Pleistocene Glaciation and Diversion of the Missouri River in Northern Montana

Student report by

William Moak, Department of Geography and Geology, University of Nebraska at Omaha.

In late May of 1805, the Lewis and Clark expedition worked its way up the Missouri River in northern Montana-the same section of river that we paddled down almost 200 years later. Lewis and Clark had set out from the mouth of the river at St. Louis to explore the newly acquired Louisiana Territory and search for a water route to the western ocean. During the Pliocene (5-1.8 million years ago), however, like-minded explorers would have set out from Hudson Bay to reach the river in Montana. The continental drainage divide during the Pliocene extended diagonally-southwest to northeast-across South Dakota (Wayne et al., 1991). Streams northwest of the divide drained into the Arctic Ocean while streams to the southeast flowed to the Gulf of Mexico. But during the Pleistocene (1.8-0.01 Mya), the Laurentide ice sheet pushed into Montana from the north and the Missouri and other streams were diverted from their beds. Streams were initially dammed up by the ice front and then flowed around the glacier.

As we set out from Coal Banks Landing (river mile 41.5), the Missouri River was confined to a narrow valley, perhaps 1-2 kilometers wide. This section of the river is called the "White Cliffs" after the massive, light-colored bluffs of Eagle Sandstone that line the river. However, the river valley widens near Judith Landing (river mile 88.5). The bluffs are set back from the river, rounder, and less imposing. The different character of the river valley results from differences in age: the narrow valley between Coal Banks Landing and Judith Landing is younger than the valley on the east side of Judith Landing.

The upper great plains were deposited during the Miocene (~24-5 Mya) by aggrading streams flowing eastward out of the Rocky Mountains. But during the late Miocene and the Pliocene, these streams began to dissect the Great Plains in response to regional uplift (Wayne et al, 1991; Trimble, 1980). The preglacial surface was more sculpted than it is today and had greater relief (Lemke et al, 1965). In central Montana, the preglacial Missouri River flowed northeast from Virgelle across the Big Sandy to Havre where it followed the present-day course of the Milk River (Figure 1). East of Glasgow, the preglacial Missouri River rejoined its current course where the Milk River today empties into the Missouri. But further east, near Poplar, the preglacial Missouri again diverted from its current path, flowing again to the northeast and eventually into Hudson Bay (Figure 2).

Figure 1: Course of preglacial Missouri River between current Missouri River and Milk River. Heavy black lines show width of preglacial river valley. (From Alden, 1932).

Figure 2: Major preglacial drainage courses. (From Lemke et al, 1965)

Howard (1958) argues that the ancestral Missouri, Yellowstone, and Little Missouri rivers all drained to the northeast prior to Pleistocene glaciation (Figure 3). He points to upland swales (low ground) which trend to the northeast and have the proper dimensions to accommodate the ancestral rivers. These swales contain fluvial gravels and are parallel to river-cut benches of Miocene age. Howard also argues that similar fluvial pebbles found in glacial till northeast of known gravel exposures must have been deposited by ancestral streams; glacial ice could only transport material southward in the direction of the advancing ice. Howard attributes variations in the width of the current Missouri River valley to changes in its postglacial course from preglacial river valleys. Assuming the regional sedimentary rock is uniformly resistant to erosion, older valleys should be wider than younger valleys. Accordingly, broad valleys occur where the river has reoccupied preglacial river valleys and narrow river valleys indicate where the river has cut a new path as a result of glacial diversion. Finally, Howard argues that barbed tributary streams, wind gaps (streamless cuts), and acute bends in current river paths all suggests postglacial realignment of river courses.

Figure 3: Ancestral Missouri, Yellowstone, and Little Missouri river valleys in eastern Montana and western North Dakota. Map also indicates current drainage pattern. (From Howard, 1958)

Alden (1932) states that segments of the Missouri River were rerouted by the Keewatin ice sheet when it pushed into Montana during the Pleistocene. As the glacier advanced, it blocked the course of the river forcing its waters to pond to the west. This created a series of glacial lakes (Figure 4)-Cutbank, Choteau, Great Falls, Musselshell, Jordan, and Glendive-along the Missouri River and its tributaries (Lemke et al, 1965). South of the ice, these lakes were confined by highlands to the east. For instance, Glacial Lake Great Falls formed west of the Highwood Mountains (Figure 5). Eventually, however, the water level raised high enough to spill over the natural barrier and carve out a new channel. Today this abandoned diversion channel is known as the Shonkin Sag (Figure 6). On the drive up to Coal Banks Landing along Montana Highway 80, we noted a deep, wide, and abandoned river valley that, in retrospect, was probably the Shonkin Sag. Trimble (1980) suggests that much of the Missouri River's present course between Great Falls, Montana, and Kansas City, Missouri, was established as an ice-marginal channel.

Figure 4: Pleistocene glacial lakes (speckled pattern) in Montana. (From Lemke et al, 1965)

Figure 5: Depiction of Great Falls Glacial Lake and Shonkin Sag during Pleistocene glaciation. (From Alden, 1932).

Figure 6: Shonkin Sag circa 1932. (From Alden, 1932)

Pleistocene glaciation occurred in several waves separated by interglacial stages when the ice sheet partially receded. Each subsequent glaciation tended to obliterate evidence of earlier glaciations. Therefore, reconstruction of glacial history is difficult. Traditionally, four glaciations were recognized: Nebraskan, Kansan, Illinoian, and Wisconsinan (oldest to most recent). Today, however, geologists conclude that there were many more episodes of glaciation than previously thought. The Missouri River was almost certainly affected by each glacial advance that reached its banks. Geologists can reconstruct the changing course of the Missouri River by creating a time sequence of events. Evidence such as stream channels and benches, fluvial deposits, glacial moraines, and glacial lake sediments can be aged absolutely or relatively. Absolute age refers to a specific time, e.g., 15,000 years ago. Relative time places an event in a sequence of events-after some events but before others. Such an endeavor is not easy, however. Evidence is often lacking and sometimes frustratingly contradictory. Alden (1932) mapped out the course of the Missouri River at four distinct stages: preglacial (Figure 7); maximum extent of the Illinoian or Iowan (?) ice advance (Figure 8); early Wisconsin ice advance (Figure 9); and postglacial (Figure 10).

Figure 7: Preglacial drainage pattern in northern Montana. (From Alden, 1932)

Figure 8: Drainage pattern in northern Montana at height of Illinoian or Iowan (?) glacial advance. Preglacial drainage pattern indicated in dashed lines. (From Alden, 1932)

Figure 9: Drainage pattern in northern Montana during early Wisconsinan stage. (From Alden, 1932)

Figure 10: Postglacial drainage pattern in northern Montana. Sections of preglacial drainage pattern indicated in dashed lines. (From Alden, 1932)

Today we can see evidence of the course changes that the Missouri River underwent as a result of Pleistocene glaciation (Figure 11). The acute turn of the river to the southeast at Virgelle, narrow versus broad river valleys, the Shonkin Sag, and barbed tributary streams such as the Judith River all suggest that the course of the Missouri was different in the past. Imagine how different life would be, in Omaha for example, if the Missouri River still emptied into Hudson Bay today.

Figure 11: Physiology and glacial geology of eastern Chouteau and western Blaine counties in northern Montana. (From Alden, 1932)


Alden, W. C., 1958, Physiography And Glacial Geology Of Eastern Montana And Adjacent Areas, U. S. Geological Survey Professional Paper 174.

Colton, R. B., Lemke, R. W., and Lindvall, R. M., 1961, Glacial map of Montana East of the Rocky Mountains, U. S. Geological Survey Miscellaneous Geological Investigations Map I-327.

Howard, A. D., 1958, Drainage Evolution In Northeastern Montana and Northwestern North Dakota, Bulletin of the Geological Society of America, v69, 575-588.

Lemke, R. W., Laird, W. M., Tipton, M. J., and Lindvall, R. M., 1965, Quaternary Geology Of Northern Great Plains, in Wright, H. E., Jr., and Frey, D. G., Eds, The Quaternary of the United States, Princeton University Press, Princeton, NJ.

Lindvall, R. M., 1962, Geology Of The Eagle Buttes Quadrangle, Chouteau County, Montana, U. S. Geological Survey Miscellaneous Geological Investigations Map I-349.

Thornbury, W. D., 1965, Regional Geomorphology Of The United States, John Wiley & Sons, Inc., New York.

Trimble, D. E., 1980, The Geologic Story Of The Great Plains, Geological Survey Bulletin 1493.

Wayne, W. J., Aber, J. S., Agard, S. S., Bergantino, R. N., Bluemle, J. P., Coates, D. A., Cooley, M. E., Madole, R. F., Martin, J. E., Mears, B., Jr., Morrison, R. B., and Sutherland, W. M., 1991, Quaternary Geology Of The Northern Great Plains, in The Geology of North America, Vol K-2, Quaternary Nonglacial Geology: Conterminous U. S., The Geological Society of America.