DECODING SEISMIC SIGNALS WITH SEAFLOOR OPTICAL FIBER STRAINMETERS (SOFS) IN OFFSHORE CASCADIA SUBDUCTION ZONE

Episodic tremor and slip (ETS) events, commonly recorded by onshore GNSS stations in the Cascadia region, typically occur where the plate interface is approximately 30-40 km deep. Notable research, has suggested the extension of these phenomena offshore, potentially increasing the risk of triggering significant earthquakes. The challenge in detecting offshore ETS lies in the ineffectiveness of GNSS technology on the seafloor. Previous efforts using seafloor optical fiber strainmeters (SOFS) have provided continuous measurements over extended periods (e.g., Zumberge et al., 2018) with an uncertainty level of about 30nε during ETS activity. However, these studies did not report any detectable strain transients, indicating no major short- term stress or slip changes in the monitored segments of the subduction zone. Building on these initial findings, we have recently deployed advanced SOFS to collect yearlong strain data from offshore areas in the Cascadia Subduction Zone, marking a pioneering step in monitoring these regions. This innovative method aims to capture the elusive signals of offshore tremors, small-scale earthquakes, and slow slip events (SSEs). We hypothesized that the updated instruments also record low-frequency seismic earthquakes (LFE, VLFE). We have applied temperature and tidal corrections, harmonic analysis, and conventional filtering and smoothing techniques to enhance data quality. In addition to the continuous strain data, our instruments detected several seismic activities, including a significant M7.6 earthquake near Mexico’s Pacific Coast on September 19, 2022, attributed to shallow thrust faulting. Concurrently, an onshore ETS event was observed in borehole strainmeters from mid-September to October 2022. Besides combining the onshore data, we have incorporated advanced time series analysis techniques, such as matrix profiling, motif, and discord discovery into our analytical processes. These novel algorithms have significantly enhanced our ability to identify patterns and anomalies within complex and noisy strain data from seismic events.