THE TIMING OF MISSOULA FLOODS ON THE COLUMBIA PLATEAU: IMPLICATIONS FOR THE AGE OF GRAND COULEE

Pleistocene floods from glacial Lake Missoula carved the Channeled Scabland landscape of eastern Washington, USA. The floods altered topography by eroding loess and incising canyons along several routes crossing the Columbia Plateau. Although recent work has advanced our understanding of Missoula flood chronology, the timing of flooding in some of the main channels that cross the plateau is unresolved, as is the timing of canyon incision. In particular, there is debate about whether Grand Coulee, the largest canyon in the Channeled Scabland, incised during or prior to the last glaciation. We will report ¹⁰Be exposure ages from granitic erratics collected in the Grand Coulee and Moses Coulee canyons, and from the less-incised Cheney-Palouse and Telford-Crab Creek scabland tracts, all of which were eroded by floodwaters that spilled from the Columbia River and its tributaries across the plateau. Our results indicate floods eroded Moses Coulee at 15–16 ka. Erratics on the east rim of Grand Coulee, approximately 200 m above the canyon floor, yield a similar 15–16 ka age, which corresponds to the time Grand Coulee was being incised via headward waterfall retreat by floods with modest discharges of <5 million m³ s^-1. Exposure ages for samples from the floor of Grand Coulee appear slightly younger than those on the canyon rim. Erratics from the head of the Telford-Crab Creek tract date to 14–15 ka. Moses and Grand Coulee were both incised when the Cordilleran ice sheet crossed the Columbia valley, causing floods to spill out of the ice-dammed Lake Columbia and across the plateau. Field relations indicate Moses Coulee was incised prior to incision of Grand Coulee, and our exposure dates suggest the two canyons were eroded in very close succession. Moreover, consistent with constraints from published hydraulic modeling and exposure ages, our results indicate Grand Coulee was incised during the last glaciation, rather than during a prior glaciation. Our results show that huge floods rapidly modified topography and dynamically altered flood routes on the Columbia Plateau, and can be used to constrain the timing of discharge in different flood channels.