What Controls Coupling and Seismogenic Potential Along Faults?

Why some tectonic plate boundary faults are locked and accumulate a sufficiently large slip deficit to produce large earthquakes remains unknown. 70% of earthquakes occur beneath the oceans, including the largest earthquakes in subduction zones (SZs), such as those in 2004 in Indonesia and 2011 in Japan. To forecast future earthquakes, it is, therefore, crucial to map offshore fault coupling and identify the frictional regimes that drive coupling. However, slow movements along offshore faults are largely hidden from onshore GPS measurements, so without seafloor data, efforts to understand coupling along these hazardous faults will remain futile. The 2011 Japan earthquake completely shifted our thinking about how faults work, leading to many subsequent studies focussing on SZs around the ‘Pacific Ring of Fire’. However, questions remain about more weakly locked faults and whether slowly deforming plate boundaries, such as those in the Atlantic region, can generate strong coupling. This uncertainty has led to a poorly estimated potential for large earthquakes in this area, with debates ongoing over how strongly coupled the faults are. Understanding the physical properties that affect coupling, such as the inherited structure of subducting plates (e.g., fluid content), is also crucial. Ocean transform faults (OTFs) provide a complementary tectonic laboratory to SZs for probing what controls coupling. OTFs are geometrically simpler and have wider fault damage zones than their continental counterparts, offering a unique environment to investigate how fluid flow along fault strands and fracture networks drive slip and coupling. I will exploit the fact that the same frictional mechanisms ultimately control large and small earthquakes. I will characterise the many weak events recorded by targeted ocean-bottom seismometer (OBS) experiments to determine coupling in multiple tectonic environments in the slowly deforming Atlantic. One of my main study areas will be the Lesser Antilles SZ to investigate whether it is capable of an M>8 earthquake, which would cause a transatlantic tsunami and local devastation, given the low-lying islands and lack of a tsunami warning system. I will also study Atlantic OTFs because they can host large (M>7) earthquakes and multi-mode slip (e.g., slow and supershear) despite assumed weak coupling. Manual analysis cannot efficiently deal with such large datasets. However, my involvement in multiple OBS experiments worldwide will allow me to develop the first machine-learning neural network model for classifying diverse signals in OBS data. Using a novel relocation technique, I will map fault zones with high accuracy and precision that implicitly accounts for Earth structure unknowns. Waveforms from small events also contain information about the material properties where they slip, so I will apply novel imaging techniques to OBS data (e.g., in-situ Vp/Vs; S-wave splitting; seismic attenuation; fault zone guided waves). I will also deploy dense nodal seismic networks to probe the material properties that lead to variable coupling. The structure of the incoming plate at the Antilles trench is well established, so I will directly relate these structures and associated material properties to plate interface coupling. My long-term vision is to continue to push our boundaries of fault zone seismic imaging. I will explore new seismic recording methodologies, such as ocean fibre cables. Finally, I will synthesise my results from studies of continental faults and Pacific-type SZs to explore unified relationships between background seismicity and coupling that will ultimately improve long-term earthquake forecasts and seismic hazard models.