight above A is initially higher than at B, and since ice is less dense than water, the ice/water interface would tilt so there would be more (lighter) ice over A and less water.
Figure 5. Relative thicknesses of ice and water in a subglacial lake. Above is the situation one might expect, and below is what actually exists in Lake Vostok and likely in all subglacial lakes. A is the northern shore of Vostok and B is its southern shore.
This probably makes a reasonable amount of sense. Where the situation becomes harder to understand is, exactly why the small tilt of the upper ice surface (131 feet) must be compensated by such a large tilt of the ice sheet at the ice/water interface (1310 feet). Why this ten-to-one ratio? The answer takes a bit a mental focus but can be understood if one imagines slowly pulling down the left side of the line defining the ice/water interface. If you pull that line down on the left side of the diagram, you are increasing the thickness of ice and decreasing the thickness of water. This will indeed reduce the pressure at A, bringing it closer to the pressure at B—so that equilibrium is approached.
However, though ice and water have different densities, they are not that different. At the pressures of the bottom of Lake Vostok, liquid water has a density of 63.43 pounds per cubic foot, while ice has a density of 57 pounds per cubic foot, a difference of only about 10 percent. Hence, pulling down the left side of the ice/water interface does add lighter ice and reduces heavier water, but since the ice is not that much lighter than the water, the interface has to be moved quite a bit before the water pressure at A and B are in equilibrium, as they must be. This is especially true because regardless of how we tilt the ice/water interface, the total height of liquid and/or solid above A is still higher than that at B.
It is a difficult bit of reasoning to wrap one’s head around. But, regardless of the reason, the tilt in the ice/water interface is about ten times larger than at the surface of the ice sheet. This is the case for all subglacial lakes. The relatively small slope on the surface of the ice sheet at Vostok causes the roof of the lake to change by about 1310 feet from the north to the south shore. That is quite a bit of change, considering that Vostok is about 3280 feet deep at its deepest location.
The change in height along the roof of a subglacial lake has significant consequences. It is widely believed that this slope is responsible for circulation in these lakes. The temperature of liquid water immediately adjacent to an ice/water interface should be quite close to the freezing point. In Lake Vostok, the northern end of the lake is 1310 feet deeper than the southern end and due to the different pressures at those depths, the freezing point is therefore about 0.5 degrees F. lower at the northern end than the southern end. Hence, the northern water near the roof is colder and denser. This denser water sinks and partially follows the bottom of the lake, traveling to the south. The interesting thing is that this cooler water is being transported to the southern end of the lake where the freezing point is higher, causing that water to freeze. Hence, ice is accreting at that end. To maintain an ice and heat balance, there would have to be melting and the formation of meltwater at the northern end of the lake. Radar soundings show this is precisely the case.
The fascinating thing about all of this is that this circulation mechanism applies not just to Lake Vostok, but to all subglacial lakes. All subglacial lakes are believed to have sloped roofs and to experience this type of circulation driven by a difference in melting points between the thick-ice end of the lake and the thin-ice end.
Now, the time scales of all of the processes described here are slow—glacial. The ice flow over the surface of Vostok is estimated to move at about 10 feet per year, taking tens of thousands of years to flow over the entire lake surface. The rate of ice accretion at the locations where water freezes is about an inch per year. And though all of these rates pale in comparison to water velocities in a temperate lake (where water velocities can be orders of magnitude larger due to wind and inflows and rain), there is still just enough vigor in these flows to enable sediment at the bottom of Vostok to be suspended in the lake water. The lake water velocities are estimated to be just strong enough to keep particles as large as 20 microns in diameter suspended in the water. Indeed, Vostok may be turbid.
Because bacteria are typically much smaller than 20 microns, it is believed that water samples from this mysterious, prehistoric Lake Vostok, should reveal not just what kind of life exists in the lake now, but via sediment particles, should also reveal the history of the life that existed in its prehistoric past. To find out, scientists would need to get a sample, which would require drilling through 2.5 miles of ice, without introducing bacteria from the surface, bacteria from the modern era, into this ancient lake. Accomplishing this was not without its challenges, but it has been done.
To date, drilling into subglacial lakes in Antarctica has been pursued at three locations. Drilling at Lake Vostok, a primarily Russian activity with support from other nations, has been ongoing. A U.K.-led effort exists at Lake Ellsworth, a narrow subglacial lake about 8.7 miles long, with a width mostly less than a mile, a maximum depth of 512 feet and a volume of about 0.3 cubic miles. It is covered by an ice sheet about 2 miles in thickness. A U.S.-led effort exists at Lake Whillans, which has the advantage of having an ice sheet “only” about 2600 feet thick. Whillans is thought to be relatively shallow, perhaps only a few yards deep, and also very hydrologically active; two periods have been recorded when the surface ice above Whillans has risen and fallen. As of 2017, of these three opportunities, two have resulted in the penetration of the ice core into the lake and have enabled retrieval of samples from the lake. These were Lakes Vostok and Whillans.
When the Russian team broke into Lake Vostok, they did so via a borehole they’d been developing for over 15 years. That borehole was initially used to obtain an ice core with an end point that came within about 500 feet of the water surface and yielded useful information in and of itself. But in February 2012, the team broke into Lake Vostok proper. When the drill bit penetrated the surface, lake water rushed upward and promptly froze. The next season, a core was taken of this lake water, referred to as fresh frozen lake water, and was made available to the scientific community. The results showed the existence of life, albeit all microbial in nature, and analyses of the bacteria and fungi in this core and in similarly obtained cores has been ongoing.
But the Russian program had always been somewhat controversial.
Drilling in Antarctica, particularly eastern Antarctica, which is colder than its western counterpart, is challenging to put it mildly. One of the difficulties with such an endeavor is simply ensuring your drill bit does not freeze to the bore hole, not an easy feat at air temperatures near −100 degrees F. One way to prevent freezing is to utilize a drilling fluid in the borehole that does not freeze. This was the approach that the Russians took, utilizing a mixture of kerosene and freon which stays liquid even at the extreme temperatures of eastern Antarctica. But kerosene is not as antiseptic as one might think; strains of bacteria call it home. So, by the time the drill bit broke through Vostok the miles-deep hole was filled with an estimated 60 tons of kerosene/freon mixture. A very real concern was that some of this mixture would fall into the lake when the bit penetrated the roof, forever contaminating the prehistoric lake with modern bacterial strains. But, because kerosene/freon is of lower density than water, when the lake surface was penetrated, lake water rushed upward. Thus the drilling fluid did not fall into the lake. From the perspective of preventing contamination of the lake, this was a plus. But at the same time, the ice cores obtained from the Lake Vostok borehole were contaminated, or such bacterial contamination could not be ruled out. Hence, although the core from Lake Vostok revealed the presence of microbial life, the significance and accuracy of these results was questioned.
A way to avoid the contamination of the Vostok approach is to use a hot-water drill. This technique prevents the drill from freezing to the ice, while also ensuring that the borehole is clean by implementing a sterilization procedure. Such an approach would ensure no external bacteria are introduced into any samples obtained from the lake. This approach was taken at Lake Ellsworth but failed due to equipment malfunctions. However, when a similar approach was applied in 2013 at Lake Whillans, it was successful. Lake samples from that work revealed a system of microorganisms that utilize chemosynthesis to extract energy from the minerals present in the lake.
In both of the subglacial lakes penetrated in Antarctica, evidence of anything larger than microorganisms has not been obtained. Moreover, truth be told, none of these drilling activities were pursued with the hope of finding evidence of aquatic dinosaurs swimming about. The expectation was for evidence of only microbial life. Bear in mind that these lakes are all completely dark, precluding the possibility of photosynthetic life. Furthermore, carbon sources in subglacial lakes are limited to organic matter scraped into the lake by the ice sheet flowing over it, or whatever may enter the lake from the ice above, that is, matter deposited onto the ice sheet that slowly made its way downward as the ice sheet formed over the interceding millennia. This is not a very large flux of carbonaceous matter and therefore sources of nutrients to sustain any life that might exist in Vostok are thought to be limited. There are also sediments that may have been left behind since before the Antarctic ice sheet existed, at the very beginning of the current ice age millions of years ago. But, by and large, whatever life exists in Lake Vostok would have to make do with minimal sources of energy and nutrients.
Nevertheless, a full understanding of the ecology of Lakes Vostok and Whillans will likely require further drilling, particularly into the floor of the lakes to further understand the sediments at the bottom. How this may be accomplished and what it might reveal are still to be determined. Even if these two lakes are found to contain nothing of especial interest, the story is unlikely to end there. It is estimated that there are more than 400 subglacial lakes, and there is nothing to say the ecosystems of these lakes have to be the same. Meanwhile, on the other side of the planet, a subglacial lake was identified in 2018 in the Canadian Arctic. This time the lake is a salt lake, suggesting that the story of subglacial lakes on Earth is far from over. Moreover, the discovery of subglacial lakes on other planets seems to be just beginning. Liquid water was detected beneath the ice near Mars’ South Pole in 2018. This is thought to be a salt lake too, though likely its salts are perchlorates, different from typical salt lakes on Earth.
Clearly, we are living in the salad days of subglacial lake research, likely with many revelations in the future. There has been absolutely no evidence to suggest the existence of dinosaurs in the subglacial lakes of our planet. But still, I keep my fingers crossed.
* It is unclear if there actually is a flow of meltwater in/out of Vostok. But this process does occur for other subglacial lakes.