A new method to determine the depth of earthquakes on oceanic transform faults using teleseismic arrays: application to the Chain transform, equatorial Atlantic

The depth distribution of earthquakes on oceanic transform faults is critical to understand how slip is accommodated at depth with increasing pressure and temperature and the mechanical properties of the oceanic lithosphere. Yet, most oceanic earthquakes have fixed depths in global catalogs, largely due to the lack of local seismic observations. Dense OBS deployments are expensive and relatively short term. They are providing detailed analysis of faster, shorter moving transforms, but are not easily extended to the greater spatial and temporal scales of the slower moving transforms of the Atlantic Ocean. In this study, we take advantage of the abundant distribution of seismic stations in the teleseismic range and develop a new method using large aperture arrays to enhance the weak signals associated with earthquakes on oceanic transform faults. These include the first arrivals and surface reflections (depth phases) that are sensitive to the depth of earthquakes. We refine the depth of the earthquakes by waveform simulation that takes into consideration of the seafloor topography. The method relies on the resemblance of earthquake signals over a large epicentral distance range, leading to coherent stacking when phase shifts between different stations are corrected. We show that the new method outperforms conventional cross-correlation-based techniques for recordings with low signal-to-noise ratio. We apply the approach to Mw>4.5 earthquakes along the Chain transform fault in the Equatorial Atlantic, the site of a recent OBS deployment. We find good agreement between the depths we obtained and those from moment-tensor inversion of the OBS recordings. Moreover, we find that the maximum centroid depth could reach about 19 km below the seafloor, which is much deeper than the 600 °C isotherm estimated using a half-space cooling model and indicates complex earthquake mechanisms at play. The method can be applied to refine the depth of moderate-sized oceanic earthquakes along the global slower transform faults. In this way, we can learn more about these major faults and also, by comparison and contrast, improve our understanding of the behavior of the shorter, faster transforms, and their more dangerous continental cousins.