Story posted December 04, 2006
Geologist Peter Lea looks at a map the way a historian might scour an ancient manuscript for hidden meaning. The clues he seeks have nothing to do with their human context, however. His Da Vinci code is nested in the landscape beneath civilization, scripted by forces such as pressure, erosion, and water.
Sometimes the landscape has stories to tell that can change the way we see history. Even the history beneath our own feet.
Such was the case with Lea's recent research on sediment cores in Merrymeeting Bay, which is part of a long-term ecological study of this unique freshwater ecosystem spearheaded by professors and students in Bowdoin's Environmental Studies program.
As Lea began poring over local maps of the Kennebec and Androscoggin valleys in search of geological study sites, he began to notice a pattern of geographic anomalies that sent the associate professor of geology on quite another path of inquiry, what he terms his "curiosity-driven research."
The ciphers included evidence of ancient riverways, which, incongruously, seemed to have flowed up ridges. There were deep, unexplained trenches in modern waterways, including The Chops in Merrymeeting Bay and at Bradley Pond. The College itself seemed to play a part, with its campus, fields and forests resting upon an aquifer of hazy origins.
They all pointed to one thing: the presence of a subglacial meltwater channelway. Meltwater channels are large tunnels beneath a glacier, where melting water collects and, driven by ice pressure and gravity, cuts deep, narrow troughs in the earth.
"It began to look as if, just within a few miles of campus, we had a nicely displayed former meltwater channel," says Lea. "No one had identified that before, and although we know that meltwater has played a part in shaping the Maine coast, researchers have not really keyed in on the erosive ability of these meltwater rivers. Most of the research in Maine has focused on eskers — sand and gravel ridges that are the deposits of these river tunnels."
Lea decided it was time to "dig deep."
The first clue that caught his eye was a ridge of curiously dissected hills in Cathance, Maine, near Gardiner. They bear all the markings of rill development — streams that branch out in veinous patterns when water flows down a slope. But there are no streams there now: The nearest waterway is the Androscoggin River, which flows to the east, in the inherited path of a glacier. How, wondered Lea, could this also have been created by the glacier?
"If you believe this channelway exists, it has to leave the Kennebec near Gardiner, actually flow up over these hills and down into the Androscoggin then come up. One of the questions is, 'How do you get water to flow uphill?' We do know that water in glaciers tends to percolate down to the bottom and flow in tunnels along the bed.
"Why would a river tend to cut through a ridge rather than go around it?"
As it turns out, these questions were fairly easy to answer, with the aid of digital tools, such as Geographic Information Systems (GIS) and other data sources. Lea decided to do some digital modeling to try to replicate the ice sheet that might have caused this and other irregularities. He took coordinates of the modern terrain, then overlaid those with models of various possible ice sheets of differing depths and flow.
When he plotted the ice sheet over the Cathance hills, the basic physics of glaciation gave him the answer.
"Generally speaking, meltwater flows in the direction of the glacier," notes Lea, "and because of pressure, it flows from areas where the ice is thick to where the ice is thinner, or sloping. Because of gravity, the water stays beneath the ice."
But the bed slope over the Cathance hills proved to be more than ten times that of the ice surface — no matter how thickly Lea plotted the ice sheet to be. As a result, he says, the gradient pressure is likely to have caused hydraulic heads to form — geyser-like eruptions of meltwater through the ice surface. Through hydraulic heads, the meltwater channel flowed up and over the hilltops.
Once Lea had plotted the flow of water uphill, the rest became clearer:
"The pressure gradient would overwhelm the topography," he says, "and as they crossed hilltops, these meltwater rills spread out into interconnected channels, which flowed southwest to Cathance and to Bradley Pond. These small channelways coalesced and dug out deeper channels — you can see the trenches in the bedrock near Bradley Pond, and Cathance itself is in a broad channel."
Eventually, he posits, the channelway came together in a single valley, ending in what is now a buried sand and gravel aquifer in Rocky Hill, near Topsham.
"Brunswick has been getting some of its town well water from that aquifer for a long time," says Lea, adding: "Geologists in the past have noted that there must have been meltwater flowing through there at some point — hence the sand and gravel — but no one put it together in the larger context."
There is only one catch to this model: In order for this to have occurred, the glacier had to flow in a different direction than it is commonly believed to have moved, which is the south-southeasterly direction in which the Androscoggin and Kennebec rivers flow.
The Cathance Meltwater Channel Way, as Lea has named it, almost certainly flowed southwesterly.
"This tells us that at some point, exact point unknown, that ice sheet was not doing its normal S-S-E thing," says Lea. "If it was, the channel would just go right down the Kennebec instead of cutting through some fairly high topography. The only way you can get water to go up that channel is to turn the ice surface slope so the ice is flowing to the southwest significantly."
If Lea is right, then his findings present a directional shift in glacial research as well.
"Researchers sort of knew that there had been some form of southwesterly flow to the glaciers at some point," notes Lea, "but nobody had documented any large-scale land forms like these channelways that related to that. The prevailing thought was maybe it veered off to the southwest when the glacier was thinning and the topography started poking up.
"If my theory is correct, the Cathance Meltwater Channel Way is telling us that no, the topography was actually driving water of sufficient amounts in that direction to cut this fairly large channelway. It's telling us that glacial history was a little more dynamic and interesting than we thought in Maine."
Lea's findings are only a small piece in the larger story of our environmental history, he says, but they may offer researchers new clues about how the earth's climate system has worked in the past. He recently presented the research at a conference of the Geological Society of America in Philadelphia.
"This generally advances our understanding of the glacial history of Maine," says Lea, grinning a bit. "But with science, you don't know whether things are going to really take off and get picked up and taken further. Sometimes they just become a historical curiosity that sits out there.
"Still, it's satisfying just to think about how things got to be how they are. It's fun to think that when you drive to Freeport, that on parts of Route 1, you're actually driving on the pathways of one of these things."