A new 3-D <i>P</i> -wave velocity model for Central America using the TeletomoDD method

SUMMARY Central America’s tectonic complexity arises from the interaction of multiple plates and diverse plate boundaries, resulting in high seismic activity and intricate subduction processes. 3-D seismic velocity models can provide critical constraints on subduction processes and associated earthquake hazard models. Although regional tomographic studies have offered insights into seismic activity and lithospheric processes in Central America, there have been few studies that image the entire region in a consistent manner, likely due to the geological complexity and numerical challenges. In this study, we develop a new high-resolution 3-D compressional (P)-wave velocity model to investigate the subduction dynamics of the region. We apply the teletomoDD method, which uses both local and global body-wave arrivals to resolve velocity structures. We use the International Seismological Center’s EHB catalogue to extract data from 6026 regional earthquakes from 1965 to 2019, recorded by seismic stations both inside and outside of our study area. Our model is further constrained by incorporating about 30 000 global events recorded by the seismic stations within our study area. We perform both checkerboard and restoration tests to assess the resolution of the model and find that the main features are resolved robustly regardless of the initial models. The model shows a coherent high-velocity anomalies along the Middle America Trench at 50 km depth, suggesting cold, dense subducting slabs. It also captures notable variations in slab geometry, including a slab window in the southern Cocos region starting at $\sim$75 km depth. Low-velocity anomalies beneath major volcanic systems such as the Central American Volcanic Arc and the Trans-Mexican Volcanic Belt point to slab dehydration, fluid migration and partial melting processes, whereas the discontinuous distribution of volcanism in Mexico and Central America appears to be influenced by the subduction of the Cocos Plate. Additionally, we identify high-velocity anomalies near the Siqueiros and Clipperton Transform Faults on the East Pacific Rise, possibly caused by mafic magmatic cumulates. The high-velocity anomaly near Swan Islands Transform Fault may reflect locally increased density inferred from previous gravity studies. Our new velocity model offers a consistent seismic structural foundation for further investigations into seismogenic processes, slab geodynamics, petrology and rheology in Central America.