Building on Perron & Venditti (2016)
In their 2016 Nature News & Views article titled "Megafloods downsized", Perron and Venditti remind us that the sizes of individual scabland megafloods remains uncertain. They question the assumption that every flood filled Scabland channels to the brim.
I am sympathetic to their ideas. Fluvial systems incise over time and leave behind flights of terraces. Vestiges of terraces occur in the Channeled Scabland. Geomorphic rules hold even in megaflood landscapes. In places, sets of flood cut surfaces separated by escarpments are ordered and mappable.
Figure 1. Early scabland floods occupied numerous, small channels following courses through a hilly landscape of virgin loess and basin fill deposits. Over time, incision, headcutting, and stream capture reduced the number of floodway channels and simplified the network to a few, large channels. The water of the latest floods was carried by really big coulees and bedrock canyons. Thick, dryland soils - loesses with thick petrocalcic horizons - fill the bottoms of many old flood coulees that were long ago abandoned by the evolving flood-channel network. The topographic expressions of the old coulees (sinuous declivities) remain in the landscape, if subtle. Paleosols preserved along the margins of scabland areas provide an important key to deciphering how the ancient floodway evolved.
By mapping flood-cut surfaces, coulee-bottom sediments, and soils at a few key localities using high-resolution lidar topography and fieldwork, could we reveal how portions of the floodway evolved? Repeated floods would certainly have focused incision in a few places. And once a bedrock canyon is established, later floods down that path would have no choice but to follow it. The paleotopography and the paleosol record preserved in the swath between Channeled Scabland and western Palouse Slope would seem a productive place to start. Calcic paleosols might be used to differentiate flood-cut surfaces that grow younger as they lose elevation (i.e., Othberg, McDaniel, and Fosberg's work on terraces of the western Snake River Plain).
Figure 2. Shallow channels through loess hills were cut by small floods (<50m) at relatively low velocities (<5 m/sec), as shown by Benito (1997, Fig. 6). Early scabland floods pioneered such channels (Time 2 and 3 in first figure), many of which were abandoned and later partially filled by loess. Thick loess and stacked paleosols indicate deposits filling the old coulees (Time 3, 4, 5) are much older than the latest Missoula flood cycle (>20ka).
Perron and Venditti's article represents out-of-the-box thinking by modern Geomorphologists. We know the Ice Age floodway evolved though time, so how did it happen?. What did the step-by-step changes look like in two and three dimensions? Think about the elevation of the channel bed, not just the maximum height of the floodwater energy surface. Scabland channel beds lowered with incision through time, didn't they? Progressive lowering of the bed and coincident hillslope retreat creates volume. New floods flow at lower elevations. Younger trimlines should occur at progressively lower elevations as well. Older, higher, abandoned surfaces should show evidence of exposure, weathering, and soil development (i.e., terraces). This pattern occurs along the Scabland-Palouse interface and likely elsewhere.
Time to get back to basics. Time to move beyond the old narratives.