The Guerrero seismic gap (GG) along Mexico's Pacific coast poses a significant seismic and tsunami hazard. A potential large-scale rupture within the GG could generate a >M_w 8 earthquake, causing devastating effects on Mexico City and coastal communities. The northwestern segment (NW-GG) hasn't experienced a M_w > 7 earthquake since 1911, only smaller events and slow slip events (SSEs). Geodetic data suggests significantly lower coupling at the plate interface in the NW-GG than in adjacent segments. Existing onshore observations, however, limit our understanding of the shallow plate interface. This study aims to investigate shallow slow earthquakes, such as tectonic tremors, in the offshore region of the NW-GG using a newly deployed ocean bottom seismometer (OBS) array. The presence and characteristics of shallow slow earthquakes in this region could offer valuable insights into the mechanics of the plate interface and the potential for future large earthquakes. Understanding the interplay between different types of slow earthquakes (tremors and SSEs), regular earthquakes, and the geological structure is crucial for accurate hazard assessment. The study employs a combination of seismological observations, analysis of residual gravity and bathymetry, and the development of a plate interface mechanical model to achieve this.

Previous research has documented shallow slow earthquakes at several subduction zones, often near the trench. These events, including episodic tremor and slip (ETS), are not fully understood but are recognized as potentially significant in the context of megathrust earthquakes. Studies in the Japan Trench have shown that ETS can precede large earthquakes, highlighting the interplay between slow and fast slip. The GG itself has been studied extensively, but primarily onshore. Past studies have identified slow slip events within the Guerrero subduction zone, which have durations longer than typical earthquakes and magnitudes up to 7.5 Mw. These events have recurrence periods around 4 years, and are accompanied by tectonic tremors. It was found that these deep tremors migrate rapidly and are closely linked to SSEs and overpressurized fluids. Geodetic data show that the NW-GG has lower coupling compared to neighboring segments. To this point, shallow slow earthquakes in the offshore GG have not been observed because of lack of offshore instrumentation.

This study utilizes data from an array of seven OBSs deployed in the NW-GG for one year starting November 2017. These OBSs were equipped with three-component 1 Hz short-period sensors and located at water depths between 980 and 2350 meters. The locations of the OBSs were estimated with high precision (mean uncertainty of 2 meters). Data processing involved correcting for time shifts in seismic recordings. Shallow tremors (STs) were detected using a modified envelope correlation method, initially identifying potential locations and then refining them using a maximum-likelihood approach. Regular earthquakes were detected from continuous OBS data using an STA/LTA method, with subsequent manual picking and location refinement using a maximum-likelihood approach and a 1D velocity model. Repeating earthquakes (repeaters) were identified using a combination of onshore data from 2001-2019 and offshore data from 2017-2018, utilizing correlation coefficient (CC) and spectral coherency (COH) analysis. The study also incorporated analysis of residual gravity and bathymetry data to infer subducting relief and its influence on the plate interface. A plate interface mechanical model was developed to integrate all the observational data, differentiating regions based on their slip behavior, inferred frictional conditions and the spatial distribution of the observed phenomena. This model is used to better understand the coupling conditions and the potential for future large earthquake nucleation in the NW-GG and neighboring segments.

The study detected over 100 STs in the NW-GG, most located between 10 and 16 km depth. No clear migration of STs was observed, suggesting mechanically isolated locked patches within a weakly coupled plate interface. STs were spatially clustered into four groups (S1-S4), with variations in magnitude and recurrence periods. These episodic tremor bursts show recurrence periods of one to three months, potentially indicating short-term SSEs. A “silent zone” was identified with a lack of seismicity, surrounded by STs, repeaters, and areas of past earthquake rupture. The distribution of STs, repeaters, and earthquakes correlate with positive residual gravity and bathymetry anomalies, potentially related to subducting relief that causes increased pore pressure and reduced coupling. The analysis divided the region into different domains based on seismic and geodetic characteristics and frictional properties. Region A (shallowest) shows episodic STs, repeaters, and near-trench tsunami earthquakes, indicative of velocity-weakening asperities within a velocity-strengthening matrix. Region A' (silent zone) is a velocity-strengthening domain with low probability of large earthquake nucleation. Region B (below the coast) contains STs, repeaters, and numerous earthquakes, suggesting frictional heterogeneity. Region C (deepest) shows long-term SSEs and deep tremors. A key difference was observed between the NW-GG and the adjacent Petatlán segment: the NW-GG is characterized by weakly coupled conditions, making the initiation of an M>8 earthquake less probable. The silent zone is found to act as a barrier, preventing large earthquakes from the Petatlán segment from propagating into the NW-GG.

The findings challenge the prevailing assumption of a high likelihood of a large earthquake in the NW-GG. The observed weak coupling, both offshore and onshore, explains the long recurrence interval of large earthquakes. The heterogeneous frictional conditions, influenced by subducting relief, contribute to the diversity of slip behaviors. While the risk of a large-scale rupture is reduced, it isn’t eliminated. The stress accumulation around the silent zone could trigger an earthquake initiated in the adjacent Petatlán segment to propagate across the gap. The coexistence of seismic and aseismic events in the same region is similar to the Japan Trench, where subducted seamounts, seismic and aseismic events coexist. These results necessitate continued efforts in disaster prevention.

This study provides the first offshore observations of shallow slow earthquakes in the Guerrero seismic gap. The findings suggest a more complex picture of the plate interface than previously assumed. Weak coupling and the presence of a velocity-strengthening silent zone decrease the likelihood of a large earthquake nucleating within the NW-GG, although the possibility of a large earthquake still remains. Future research should focus on validating these findings through further offshore and onshore observations, combined with physics-based source modeling.

The study's duration (one year) limits the assessment of long-term slip behavior. The depth resolution for tremor locations is relatively low, mostly because the used 1D velocity model does not include a shallow sedimentary layer. The interpretation of frictional conditions relies on inferences from observed seismic and geodetic data. The model is a simplification of a complex system, and further investigation may reveal additional subtleties.