QUATERNARY MAPPING AND PALEOSEISMOLOGY AS A TOOL TO PREDICT EARTHQUAKE RUPTURE LENGTH AND DIRECTIVITY: OBSERVATIONS FROM JAPAN AND ELSEWHERE AROUND THE GLOBE

This paper proposes hypothetical methods to predict segmentation of fault systems and directivity of their rupture propagation based on geometric criteria such as branching of fault traces and dip-slip distribution along the fault traces. Defining fault-zone segments and evaluating the reliability of individual segment boundaries to stop propagating ruptures are essential for accurate size estimation of future earthquakes generated from active faults. Direction of rupture propagation is closely related to strong ground motions and resulting earthquake damage. Therefore, predicting earthquake rupture length and directivity are crucial in mitigating earthquake damage. However, they were mostly ascertained only after earthquakes from the observed seismological records and not before the earthquakes. Nakata and Goto (1998) proposed new criteria for identification of segments for active strike-slip fault systems based on geometric surface rupture heterogeneity; the characteristic pattern of vertical-slip distribution along strike-slip faults and fault branching. Regarding dip-slip distribution along strike-slip faults, up-thrown sides of the faults along strike-slip faults are, in general, located on the fault blocks in the direction of relative strike-slip. For example, along an E-W trending right-lateral strike-slip fault, up-thrown sides are located on the north in the eastern section and on the south in the western section. Therefore, a fault segment may be identified based on a set of the vertical slip distribution. If branching of two faults faces each other, we may expect a segment boundary between these branching. Nakata and others (1998) found an interdependent correlation between the branching of the surface ruptures and the direction of their propagation from an investigation of recent earthquake faults. The branching of faults during rupture propagation is regarded as an effective energy dissipation process and could result in final rupture termination. We applied these criteria successfully to several prominent recent earthquake fault ruptures as well as active faults in Japan and elsewhere around the globe.