[1] GPS measurements of interseismic horizontal surface velocities reveal the degree of kinematic coupling of the plate boundary thrust along the Kamchatka subduction zone from about 51°to 57°N latitude. Inversions for the distribution of aseismic slip rate along the $15°NW dipping underthrust suggest a nonslipping plate interface in southern Kamchatka above $50 km depth, along the segment that ruptured in the M w = 9, 1952 earthquake. North of $53°N, the subduction interface experiences significant aseismic slip, consistent with the lower seismic moment release in M 8.5 earthquakes along this portion of the subduction zone. The GPS velocities are consistent with a boundary element forward model in which historic earthquake rupture zones are represented as locked asperities, surrounded by a zero shear stress subduction interface loaded by plate convergence. Models in which the complete rupture zones of historic earthquakes are considered locked greatly overpredict the degree of kinematic coupling. Reducing the area of the locked model asperities to the central 25% area of historic rupture zones fits the data well, suggesting that large earthquakes involve small fully locked core asperities surrounded by conditionally stable portions of the plate interface. Areas of low aseismic slip rate appear to be roughly correlated with areas of low isostatic gravity anomalies over offshore forearc basins, while less coupled portions of the Kamchatka subduction zone coincide with high-gravity anomalies offshore of two peninsulas, possibly related to the subduction of the Emperor-Meji seamount chain and the Kruzenstern fracture zone.Citation: Bürgmann, R., M. G. Kogan, G. M. Steblov, G. Hilley, V. E. Levin, and E. Apel (2005), Interseismic coupling and asperity distribution along the Kamchatka subduction zone,

Abstract.Model (Fig. 1A). Historically, the Kamchatka subduc-

SUMMAR YSince 1996, a network of nine permanent GPS receivers has been recording, in continuous mode, the deformation on Kamchatka peninsula. The velocity and direction of the relative displacement of observation points are estimated from the entire data set for 1996-2000. The deformation related to the large Kronotskoe earthquake (M w =7.8, 1997 December 5) was identified at distances up to a few hundred kilometres. Half a month before this major event, large-amplitude pre-seismic deformation appeared at stations closest to its epicentral area. The deformation corresponds approximately to a double-couple slow event with M w =7.7 located in the foreshock area, with an orientation that differs significantly from that of the main shock. Clear coseismic displacements were also observed. They match well those predicted by the published Harvard CMT solution. Large-scale post-seismic deformation was also recorded, again with a duration of about half a month, and a cumulative moment comparable to that of the 'main' shock. In addition to the transient effects of a large earthquake, data show a secular trend that reflects both the continuous plate motion and the elastic response to interplate coupling. Preliminary estimates of relative plate velocities for the North America, Okhotsk and Beringia plates are given based on the data from stations distant from the most active plate boundaries. Other stations seem to show mainly the elastic response of the Okhotsk and Beringia plates to their coupling with the subducting Pacific plate. At one station at least, the velocity of continuous motion underwent a significant change at the time of the Kronotskoe earthquake, probably indicating a related change in interplate coupling or the effects of mantle rheology.

Abstract. To investigate the current crustal movements in and around the Sea of Okhotsk and Sea of Japan regions, we have established a continuous GPS network. By the end of 1997, the network had been expanded to include 12 new stations. Data for the period from July 1995 to November 1997 were analyzed together with data from International GPS Service for Geodynamics (IGS) global stations. To fix the estimated coordinates to the terrestrial reference frame, the Tsukuba IGS station was assumed to be moving westward relative to the stable Eurasian continent at •2cm/yr according to Heki's[1996] estimate. We find that: (1) stations in the western margin of the Sea of Japan have eastward velocity vectors, (2) the pole position of the Okhotsk plate is located near Okha, which reconfirms the Okhotsk micro plate, (3) a plate boundary of the Okhotsk and Amurian plates between southen Sakhalin and Hokkaido is suggested. IntroductionThe tectonic plate motions in the vicinity of the Sea of Okhotsk and Sea of Japan are very complex. Because of low seismicity and having no clear geographical boundary except for Kuril-Japan trench, it has been difficult to describe the plate tectonics in this region. Seno et al. [1996] Network and Data analysisWe established two GPS stations in Sakhalin on July 1995 at first, and further developed the station network in 1996 and 1997. By the end of 1997, 12 continuous observation stations in this region had been installed (Figure 1). We selected the station sites so that a stable power supply is available and a rigid installation of antenna can be made. The sampling rate of GPS observation in this network is every 30 seconds and the data are stored on the 100MB removable disks through a personal computer. All data collected at these stations are sent to the data center at the Hokkaido University within a half years delay at most. The data are converted to the standard receiver independent exchange (RINEX) format. We have analyzed the archived data using the Bemese GPS Software Ver- ResultsWe show two examples of time series of horizontal and vertical coordinates with respect to TSKB in Figure 2. These can 2533

No abstract