- The two largest earthquakes ever recorded, the 1964 M[subscript w] 9.2 Alaskan and 1960 M[subscript w] 9.5 Chilean, occurred on seismogenic plate interfaces at subduction zones. It has been theorized that the catastrophic failure of a locked zone along the contact between the downgoing slab and the upper plate causes these earthquakes, although determinations of
the position, attitude and extent of this locked zone vary from model to model. Four methods used to constrain the positions of the locked zones are: (1) historical great earthquake rupture extents, (2) heat flow/thermal profiles along the seismogenic plate interface, (3) patterns of surface deformation across the subduction zone forearc, and (4) spatial patterns of upper plate seismicity. Secondary parameters, such as subducted sediment thickness, upper plate lithology, and dip angle of the subducting slab likely play a role in locked zone location as well. In addition to a locked zone, the upper plate of most subduction zones is marked by paired inner and outer forearc highs and basins between the deformation front (trench) and the volcanic arc. Although such surface morphological features are easy to recognize, their spatial and geometric relationships to the locked zone have not been investigated systematically. This thesis investigates correlation between the spatial position of these morpho-tectonic features and the underlying locked zone at the Aleutian, Alaskan, Cascadia, Costa Rican, Javanese, Sumatran, Nankai, and Southern Chilean subduction zones. For all subduction zones other than Cascadia, which has yet to experience a great earthquake in historical times, the applied means of determining the position of the locked zones place them on plate interface regions between the inner and outer forearc highs. A strong correlation exists between dip of the downgoing plate and the width of both the locked zone and the spacing of the forearc morphologic elements for each of the subduction zones examined. The concept of comparative subductology is updated and enhanced in this study by creating GIS databases incorporating geological, seismological, geodetic, and geophysical observations. Correlations between surface morphological features and geologic and geophysical observations provide insight into controls on the position of the locked zone responsible for great earthquakes within the eight subduction zones examined, indicating that forearc morphology and interplate coupling are related via basic subduction parameters and the structural-tectonic regime of the forearc region.