“…Soon after, Cloos () proposed that subducting seamounts were likely candidates for large earthquake triggers. Based on additional marine observations and physical modeling, alternative processes have been proposed involving multiple seismogenic behaviors depending on the type of subducting topographic features (Abercrombie et al, ; Bassett & Watts, , ; Bilek, ; Bilek et al, ; Carena, ; Cloos, ; Das & Watts, ; Dominguez et al, ; Geersen et al, ; Gutscher et al, ; Henstock et al, ; Kodaira et al, ; Konca et al, ; Kopp, ; Landgrebe & Müller, ; Marcaillou et al, ; Métois et al, ; Mochizuki et al, ; Morgan et al, ; Müller & Landgrebe, ; Robinson et al, ; Scholz & Small, ; Sparkes et al, ; Wang & Bilek, , ; Yang et al, ), the trench sediment thickness (Heuret et al, ; Jarrard, ; Ruff, ; Scholl et al, ), the state of stress within the upper plate (Heuret et al, ; Jarrard, ; Schellart & Rawlinson, ), the possible occurrence of tectonic erosion (Bilek, ; Sage et al, ; Scholl et al, ), the friction, normal stress, and fluid pressure along the subduction interface (Chlieh et al, ; Corbi et al, ; Lin et al, ; Ranero et al, ; Ruff, ; Saffer & Tobin, ; Saillard et al, ; Scholz, ) or the geometry or kinematics of the subduction zone (Bletery et al, ; Gutscher & Westbrook, ; Jarrard, ; McCaffrey, ; Schellart & Rawlinson, ; Uyeda, ). Some of these studies argue that along‐trench segments exhibiting low topographic roughness at long spatial wavelength should be prone to propagate ruptures over large distances and, consequently, be the location of very large earthquakes.…”