Giant landslide triggerings and paleoprecipitations in the Central Western Andes: The aricota rockslide dam (South Peru).

• The Aricota landslide dam is an enigmatic giant landslide of the hyper arid Andes • Its rock volume was 2 km3, producing a dam of 600 m height and a lake of 6 km long • 13 new 10Be exposure ages show two main events at 17.9 ± 0.7 and 12.1 ± 0.2 ka • This timing is consistent with periods of regional paleoprecipitation increases The central part of the Western Andes holds an exceptional concentration of giant paleolandslides involving very large volumes of rock material (v > km3). While those gravitational slope failures are interpreted consensually as an erosional response to the geodynamic activity of the Andes (relief formation and tectonic activity), the question of their triggering mechanisms remains enigmatic. To clarify the respective roles of climatic versus seismic forcing on the Andean landslides, new temporal constraints on paleo movements are essential. Here, we focus on one of those giant slope failures, the Aricota giant rockslide that damned the Locumba valley in southern Peru. We conducted fieldwork, high-resolution DEM analysis and cosmogenic nuclide dating to decipher its development history and failure mechanisms. Our results point to the occurrence of two successive events. A giant failure mobilizing a rock volume of ca. 2 km3 first produced a dam at 17.9 ± 0.7 ka. Considering its height of ca. 600 m, the Aricota rockslide dam is one of the three largest landslide dams worldwide. At 12.1 ± 0.2 ka, a second event produced ca. 0.2 km3 of material, and the rock-avalanche debris spread out over the dam. As the chronology of those two events is pointing to the two main paleoclimatic pluvial periods in this region (Heinrich Stadial 1a and Younger Dryas), we favor the interpretation of a climatic forcing. At a regional scale, the concomitant aggradation of alluvial terraces and fan systems along the nearby valleys highlights higher regional erosion, sediments supply and mass-wasting events during those paleoprecipitation events and strengthens this conclusion. [ABSTRACT FROM AUTHOR]