The Red River Valley

Tilted Shorelines and
Rebounding Lake beds
By: Don McCollor


Click on highlighted words for images
and additional information

The exceptionally flat terrain of the Red River Valley is a legacy of glacial Lake Agassiz waxed and waned with the advances and retreating of lobes of the Laurentide Ice sheet of the last ice age. With massive ice layers to the north, drainage was through the Minnesota River Valley as the ice advanced and toward Lake Superior as the ice retreated. By 8500 years ago, the lake had drained away and dried up completely.

Besides the flat floor and deep rich sediment of the lake bed forming the fertile farmland of the Valley, Lake Agassiz left other traces of its existence. Lake Winnipeg, Red Lake, and Lake of the Woods are remnants of the lake. Less impressive, but easily seen are several "beaches" or shorelines which delimitate the edges of the Valley. These are low ridges of sand and gravel representing "strandlines" when the level of the lake was stable. In places they are crossed by deltas extending into the lake plain marking where ancient rivers entered the lake. The two most well-developed beaches are the Herman Beach, formed 11,700 years ago and the Campell Beach, formed 11,200 years ago. Both can be traced for hundreds of miles through Minnesota and North Dakota. In Grand Forks Country, ND, the western beaches are in the Emerado area, and near Red Lake Falls in MN.

There is a curious feature about these beaches. The beaches at the northern end of the Valley near the Canadian border are higher than at the southern end at the namesake Minnesota towns of Herman and Campbell. Not that the beach features themselves are piled higher, but that they are higher above sea level. The slope is very gentle, impossible to observe without accurate surveying techniques. Even on a topographic map the change is not apparent until the fine print giving the elevations of the contour lines is scrutinized. Careful measurements show that the western Herman Beach is 180 feet (55 meters) higher at the Canadian border than at the southern end.

Clearly, something is amiss. The "level" in "water level" implies just what it says-that all parts of a body of water are at the same elevation. Of course there are variations around the average water level due to tides, storms, and seiches. Tides are negligible in landlocked water bodies smaller than oceans, and wave action due to storm winds intermittent. To anyone who has experienced the strength and persistence of the wind in the Valley, the image of the wind keeping the water pushed to one end of Lake Agassiz is almost in reach of the imagination. However, given the prevailing northwestern winds, the water and beach lines should be higher in the southeastern corner. Indeed the eastern Agassiz beaches show greater working indicating that the prevailing direction of severe lake storms was from the northwest. No natural force in our common experience explains the shorelines.

The answer to the tilted shorelines is a phenomenon called isostatic rebound. Basically, the land sinks when a heavy weight, like a glacier or lake water and sediments are place on it, and rises again when the weight is removed. An analog would be a heavy box placed on a vinyl car seat. Removal of the box after a few hours reveals a box-shaped depression in the seat. After a few hours the seat "rebounds" and returns to its original comfy contours. This rising and sinking of large land areas is a common concept in geology, as in the uplifting of a mountain range as erosion removes some of the weight. The concept of solid rock being springy like a car seat is difficult to imagine, especially in a geologically stable area such as the Valley. In places like California, where the landscape occasionally (and suddenly) moves up, down, and sideways; it is easier to grasp. Rock is stiffer, but it does deform and rebound, albeit over hundreds and thousands of years rather than hours. In the Lake Agassiz plain, the effect of the rebound is complicated by weight of water and lake sediments which accumulated as the ice sheet retreated. And although both the northern and southern ends of the beach line presumably raised, there is no reference point to measure this by. The only measurement possible is the difference between heights of the northern and southern beaches, which sets the minimum value for the rebound of 180 feet (55 meters).

The difference in elevation of the beach lines (and an estimation of how deformable the underlying rock is) has been used to estimate the minimum thickness of glacial ice in the Valley. In the Grand Forks area, the ice layer was probably between 920 feet (280 meters) to 3400 feet (1040 meters) thick. As the ice melted from south to north, the weight of lake water replaced some of the ice weight with an estimated 325 feet (100 meters) of water depth, delaying some of the rebound until the lake eventually drained. However, during this time up to 150 feet of sediment (46 meters) accumulated in the lake bed. This remained permanently, weighing down the underlying rock and preventing the completion of rebound from occurring. It is believed that rebound has been completed in the valley as far north as Lake Winnipeg, with the area further north still rebounding.

The phenomena of rebound with a general uplift of the terrain is of modern interest, especially since it is still occurring in the far north. The Red River of the North is a gently meandering river with a very low gradient (drop) of about 0.5 foot per mile (0.1 meter/kilometer) between Grand Forks and Pembina. The current gradient of the Red is now half what it would have been had rebound not occurred. (Of course without the glacier ice, Lake Agassiz and the current Red River water course would not have formed either.)The continued rebound in the Canadian north where the Red River water eventually reaches Hudson's Bay may have interesting future consequences. A decrease in gradient may make the Red (even) more prone to flooding. Or will Lake Agassiz return? Or perhaps large-scale canal building be needed in a few centuries to maintain the present drainage.


Ojakangas, Richard W., and Matsch, Charles L., Minnesota's Geology, University of Minnesota Press, Minneapolis, 1982.
Brevik, Eric C., Isostatic Rebound in the Lake Agassiz Basin Since the Late Wisconsinan, Masters Thesis, University of North Dakota, 1994.

A seiche (pronounced SAY-sh) is a periodic oscillation of a ocean, lake, or other body of water apparently caused by storms or changes in atmospheric pressure. The phenomenon is like the rhythmic rocking of water in a bucket when it is carried. In appearance, the seiche resembles a tide rather than a wave, with a change in water level of inches to several feet over a period of minutes to hours. Unlike a tide, it occurs unpredictably. A seiche can be particularly treacherous, snatching up shore fishermen and pier strollers, since the weather effect which causes it may be well out of sight, and the general rise of the whole water surface not immediately noticed.


Galley, Daniel V., U-505, Paperback Library, 3rd ed., New York, 1971, pp 288-289.

Storms on Lake Agassiz must have been awesome, being larger than the five present Great Lakes combined and with a fetch across hundreds of miles of open, shallow water for waves to build. Fresh water, lacking the salt content of ocean water is less dense, producing livelier, quicker waves. Waves on the Great Lakes are steeper and sharper than their rolling ocean cousins, leaping and tumbling. Waves thirty feet (9 meters) from crest to trough have been reported on Lake Superior, driven by seventy-knot winds (81 mph, 36 m/sec). A bleakly comforting sight during an ordinary Lake storm are the "Christmas trees"-steep triangular waves silhouetted against the horizon, appearing like a forest of pine trees. In a severe storm, the "Christmas trees" disappear as the wave tops are torn off by hurricane-force winds.


Ships Gone Missing, Hemming, Robert J., Contemporary Books, Chicago, 1992.
Great Lakes Shipwrecks and Survivals, Ratigan, William, WM. B. Eerdmans Publishing Co., Grand Rapids, Michigan, 1977.

The sinking and rebound of the earth is a factor to consider when large-scale engineering projects concentrate large masses, such as dams and associated reservoirs or removed large quantities of material in the course of open-pit mining and canal construction. During the building of the Panama canal, Culebra Cut, besides being noted for landslides (the most notorious being Cucaracha [the cockroach] Slide, named for its creeping habits) also had the floor of the Cut heave upward as over 100,000,000 cubic years of overburden to a depth of 260 feet was removed. The unstable rock of the Cut and the weight of the sides contrived to heave the bottom up twenty feet (6 meters) in several places, and thirty(9 meters) in one area. One section of the cut, complete with stream shovel sitting atop it lifted nine feet (2.7 meters) in one afternoon.


The Panama Canal, Hammond, Rolt and Lewin, C.j., Fredrick Muller Ltd., 1966.


Disclaimer and Copyright