An asperity model of large earthquake sequences

The variation in maximum rupture extent of large shallow earthquakes in circum-Pacific subduction zones is interpreted in the context of the asperity model of stress distribution on the fault plane. Comparison of the historic record of large earthquakes in different zones indicates that four fundame...

Full description

Bibliographic Details
Main Authors: Lay, Thorne, Kanamori, Hiroo
Other Authors: Simpson, David W., Richards, Paul G.
Format: Book Part
Language:English
Published: American Geophysical Union 1981
Subjects:
Online Access:https://authors.library.caltech.edu/51570/
https://authors.library.caltech.edu/51570/1/HKep81.pdf
https://resolver.caltech.edu/CaltechAUTHORS:20141111-105135960
Description
Summary:The variation in maximum rupture extent of large shallow earthquakes in circum-Pacific subduction zones is interpreted in the context of the asperity model of stress distribution on the fault plane. Comparison of the historic record of large earthquakes in different zones indicates that four fundamental categories of behavior are observed. These are: (1) the Chile-type regular occurrence of great ruptures spanning more than 500 km; (2) the Aleutians-type variation in rupture extent with occasional ruptures up to 500 km long, and temporal clustering of large events; (3) the Kurile-type repeated failure over a limited zone of 100–300 km length in isolated events; and (4) the Marianas-type absence of large earthquakes. Southern Chile, Alaska, Southern Kamchatka, and possibly the Central Aleutians are grouped in the first category. The Rat Island portion of the Aleutians, Colombia, Southwest Japan, and the Solomon Islands zones demonstrate the temporal variation of rupture length and multiple earthquake sequences that characterize category 2. The New Hebrides and Middle America have earthquake clustering on a more moderate scale, and are intermediate between categories 2 and 3. Category 3 includes the Kurile Islands, Northeast Japan, Peru and Central Chile. Zones lacking large earthquakes (category 4) include the Marianas, Izu-Bonin, and large portions of Tonga-Kermadec. By loosely grouping each subduction zone into these categories and comparing the general range in behavior with a simple fault model, which is used in a numerical simulation, the parameters governing large earthquake development are clarified. Interpretation of the four categories in terms of asperity distribution and interaction permits some inferences of the nature of stress distribution in particular zones. Two factors appear to dominate in the development of large earthquake failure zones; the nature and degree of coupling on the fault contact, and the extent of lateral segmentation of the subduction zone by transverse stress barriers. Strong coupling and uniform stress distribution on the fault plane produces larger events, whereas more heterogeneous stress distributions produce smaller ruptures and temporal variation in rupture length. Segmentation of the subduction zone that may result in stress barriers affecting rupture length is produced by subduction of transverse structures such as aseismic ridges, and is reflected by submarine canyons and geometric variations in trench configuration.