Lake Agassiz and Lake Missoula were glacier-dammed lakes that periodically overflowed the lowest sill of their basins to cause massive floods. Lake Agassiz is thought to be responsible for the deeply incised Minnesota River valley.

The shifting of glaciers and the ice sheet altered the effective relief of the basins over time, so simulating these lakes today takes a lot of work.

In the Great Basin, away from glaciers, Lake Bonneville and Lake Lahontan were the largest lakes that formed during a colder, wetter time, as endorheic valleys filled with water and repeatedly overflowed the lowest sill of their basins. Lake Bonneville is thought to have overflowed the outer edge of Great Basin itself at Red Rock Pass near Downey, Idaho; eroding the gap and releasing a huge flood into the Snake River basin.

Because a moving ice sheet wasn't a factor in their case, DEM shading can be used to approximate their overflows. This doesn't account for isostatic rebound or tilt in terrain, but gives a visually enlightening approximation, and lets one interactively explore how these outflows may have worked.

One can shade a topo map of the Great Basin at 4785 feet -- today's elevation of Red Rock Pass -- to approximate how Lake Bonneville's floods may have worked. Or, shade at these key sill elevations, in feet, to see how Lake Lahontan may have outgrown one valley after another: 3878, 3933, 3976, 4154, 4180, 4301, 4386.

Example with Caltopo at 3976 feet: https://caltopo.com/map.html#ll=40.28729,-118.02612&z=8&b=t&...

> The last major shift in drainage occurred around 8,200 years ago. The melting of remaining Hudson Bay ice caused Lake Agassiz to drain nearly completely. This final drainage of Lake Agassiz has been associated with an estimated 0.8 to 2.8 m (2.6 to 9.2 ft) rise in global sea levels. [...] A recent study by Turney and Brown links the 8,500-years ago drainage to the expansion of agriculture from east to west across Europe; they suggest that this may also account for various flood myths of prehistoric cultures, including the Biblical flood narrative.

Reading things like this is very humbling. Humankind has been good at keeping records of things for the last, what? 300 years? Nearly by accident, we have successfully retained some accounts from the ancient Greeks, which were produced not even 3000 years ago. So 3000 years is a long-ass time, and really stretches our capabilities of preserving records. And this was after we invented some forms of writing, as far as I understand it.

Yet historians have reason to believe that in the 5000 something years before that, humans told stories about actual great floods, passing them down one generation at a time, to a point where they still form part of our cultural heritage today? That is absolutely mind-boggling to me. Obviously, I can't tell whether or not that is actually the case, but that there's even a possibility that it is the case is just beyond me.

Of course, for the generations living around 8,500 years ago, even just a 3 ft flood must have really been something if it happened all over the world at once – even if it takes multiple generations for it to reach peak levels. That I can sort of get my head around. Obviously, you talk about that with your contemporaries. But that it would have been significant enough that it would somehow make it into folklore that two hundred generations later, they still talk about it, that just... no words.

Agassiz fell and rose multiple times over thousands of years as the ice sheet retreated and regrew. During that time it found many paths to escape its confinement, and left behind several beach-lines.

One of the most clearly-written and illustrated professional Agassiz stories is Wright's lightly-technical 1990 geology paper, available in PDF [0]. A lot more has been learned in the past 30 years, but by that time its effects in Minnesota were quite well known.

[0] https://conservancy.umn.edu/handle/11299/57272 Wright, H.E. Jr (Minnesota Geological Survey, 1990)

Lake Missoula wasn’t as big, but it’s floods were more spectacular.

https://en.wikipedia.org/wiki/Missoula_floods

“After Pardee studied the canyon of the Flathead River, he estimated that flood waters in excess of 45 miles per hour (72 km/h) would be required to roll the largest of the boulders moved by the flood. He estimated the water flow was 9 cubic miles per hour (38 km3/h), more than the combined flow of every river in the world. Estimates place the flow rate at ten times the flow of all current rivers combined.”

There was a similar lake in Siberia: https://en.wikipedia.org/wiki/West_Siberian_Glacial_Lake

It likely was so dammed up it overflowed back down to the south, through the Aral, Caspian, and Black Seas, to the Mediterranean: https://folk.uib.no/ngljm/PDF_files/Mangerud_et_al_2004,_QSR...

I'm sure I remember reading a good article about it in National Geographic or Scientific American or something, but I can't find it.

Fascinating! Also worth a look is Lake Lahontan, which covered much of northern CA and southern OR about 12,000 years ago: https://en.wikipedia.org/wiki/Lake_Lahontan

Interesting reading. As a lifelong alpine climber with intimate knowledge of glaciers, I always get weirdly sad at the thought of receding ice sheets (no matter the reason or epoch.)