JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 96, NO. B2, PAGES 2205-2221, FEBRUARY 10, 1991 Rupture Process of the February 4, 1965, Rat Islands Earthquake SUSAN L. BECK 1 Earth Sciences Department and Institute of Geophysics and Planetary Physics Lawrence Livermore National Laboratory, Livermore, California DOUGLAS H. CHRISTENSEN Geophysical Institute, University of Alaska, Fairbanks The great Rat Islands underthrusting earthquake (Mw = 8.7), of February 4, 1965, represents subduction of the Pacific plate beneath the North American plate along a 600-km segment of the western end of the Aleutian Islands. Body wave inversion techniques are used to determine the spatial and temporal heterogeneities associated with the Rat Islands earthquake. We have deconvolved World-Wide Standard Seismograph Network long-period teleseismic P wave seismogramsto obtain source time functions. Directivity associatedwith the three major pulses of moment release in the source time functions indicates a total source duration of 160 s, unilateral rupture in the direction 300 ø, fault length of 420 km, and averagerupture velocity of 2.5 km/s. The three pulsesof moment release are located along the fault, and these regions of high moment release are interpreted as asperities. The first asperity extends from the epicenter to 100 km to the WNW. This is the largest asperity and corresponds to a smoothpulse of moment release in the sourcetime function with a duration of 50 s. The secondpulse of moment release is very jagged, is less coherent between stations, and is centered -200 km WNW of the epicenter. The third pulse of moment release extends from 360 to 420 km WNW of the epicenter. Although the aftershock area is -600 km in length, we can not resolve any moment release from the P waves beyond -420 km WNW of the epicenter. The Rat Islands event was closely followed by a large tensional outer-rise event on March 30, 1965, (Ms - 7.5), which is located oceanward of the largest moment release associated with the Rat Islands mainshock rupture. Detailed analysis of the P waves for this large outer-rise event confine the depth extent to the upper 30-35 km of the crust. The spatial and temporal associationbetween the February 4 mainshock and the March 30 tensional outer-rise event suggeststhe tensional event may have been triggered by the large displacement near the mainshock epicenter. The overriding plate along the western Aleutian subduction zone is laterally segmented into a series of rigid tectonic blocks separated by fault controlled canyonsand extensionalbasins(Geist et al., 1988). We suggest that the central undeformed parts of the blocks have the strongest couplingwith the down-goingplate and hence are the sitesof the largestmoment releaseduring an underthrusting earthquake. The three asperities determinedfrom the P waves correspond to the Rat, Buldir, and Near tectonic blocks respectively. Hence the P wave seismicmoment release of the Rat Islands earthquake is controlled by the lateral segmentation of the overriding plate. INTRODUCTION Many of the variations associated with the occurrence of earthquakes can be explained by spatial heterogeneity in the mechanical properties of the fault plane [Kanamori, 1981]. In this model a fault consists of patches having different failure strengths. The heterogeneity in failure strengthsalong a fault segmentwill not only control the rupture of the largest earthquakes but also control the observed seismicity pat- terns. Recently many studies have identified temporal and spatial heterogeneity associated with the rupture of an individual earthquake [e.g., Beck and Ruff, 1984; Schwartz and Ruff, 1987; Boyd and Nabelek, 1988; Choy and Dewey, 1988; Houston and Engdal, 1989]. The asperity model is a simple fault heterogeneity model which provides a good framework in which to interpret the complex occurrence of earthquakes [Lay et al., 1982]. The fault consists of "strong" and "weak" regions. The stron- 1Nowat Department of Geosciences, University of Arizona, Tucson. Copyright 1991 by the American Geophysical Union. Paper number 90JB02092. 0148-0227/91/90JB-02092505.00 gest regions are often termed asperities and have a higher failure stress than the weaker regions. The essential feature of the asperity model is that the largest earthquake occurs when the dominant asperity breaks. The weaker regions slip at a lower level of applied stress and may slip seismically with the large earthquakes but with smaller displacementsor they may slip aseismically. A primary feature of the asperity model is that the coseismic displacement during an earth- quake is largest at the asperity. For large earthquakes, teleseimic P waves can be used to determine the temporal and spatial heterogeneity in seismic moment release associ- ated with the earthquake rupture. Regions of relatively high moment release are interpreted as asperities. However, the physical properties of these seismically determined asperi- ties are not well understood. Efforts to identify bathymetric features on the downgoing plate that correlate with the asperitieshave generally been unsuccessful. In addition, we do not understand the interaction of adjacent asperities along a plate boundary segment or the variations between succes- sive earthquake cycles. The February 4, 1965, Rat Islands earthquake (Mw = 8.7) yields informationabout not only the details of the rupture of a truly great earthquake but also some insight into asperity interaction and correlation with the lateral segmentation along the plate boundary. 2205