DID HEADWARD EROSION COMPLETE THE GRAND COULEE EARLY IN LATE WISCONSIN TIME OR LONG BEFORE THEN?

Did carving of this great tandem canyon require multiple Columbia River diversions and megaflooding episodes over multiple Pleistocene glaciations, or was Grand Coulee mostly or entirely eroded early during the 20–15 ka last-glacial period by a few large Missoula floods? Several findings show the upper coulee floor had approached its elevation before much, if not all, of the last glaciation. (1) Striae and till on the coulee floor indicate that at its maximum last-glacial extent, the Okanogan lobe invaded upper Grand Coulee at or near its present depth near Steamboat Rock. (2) The level of last-glacial Lake Columbia, which the Okanogan lobe impounded for roughly 3 kyr, was usually controlled by the coulee floor near its present elevation. (3) The lobe’s occupation of the upper coulee suffices to explain higher lake levels that are evidenced by faint shorelines and glacial-lake stratigraphy. But these findings also allow a plausible scenario for floods from glacial Lake Missoula to have done much or all of the coulee erosion early in the last-glacial flood episode, during the time when the advancing the Okanogan lobe blocked the Columbia River, first shunting floods to Moses Coulee, and then eroding the Grand Coulee pathway as the Moses Coulee pathway was blocked by the continuing ice advance. Some evidence accords with this last-glacial erosion alternative: (1) Early last-glacial Missoula flood stages in the Columbia valley and Moses Coulee, as simulated in two-dimensional hydraulic models of the entire Missoula flood system, are only attained if no flow is diverted through Grand Coulee. (2) Unweathered last-glacial flood gravel in Hartline basin, 8 km northeast of Dry Falls, is devoid of granitic clasts, despite abundant exposed granitic basement on the coulee floor north of Steamboat Rock, suggesting flood erosion prior to basalt removal. (3) Basalt-erosion modeling permits kilometers of cataract recession by a few large floods. (4) No flood or Columbia River deposits indicative of pre-Wisconsin age have yet been found in the Grand Coulee. Whatever the chronology, much remains to be learned about the erosional history that the Grand Coulee so spectacularly records.