REVISITING IDEAS ABOUT GIANT EARTHQUAKES, THE SUBDUCTION PROCESS, AND THEIR RELATION TO ANDEAN MOUNTAIN BUILDING IN THE LIGHT OF THE SUMATRA MEGAEARTHQUAKE

The December 2004 Sumatra earthquake, the largest in 40 years, gives new insights into the largest subduction zone earthquakes. It seems that earlier ideas about where such earthquakes can occur, based on convergence rate, age of the subducting lithosphere, and trench sediment thickness, do not apply. Many authors have investigated variations, with "coupling" defined in terms of the fraction of plate motion released in earthquakes or the occurrence of larger earthquakes. It has been proposed that the largest (Mw = 9) earthquakes occur only when young lithosphere subducts rapidly. However, much of the correlation vanishes using new data reflecting better known convergence rates (for example, Nazca-South America significantly slower than previously assumed) and reassessed earthquake magnitudes (including the fact that some large trench earthquakes are normal faults). This result has important implications for the relation between coupling at the Peru-Chile trench and Andean uplift. It has been suggested argued that the south Peru and north Chile trench segments, seaward of the high Andes, have higher shear stresses than segments to the south such as the south Chile segment where the 1960 Mw = 9.6 earthquake occurred, because Cenozoic climate change deprived these segments of sediment and thus raised stresses on the interface, causing uplift of the high Andes. However, we find that the size of the largest trench earthquakes and seismic coupling fractions depend at best weakly on trench sediment thickness. In particular, despite differences in sediment thickness, the south Peru and north Chile trench segments are similar in seismic coupling to segments to the south. They also have giant (Mw = 9) earthquakes, as paleotsunami analysis shows that 1868 earthquake ruptured farther to the north than traditionally assumed, implying Mw ~ 9.2. Hence it appears that the location of the high Andes is not controlled by unusually strong coupling.