The Lake That Broke Its Chains:

Glacial Lake Missoula and the Floods That Remade a Continent

A LANDSCAPE WITHOUT MEMORY

Western Montana keeps few visible reminders of the strangest chapter in its geological past. Drivers on Highway 200 or hikers on Mount Jumbo above Missoula pass faint horizontal lines etched into the hillsides without necessarily registering what they represent. Those lines are shorelines, the ghost marks of a lake that no longer exists, one that at its height held more water than Lake Erie and Lake Ontario combined and then, more than once, let it all go in a matter of days. The story of Glacial Lake Missoula is not simply a footnote in the natural history of Montana. It is also a case study in how science corrects itself, sometimes slowly and sometimes stubbornly, when the physical evidence contradicts inherited assumptions.

THE ICE THAT BUILT A SEA

During the final advance of the Pleistocene ice age, a lobe of the Cordilleran Ice Sheet pushed southward through the Purcell Trench in what is now northern Idaho. This lobe of ice moved through the Purcell Trench and created an ice dam roughly 2,500 feet high across the Clark Fork drainage. Behind that frozen barrier, meltwater and river flow had nowhere to go but up and back into the valleys of western Montana. The resulting lake occupied the Mission, Jocko, and Little Bitterroot valleys, drained by the Flathead River, along with the Missoula, Ninemile, and Bitterroot valleys, drained by the Clark Fork River. At its greatest extent the lake covered roughly 2,900 square miles and held more than 500 cubic miles of water. That volume, spread across the intermountain basins of western Montana, is why later observers would compare it in scale to two of the Great Lakes at once, though the comparison appears mainly in popular accounts rather than in the technical literature and should be treated as an approximation rather than a precise figure.

The dam itself was not a permanent structure. Ice floats, and as water rose behind it, the dam eventually became buoyant enough, or fractured enough along its base, to fail. When it did, the entire impounded lake emptied through the gap at extraordinary speed. Sources differ on exactly how many times this cycle repeated. One early U.S. Geological Survey account concluded that the Purcell lobe had advanced and blocked the Clark Fork River at least four times, and perhaps as many as seven, with the largest lake forming around 18,000 to 20,000 years ago and its draining producing the last great flood. A National Park Service landmark description, by contrast, states more conservatively that at least three separate collapses of the ice dam, probably coinciding with periods of ice retreat, caused the lake to drain and produced catastrophic floods across the Columbia Plateau and down the Columbia River Basin. This is one of several points on which the record is not fully reconciled: the precise number of dam failures remains an estimate built from incomplete field evidence rather than a settled count, and readers should understand the “three,” “four,” and “seven” figures as competing approximations rather than as agreement.

RUPTURE AND RELEASE

When the ice dam gave way, the discharge was not a flood in any ordinary sense of the word. Geologists estimate that the waters of Glacial Lake Missoula, roughly 1,000 feet deep at Eddy Narrows, were withdrawn at a peak rate near ten cubic miles per hour, with approximately 380 cubic miles of water passing through the narrows during a single withdrawal event. The surge tore across what is now the Idaho Panhandle and the Columbia Plateau of eastern Washington, stripping soil down to bedrock and carving a maze of dry channels that would later be called the Channeled Scablands. Geologist David Alt, who spent decades studying the lake and its floods, described the impact of that torrent striking the ice-marginal Glacial Lake Columbia as comparable to an elephant jumping into a mud puddle. The image is vivid, but it is worth noting that it appears as Alt’s own characterization rather than as a measured finding, and it survives in the secondary literature mainly because it captures, in a single phrase, a collision of forces that is otherwise difficult for a general reader to picture.

Physical traces of that violence remain scattered across the outflow path. Among the most distinctive are giant ripple marks, arcuate ridges of gravel between fifteen and fifty feet high, one hundred to two hundred fifty feet wide, and up to a half mile long, formed on the floor of the draining lake near what is today Montana’s Camas Prairie. These ripples are, in effect, ordinary streambed ripple marks scaled up by a factor of a thousand, evidence of currents moving with a force well outside the range of any river flowing today.

THE SCHOLAR WHO WOULD NOT YIELD

The scientific recognition of this history did not come easily, and its slow acceptance says as much about the culture of early twentieth-century geology as it does about the flood itself. J Harlen Bretz, born in 1882 in Ionia County, Michigan, trained first in biology at Albion College before earning a doctorate in geology at the University of Chicago in 1913. Beginning in the summer of 1922 he undertook seven years of field research across the Columbia River Plateau, eventually producing fifteen papers on the region between 1922 and 1931. In 1923 the Journal of Geology published his paper “The Channeled Scablands of the Columbia Plateau,” in which he argued that this unusual landscape had been carved by a single, catastrophic flood he called the Spokane Flood.

The claim ran directly against the doctrine of uniformitarianism then dominant in geology, the assumption that landscapes are shaped gradually, by processes still observable today, rather than by sudden catastrophic events. The academic establishment was unpersuaded, and in 1927 the Washington Academy of Sciences convened a meeting specifically to have Bretz defend his hypothesis before his peers. He held his ground, but the missing piece of his argument, an actual source large enough to generate such a flood, remained unresolved for years. Bretz recognized that the boulders scattered across the scablands could only have traveled hundreds of miles if carried in rafts of glacial ice, which meant the floodwaters had to have been deep, fast, and immense in volume, yet he could not identify where that volume of water had come from.

PARDEE’S PROOF

The answer had, in fact, been building in Montana for over a decade, developed independently by a different Geological Survey scientist. Joseph Pardee, studying the Channeled Scabland and the intermountain basins of Montana as early as 1910, identified high water marks near Missoula that pointed to the former presence of a large glacial lake, and later, in the Camas Prairie of northwestern Montana, discovered giant ripple marks left by powerful currents that had once flowed across the lake bottom. Pardee did not immediately connect this Montana lake to Bretz’s Washington scablands, and the two men appear, by most accounts, to have coordinated only loosely for years. It was not until 1940 that Pardee formally demonstrated that Lake Missoula was in fact the source of Bretz’s flood. Had Bretz and Pardee communicated more closely earlier in their careers, they might have persuaded the broader geological community sooner, but full acceptance took much longer still, arriving only gradually through additional fieldwork in the 1940s, aerial photography in the 1950s, and finally satellite imagery in the late 1970s that made the scale of the flood undeniable. Bretz received the Geological Society of America’s Penrose Medal in 1979, by which point, as he reportedly told his son, all his scientific opponents had died. That anecdote appears in only one of the sources consulted for this account and should be read as a well-circulated detail of Bretz’s later life rather than as independently verified fact.

HOW MANY FLOODS? A DEBATE NOT SETTLED

Even after Bretz and Pardee were vindicated, a further question remained, and remains today only partly resolved: how many times did the lake fill and drain. The evidence for repeated flooding comes largely from rhythmites, layered sediment sequences left behind in backwater basins along the flood path, each layer presumably marking a separate flood event. Richard Waitt, a research geologist who examined varved sediment sequences in southern Washington, argued in a 1985 paper published in the Geological Survey of America Bulletin that more than forty successive flood-laid rhythmites had accumulated in back-flooded valleys, separated by layers of windblown loess and volcanic ash. Later work pushed the count higher still. Subsequent research using varves, burrows, mudcracks, and datable volcanic ash layers led to the hypothesis that there may have been as many as eighty-nine separate releases from Glacial Lake Missoula, though this higher estimate has itself been contested by other researchers and defended in turn by its original proponents. At the well-studied Ninemile locality near Missoula, roughly forty rhythmites can be identified, each consisting of lake-bottom varves overlain by a sand or silt layer marking a period when the lake bottom was briefly exposed to air, while the most complete record, in the Sanpoil Valley of northeastern Washington, appears to preserve as many as eighty-nine distinct flood events.

This is, in short, an argument that has never fully settled, and readers should understand the numbers forty and eighty-nine not as rival final answers but as figures tied to different sediment sequences studied by different research teams, each working from a different, and necessarily incomplete, stretch of the flood’s geologic record. What is not seriously disputed any longer is that Glacial Lake Missoula filled and emptied on a recurring basis over a span of several thousand years near the close of the last ice age, roughly between fifteen thousand and thirteen thousand years before the present.

READING THE LAND TODAY

The organizations most invested in preserving this record today are themselves regional rather than national in character. The Glacial Lake Missoula Chapter of the Ice Age Floods Institute has installed a series of engraved stone high water markers around the former lake basin and works closely with the Montana Natural History Center to interpret the site for visitors, having gained access to many of Joseph Pardee’s original papers and lantern slides in the process. Because the chapter’s focus lies specifically on the Montana side of the story, its programming tends to foreground Pardee’s contribution over Bretz’s, a regional emphasis that is itself worth noting as a difference in perspective rather than a difference in the underlying science.

What remains on the ground for a visitor is modest and easy to miss: faint terraces on the hillsides above Missoula, oddly regular gravel ridges on Camas Prairie, boulders far from any plausible source. None of it announces itself the way a canyon or a waterfall does. But the discovery of what produced these features, and the decades-long argument over how it happened and how often, stands as one of the more instructive episodes in American earth science, a reminder that even a landscape as quiet as western Montana’s valleys can hold the record of an event almost too large to credit.

WORKS CITED

Alt, David. Glacial Lake Missoula and Its Humongous Floods. Mountain Press Publishing, 2001.

Ice Age Floods Institute. “Glacial Lake Missoula Chapter.” Ice Age Floods Institute, https://iafi.org/chapters/glacial-lake-missoula-chapter/. Accessed 2 Aug. 2026.

Ice Age Floods Institute. “Missoula Flood Rhythmites.” Ice Age Floods Institute, https://iafi.org/missoula-flood-rhythmites/. Accessed 2 Aug. 2026.

“J Harlen Bretz.” Graph Search, École Polytechnique Fédérale de Lausanne, https://graphsearch.epfl.ch/concept/1245472. Accessed 2 Aug. 2026.

Reidel, Stephen P., and Karl R. Fecht, editors. “Section 4: The Missoula Floods.” Geologic Story of Grand Coulee, U.S. Geological Survey / National Park Service Online Books, https://www.nps.gov/parkhistory/online_books/geology/publications/inf/72-2/sec4.htm. Accessed 2 Aug. 2026.

“Who Was Harlen Bretz?” Spokane Historical, Eastern Washington University Public History Program, https://spokanehistorical.org/items/show/855. Accessed 2 Aug. 2026.

U.S. National Park Service. “Glacial Lake Missoula National Natural Landmark (Camas Prairie Ripples).” National Park Service, U.S. Department of the Interior, https://www.nps.gov/places/glacial-lake-missoula-national-natural-landmark-camas-prairie-ripples.htm. Accessed 2 Aug. 2026.

U.S. National Park Service. “People.” Ice Age Floods National Geologic Trail, https://www.nps.gov/iafl/learn/historyculture/people.htm. Accessed 2 Aug. 2026.

Waitt, Richard B. “Case for Periodic, Colossal Jökulhlaups from Pleistocene Glacial Lake Missoula.” Geological Society of America Bulletin, vol. 96, no. 10, 1985, pp. 1271-1286.