Predicting flooding from glacial lakes

Science Writing

What in water did Bloom, waterlover, drawer of water, watercarrier, returning to the range, admire?

—Its subsidence after devastation: its violence in seaquakes, waterspouts, Artesian wells, eruptions, torrents: its persevering penetrativeness in runnels, gullies, inadequate dams: its solidity in glaciers, icebergs, icefloes.

Scientists in Switzerland, Alaska, and Iceland are studying the nature of water as it drains from glaciers in what are known as outburst floods. As in Joyce’s description, water penetrates the glacial dam until it becomes “inadequate,” and scientists hope to someday be able to predict flooding events to give local residents warning.

Ancient floods

Outburst floods are a natural phenomenon. In fact, the largest floods known to geologists were a result of outburst events. About 12,000 years ago, the floods from Lake Missoula in North America carved the landscape into the Channeled Scablands seen today in Washington state. The ice dam was estimated to be over 600 meters tall, and it held back as much water as Lakes Erie and Ontario combined (2,120 cubic kilometers, which would make it the tenth-largest continental lake in the world today). When the dam broke, colossal waves sped towards the ocean at 50 to 80 kilometers per hour, carrying pieces of ice ripped from the glacier and enormous boulders. This happened dozens of times over the centuries.

An exception to the rule in geology, glacial outbursts like the Missoula floods are an example of catastrophism, which is the idea that the surface of the Earth is shaped by occasional large events. Most Earth processes, on the other hand, take place slowly over thousands and even millions of years. The erosion accomplished by one Missoula flood was like that of sixty steadily-flowing Amazon Rivers.

Today, glacial floods are smaller by three orders of magnitude, but the process by which they happen is similar to that of the Missoula floods. Glaciologists study the timing and intensity of modern glacial flooding primarily through seismic and hydrologic measurements.

Ice-quakes help researchers “look inside” glaciers

The first method of studying glacial flooding is seismology. In their recently published study, Fabian Walter and David Heeszel call this glacial “humming” (Geology, 24 October 2014). Walter, currently a research assistant at the Federal Institute for Forest, Snow and Landscape Research in Birmensdorf, Switzerland, spent one summer measuring seismic activity at different points along the Gorner glacier.

This particular glacier is part of the second-largest glacial system in Switzerland. Every year in spring, snowmelt collects at the edge of the glacier. The resulting ice-marginal lake, Gornersee, is one of the few glacial lakes in the Alps that drains every summer by flooding.

Image
Gornersee before and after summer drainage event. Credit to Mauro Werder.

In their study, Walter and his colleagues placed seven recording devices along the glacier, a few hundred meters away from the lake margin. They installed a final sensor close to the bottom of the glacier, 150 meters deep. The research team had to adjust these tripod sensors every day during the field season because of the melting surface of the glacier. The devices collected data about water moving through cracks in the glacier.

These data revealed a phenomenon called gliding harmonic tremor; this is the first time such tremors have been recorded at glaciers. The researchers in this study hypothesize that the harmonic tremor is due to water flowing through cracks in the glacier. When water starts draining from the lake, the pressure of water in the cracks causes them to vibrate at a measurable frequency, much like organ pipes.

From six thousand miles away, David Heeszel from the Scripps Institute of Oceanography at the University of California–San Diego wrote the code for transforming raw data into spectrograms. “The seismic approach gives us new insights into the plumbing system of the glacier,” says Heeszel. “Seismology is a perfect tool for looking inside a glacier.” Especially as the recording equipment becomes cheaper and more portable, seismology is becoming an increasingly attractive option to glaciologists.

The spectrograms reveal patterns like those obtained from hydraulic fracturing, also known as fracking. This is because in both glaciers and fracking, pressurized fluids enlarge cracks, creating different frequencies of vibrations when water flows through them.

“We want to be open-minded about looking to other fields for good ideas and applying them to our field,” says Heeszel. The parallel of glacial hydrology to fracking is a reminder that great science applies to more than one field.

Understanding and predicting water flow in glaciers

In addition to seismologic methods, glaciologists can measure the volume of water flowing from the glacier. This way, they can determine how long it takes water to drain from the glacier.

In this hydrologic approach to understanding glacial lake drainage, Mauro Werder from the University of Bristol, UK, focuses on how quickly the lake fills and drains. For his post-doctoral work, published in the Journal of Glaciology, he and his colleagues compared the timeline of outburst events every summer from 1950 to 2005, with detailed analysis from the summers of 2004 and 2005.

This study found a long-term trend towards flooding earlier in the summer, but otherwise there was no pattern. “The only constant in this field is change,” says Werder.

Lake basin after drainage. Photo credit to Mauro Werder,  University of Zurich.

Lake basin after drainage. Photo credit to Mauro Werder.

“The goal is to understand how the floods work to mediate the potential hazards, but it can be difficult to make sense of measurements. The mechanisms of glacial outburst floods are still not well understood.” Right now, he says, glaciologists are unable to predict the timing and volume of outbursts.

Gornersee is a perfect example of why glaciologists find it difficult to predict the next flood. The lake drained in four different ways over four different field seasons. The first year, 2004, flooding started via a process called ice-dam flotation. The water dammed by the ice increases in pressure until the water makes the glacier float. This drainage process can happen when the lake depth is 90 percent of the ice thickness. The second year, water tunneled through cracks in the glacier, which Walter says were left over from previous years. The third year, water overflowed the glacier like an over-full bathtub. The fourth year, it was a combination of crack and overflow drainage. “Four different draining mechanisms in four years––this is why it’s so difficult to predict the onset of flooding,” says Walter.

The first sign of lake drainage is the appearance of clean ice along the rim of the glacier. This is part of the reason why Walter suggests that the best bet for a warning system at Lake Gorner would be to place a camera next to the lake and watch it remotely.

Another indicator of the beginning of the lake drainage process at Gorner is calving. Huge chunks of ice up to 800 meters in width can break off from the glacier and rise to the surface of the lake. Later, when the lake finishes draining, the basin lies covered with these calved icebergs.

“Once drainage begins, it takes about a day for the water to emerge from the glacier,” says Werder. “We call it an outburst, but it’s not particularly dramatic.” The community down-valley from the glacier, Zermatt, currently harnesses meltwater to produce hydroelectric power. Even so, a large flood would pose a threat to this populated residential area.

This is exactly what happened this past July in Juneau, Alaska. Meltwater from Suicide Glacier threatened local infrastructure by swelling the Mendenhall River from 75 cubic meters per second to 476 cubic meters per second at its peak. The town, which lies 22 kilometers from the glacier, sustained minimal damage, but several roads were forced to close. Scientists from the University of Alaska Southeast are currently monitoring the glacier using sensors submerged in Suicide Basin.

The floods marked the fourth summer in a row that Juneau has experienced glacial outburst floods. The floods started in 2011 as a result of the increased amount of glacial meltwater at Suicide Basin. Elsewhere, ice dams can also melt until they become too weak to hold back their lakes. It’s clear that the activity of glacial lakes depends in part on climate; as the Earth warms, glaciers are melting, and outburst floods are expected to become more frequent.

The jökulhlaup

Compared to Alaska, Iceland experiences outburst events that are larger and more spectacular. The land is relatively unpopulated, so fatalities are uncommon, but the floods still pose a challenge to civil engineers.

The Icelandic Glaciers research project, funded by the Earthwatch Institute, seeks to understand how water flows through glaciers. Most glaciers drain their lakes when water erodes a tunnel underneath the glacier. During this process, water flowing through tiny cracks in the ice enlarges them as friction from the movement of the water melts the ice. Eventually, the cracks weaken the ice until it can no longer hold back the dammed lake. The word for these outburst events in Icelandic is “jökulhlaup” (how to pronounce), which literally means “glacier burst.”

But sometimes the glacial lake drains when sudden volcanic activity underneath the glacier melts a large amount of ice. In the summer of 1996, a jökulhlaup burst from the Vatna glacier, the largest in Europe, when volcanos underneath the glacier erupted. The resulting flooding damaged local infrastructure such as power lines, roads, and bridges, at a cost of 9 or 10 million dollars. Andrew Russell, the spearhead of the Icelandic Glaciers research project, likens the event to a flash flood because of its unpredictable nature.

A second outburst also took scientists by surprise when it burst from the Solheim glacier (Sólheimajökull) in July 1999, about 160 kilometers from Vatna. Once again, heat from a volcano under the glacier rapidly melted the ice.

This volcano is Katla. It erupts roughly every 50 years, but the last major blast was almost one hundred years ago, leading geologists to believe that it’s “due” for another eruption. The 1999 incident seemed like the rumblings of an awakening giant, and glaciologists are now closely monitoring seismic activity in the area. A larger eruption would melt enough of the glacier to wash out main roads, cutting off the east side of the island from the west for months. The flood resulting from the last major eruption swept up boulders up to 11 meters in diameter.

While glaciologists continue to watch the formidable volcano in Iceland for eruption potential, scientists in Switzerland are not optimistic about the future of an effective early-warning system for floods triggered by seasonal melting. Heeszel believes that while it is technologically feasible to create such a system, it would take a lot of resources and work on behalf of researchers. “It’s a huge challenge.”

Citations

Clarke, G. K. C., Matthews, W. H. and Pack, R. T., 1984. Outburst Floods from Glacial Lake Missoula: Quaternary Research, v. 22, pp. 289-299.

Evans, Elena. Jökulhlaups past and present: Key Concepts in Geomorphology.

Forgey, P., 2014. Juneau’s Mendenhall River crests after glacial outburst; flooding considered minor: Alaska Dispatch News.

Heeszel, D. S., Walter, F. and Kilb, D. L., 2014. Humming glaciers: Geology, v. 42, pp. 1099-1102, http://dx.doi.org/10.1130/G35994.1.

Huss, M., Bauder, A., Werder, M., Funk, M. and Hock, R., 2007, Glacier-dammed lake outburst events of Gornersee, Switzerland: Journal of Glaciology, v. 53, pp. 189–200, doi:10.3189/172756507782202784.

Roberts, M. J., 2005. Jökulhlaups: A reassessment of floodwater flow through glaciers:Reviews of Geophysics, v. 43, doi:10.1029/2003RG000147.

Russell, A. J., Tweed, F. S. and Knudsen, Ó., 2000, Flash flood at Sólheimajökull heralds the reawakening of an Icelandic subglacial volcano: Geology Today, v. 16, pp. 102–106. doi: 10.1046/j.1365-2451.2000.00005.x.

Russell, A. J., Tweed, F. S., Roberts, M. J., Harris, T. D., et al., 2010, An unusual jökulhlaup resulting from subglacial volcanism: Quaternary Science Reviews, v. 29, pp. 1363-1381, http://dx.doi.org/10.1016/j.quascirev.2010.02.023.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.