Hydroacoustic Monitoring of Earthquake Sequences on the Blanco Oceanic Transform Fault

Understanding how earthquakes initiate lies at the heart of earthquake physics and analyzing foreshock sequences is one way to probe the initiation process of large earthquakes. A few oceanic transform fault (OTF) earthquakes have been observed to have more foreshocks compared to continental earthquakes. It has also been proposed that OTFs accommodate plate motion primarily by slow creep instead of rapid seismic slip, though with significant along-fault variability. These characteristics make OTFs unique laboratories for probing the physical processes underlying foreshocks and their relations with slow slip events. However, in the past, detailed studies at OTFs have been limited due to their distance from land-based seismic stations. Since 2015, small arrays of ocean-bottom seismometers and hydrophones have been permanently deployed on cabled seafloor observatories in the northeast Pacific Ocean, allowing for monitoring of seismicity on the Blanco Transform Fault (BTF) using the earthquake’s radiated hydroacoustic energy (T-phase). T-phases propagate through the SOFAR channel in the world’s oceans with little attenuation, allowing for the detection of low-magnitude earthquakes at large distances. In this study, we apply a suite of techniques to detect, associate, and locate foreshocks and aftershocks of large mainshocks occurring along the BTF since 2015. We define a mainshock as the largest event occurring within two weeks and a radius of 50 km. 19 Mw  5.0 mainshocks are selected from the GCMT catalogue. However, we are only able to analyze 12 mainshocks due to data availability issues. We use the STA/LTA algorithm to detect T-phase arrivals one week before and after each mainshock through the continuous waveforms recorded on both OBSs and hydrophones. For each detection, we then use the relative station arrival times compared to the mainshock to make sure we only retain events close to the mainshock, i.e., its foreshocks and aftershocks. We then employ the GLOBAL mode Non-Linear Location (NLLoc) program for event localization using a 3D ocean sound velocity model. Compared to the IRIS catalogue which only has a minimum detection level of magnitude 3, lower-magnitude T-phase events are successfully detected by our method. Using our T-phase catalogue, we quantify how the BTF sequences compare with earthquake sequences observed at continental transform faults and test the various proposed models to explain foreshock spatiotemporal behavior and the partitioning of seismic and slow slips along the BTF.