Late Pleistocene landslide-dammed lakes along the Rio Grande, White Rock Canyon, New Mexico

Massive slump complexes composed of Pliocene basaltic rocks and underlying Miocene and Pliocene sediments flank the Rio Grande along 16 km of northern White Rock Canyon, New Mexico. The toe area of at least one slump complex was active in the late Pleistocene, damming the Rio Grande at least four times during the period from 18 to 12 14C ka and impounding lakes that extended 10-20 km upriver. Stratigraphic relationships and radiocarbon age constraints indicate that three separate lakes formed between 13.7 and 12.4 14C ka. The age and dimensions of the ca. 12.4 ka lake are best constrained; it had an estimated maximum depth of nearly equal 30 m, a length of nearly equal 13 km, a surface area of nearly equal 2.7 km (super 2) , and an initial volume of nearly equal 2.5 X 107 m (super 3) . The landslide dam responsible for this lake was apparently stable, and the lake filled completely with laminated silt-rich lacustrine sediment and overlying coarse sands and gravels that represent bed load of the Rio Grande. A lake that formed at < or =17.5 ka was nearly equal 21 km long and apparently failed catastrophically, recorded by probable outburst flood deposits that extend downriver from the landslide dam. An older (> or =43 14C ka) landslide-dammed lake is less well constrained, but it had a maximum depth of at least 57 m, an estimated minimum length of 25 km, and a surface area of 25 km (super 2) . Deposition of coarse sands and gravels of the Rio Grande over lacustrine sediments indicates that the landslide dam responsible for this older lake also was stable; extrapolation of historic Rio Grande sediment yield data suggests that this lake persisted for 100-1000 yr. The stability of the dam may have been due to armoring of the outflow channel with basalt boulders. The youngest landslide-dammed lakes formed during a period of significantly wetter regional climate, strongly suggesting that climate changes were responsible for reactivation of the slump complexes. We are not certain about the exact triggering mechanisms for these landslides, but they probably involved removal of lateral support due to erosion of the slope base by the Rio Grande during periods of exceptionally high flood discharge or rapid incision; increased pore pressures associated with higher water tables; higher seepage forces at sites of ground-water discharge; or some combination of these processes. Seismic shaking could also have contributed to triggering of some of the landslides, particularly if aided by wet antecedent conditions.