Description: Some of the largest earthquakes worldwide, including the 1964 9.2 Mw megathrust earthquake, occurred in Alaskan subduction zones. To better understand rupture processes and their mechanisms, we relate seafloor morphology from multibeam and regional bathymetric compilations with sub-seafloor images and seismic P-wave velocity structures. We re-processed legacy multichannel seismic (MCS) data including shot- and intra-shotgather interpolation, multiple removal and Kirchhoff depth migration. These images even reveal the shallow structure of the subducting oceanic crust. Traveltime tomography of a coincident vintage (1994) wide angle dataset reveals the P-wave velocity distribution as well as the deep structure of the subducting plate to the ocean crust Moho. The subducting oceanic crust morphology is rough and partly hidden by a thick sediment cover that reaches ~3 km depth at the trench axis. Bathymetry shows two major contrasting upper plate morphologies: the shallow dipping lower slope consists of trench-parallel ridges that form the accreted prism whereas the steep rough middle and upper slopes are composed of competent older rock. Thrust faults are distributed across the entire slope, some of which connect with the subducted plate interface. A subtle change in seafloor gradient from the lower to the middle slope coincides with a thrust fault zone marking the boundary between the margin framework and the frontal prism. It corresponds to the most prominent lateral increase in seismic P-wave velocities, ~25 km landward of the trench axis. Major thrusts in several MCS-lines are correlated with bathymetric data, showing their 〉 100 km lateral extent, which might also be tsunamigenic paths of earthquake rupture from the seismogenic zone to the seafloor.

Description: Some of the largest earthquakes worldwide, including the 1964 9.2 Mw megathrust earthquake, occurred in Alaskan subduction zones. To better understand rupture propagation, with a focus on barriers that limit rupture, i.e. the southwest end of the Albatross segment at the boundary with the 1938 Semidi Mw 8.3 earthquake segment, we relate multibeam seafloor morphology with sub-seafloor images and seismic P-wave velocity structure. We re-processed legacy multichannel seismic (MCS) data including shot- and intra-shotgather interpolation, multiple removal and Kirchhoff depth migration and/or MCS traveltime tomography. These images even reveal the shallow structure of the subducting oceanic crust. Traveltime tomography of a coincident vintage wide-angle dataset reveals the P-wave velocity and the deep structure of the subducting plate to depths of the ocean crust Moho. The subducting oceanic crust morphology is rough and partly hidden by thick sediment cover that reaches ~3 km in the Albatross and ~1.5-2 km thickness in the Semidi segment at the trench axis. In both segments, bathymetry shows two major contrasting upper plate morphologies: trench-parallel ridges form the accreted prism of the shallow dipping lower slope whereas the steep rough middle and upper slopes are composed of competent older rock. Thrust faults are distributed across the entire slope, some of which connect with the subducted plate interface. A subtle change in seafloor gradient from the lower to the middle slope coincides with a splay fault zone (SFZ) marking the boundary between the margin framework and the frontal prism. This SFZ corresponds to the most prominent lateral increase in seismic P-wave velocities, ~25 km landward of the trench axis. Major differences in the Albatross/Semidi segments are 1.) origin of subducting sediment (Surveyor vs. Kodiak fan) 2.) geometry of the subducting plate (gentle vs. steep dip) beneath the lower and middle slopes.

Description: The Costa Rica Seismogenesis Project (CRISP) is designed to explore the processes involved in the nucleation of large interplate earthquakes in erosional subduction zones. On 16 June 2002 a magnitude Mw=6.4 earthquake and its aftershocks may have nucleated at the subduction thrust to be penetrated and sampled by CRISP, ~40 km west of Osa Peninsula. Global event locations present uncertainties too large to prove that the event actually occurred at a location and depth reachable by riser drilling. We have compiled a database including foreshocks, the main shock, and ~400 aftershocks, with phase arrival times from all the seismological networks that recorded the 2002 Osa sequence locally. This includes a temporal network of ocean-bottom hydrophones (OBH) that happened to be installed close to the area at the time of the earthquake. The coverage increase provided by the OBH network allow us to better constrain the event relocations, and to further analyze the seismicity in the vicinity of Osa for the six months during which they were deployed. We derived a minimum 1D model and used probabilistic earthquake relocation. Moreover, we undertook teleseismic waveform inversion to provide additional constraints for the centroid depth of the 2002 Osa earthquake. The latter, together with the maximum likelihood hypocenter, places the main shock origin at 5 to 10 km depth, ~30 km landward from the trench. Along the Costa Rican seismogenic zone, the 2002 Osa sequence is the most recent. It nucleated in the SE region of the forearc where this erosional margin is underthrust by a seamount covered ocean plate. A Mw=6.9 earthquake sequence occurred in 1999, co-located with a subducted ridge and associated seamounts. The Osa mainshock and first hours of aftershocks began in the CRISP area, ~30 km seaward of the 1999 sequence. In the following two weeks, subsequent aftershocks migrated into the 1999 aftershock area and also clustered in an area updip from it. The Osa updip seismicity apparently occurred where interplate temperatures are ~100°C or less. In this study, we present the relocation of the 2002 Osa earthquake sequence and background seismicity using different techniques and a moment tensor inversion for the mainshock, and discuss the corresponding uncertainties, in an effort to provide further evidence that the planned Phase B of CRISP will be successful in drilling the seismogenic coupling zone.

Description: Some of the largest earthquakes worldwide, including the 1964 9.2 Mw megathrust earthquake, occurred in Alaskan subduction zones. To better understand rupture processes and their mechanisms, we relate seafloor morphology from multibeam and regional bathymetric compilations with sub-seafloor images and seismic P-wave velocity structures. We re-processed legacy multichannel seismic (MCS) data including shot- and intra-shotgather interpolation, multiple removal and Kirchhoff depth migration. These images even reveal the shallow structure of the subducting oceanic crust. Traveltime tomography of a coincident vintage (1994) wide angle dataset reveals the P-wave velocity distribution as well as the deep structure of the subducting plate to the ocean crust Moho. The subducting oceanic crust morphology is rough and partly hidden by a thick sediment cover that reaches ~3 km depth at the trench axis. Bathymetry shows two major contrasting upper plate morphologies: the shallow dipping lower slope consists of trench-parallel ridges that form the accreted prism whereas the steep rough middle and upper slopes are composed of competent older rock. Thrust faults are distributed across the entire slope, some of which connect with the subducted plate interface. A subtle change in seafloor gradient from the lower to the middle slope coincides with a splay fault zone marking the boundary between the margin framework and the frontal prism. It corresponds to the most prominent lateral increase in seismic P-wave velocities, ~25 km landward of the trench axis. Major thrusts in several MCS-lines are correlated with bathymetric data, showing their 〉 100 km lateral extent, which might also be tsunamigenic paths of earthquake rupture from the seismogenic zone to the seafloor. The splay fault zone has been recognized as a potential tsunamigenic structure in the 1938 and 1946 earthquake rupture areas.

Description: The Costa Rica Seismogenesis Project (CRISP) is designed to explore the processes involved in the nucleation of large interplate earthquakes in erosional subduction zones. On 16 June 2002 a magnitude Mw=6.4 earthquake and its aftershocks may have nucleated at the subduction thrust to be penetrated and sampled by CRISP, ~40 km west of Osa Peninsula. Global event locations present uncertainties too large to prove that the event actually occurred at a location and depth reachable by riser drilling. We have compiled a database including foreshocks, the main shock, and ~400 aftershocks, with phase arrival times from all the seismological networks that recorded the 2002 Osa sequence locally. This includes a temporal network of ocean-bottom hydrophones (OBH) that happened to be installed close to the area at the time of the earthquake. The coverage increase provided by the OBH network allow us to better constrain the event relocations, and to further analyze the seismicity in the vicinity of Osa for the six months during which they were deployed. Moreover, we undertook teleseismic waveform inversion to provide additional constraints for the centroid depth of the 2002 Osa earthquake, allowing further study of the focal mechanism. Along the Costa Rican seismogenic zone, the 2002 Osa sequence is the most recent. It nucleated in the SE region of the forearc where this erosional margin is underthrust by a seamount covered ocean plate. A Mw=6.9 earthquake sequence occurred in 1999, co-located with a subducted ridge and associated seamounts. The Osa mainshock and first hours of aftershocks began in the CRISP area, ~30 km seaward of the 1999 sequence. In the following two weeks, subsequent aftershocks migrated into the 1999 aftershock area and also clustered in an area updip from it. The Osa updip seismicity apparently occurred where interplate temperatures are ~100°C or less. In this study, we present the relocation of the 2002 Osa earthquake sequence and background seismicity using different techniques and a moment tensor inversion for the mainshock, and discuss the corresponding uncertainties, in an effort to provide further evidence that the planned Phase B of CRISP will be successful in drilling the seismogenic coupling zone.