How seismotectonics of convergent margins reconciles with the force balance of subduction is contentious. The comparison of seismotectonics and available slab pull forces along the Sunda convergent margin shows an enigmatic inverse relationship: upper plate thickening and seismicity magnitude are highest along Sumatra and Andaman, where the slab is shorter than ∼300 km; conversely, these are negligible along the Java segment, where the slab reaches deeper, ∼660 km. Using numerical models, we test the role of such slab pull gradients in the force balance of subduction in three‐dimensions, where the slab depth, and therefore its net pull, varies along the trench. We show that in the presence of a “slab step,” the deeper slab drives the convergence of the rigid plate, causing upper plate compression and trench advance in the neighboring trench segments, where a short slab may have no pull to subduct the incoming plate. While neglecting convergence obliquity, the simplified models show relevant along‐trench variations of coupling, trench rotations, and minor strike‐slip shearing due to the slab step, providing a diagnostic strain pattern, with compression/extension atop the short/long slab and minor strike‐slip, increasing in magnitude with depth difference. The modeled tectonic patterns are compared to Sunda margin deformation across scales, from the Cenozoic tectonics to the seismic strain rates, showing remarkable consistency with deformation gradients from Sumatra to Java, potentially illustrating the contribution of the slab step to the seismotectonics of the region.
<p>Understanding the controls on large magnitude seismicity occurrence remains an open challenge, yet a pressing one, for the exceptional hazard associated with earthquakes. Different parameters are proposed to exert control on the generation and propagation of megathrust earthquakes and untangling their complex interactions across scales remains challenging. Here, we use explainable artificial intelligence to unravel the interactions between different parameters and elucidate the underlying mechanisms. We use three types of datasets from a number of convergent margins: <em>a</em>) a catalogue of earthquake hypocentre and rupture, <em>b</em>) geophysical observations of subduction zones properties (e.g., gravity, bathymetric roughness, sediment thickness), and<em> c</em>) the distribution of stress within the slab due to slab pull calculated from flexure models. These constitute the three types of nodes in the input layer of a Fully Connected Network (FCN) trained to classify earthquake magnitude embedding the state of the system (<em>b</em>), the driving mechanism (<em>c</em>) and the resulting seismicity (<em>a</em>). We then analyse the trained network using Layer-wise Relevance Propagation (LRP) to determine the relative weights of the input nodes, providing relevant constraints on the mechanisms that dominate the seismicity in a region, their scale and likelihood.</p>
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.