Search for: All records

Abstract With the rise of data volume and computing power, seismological research requires more advanced skills in data processing, numerical methods, and parallel computing. We present the experience of conducting training workshops in various forms of delivery to support the adoption of large-scale high-performance computing (HPC) and cloud computing, advancing seismological research. The seismological foci were on earthquake source parameter estimation in catalogs, forward and adjoint wavefield simulations in 2D and 3D at local, regional, and global scales, earthquake dynamics, ambient noise seismology, and machine learning. This contribution describes the series of workshops delivered as part of research projects, the learning outcomes for participants, and lessons learned by the instructors. Our curriculum was grounded on open and reproducible science, large-scale scientific computing and data mining, and computing infrastructure (access and usage) for HPC and the cloud. We also describe the types of teaching materials that have proven beneficial to the instruction and the sustainability of the program. We propose guidelines to deliver future workshops on these topics. 

The January 1st, 2024, moment magnitude (Mw) 7.5 Noto Peninsula earthquake ruptured in complex ways, challenging timely analysis of the tsunami generation. We present rapid and accurate tsunami models informed by a 6-subevent centroid moment tensor (CMT) model that we obtain by inverting teleseismic and strong motion data and validation against geodetic observations. We identify two distinct bilateral rupture episodes, including six subevents and a re-nucleation episode at its hypocenter 20 seconds after its initiation, likely aided by fault weakening. We construct a complex uplift model that aligns with known fault system geometries and is critical in modeling the observed tsunami. Our tsunami simulation can explain wave amplitude, timing, and polarity of the leading wave, which are crucial for tsunami early warning. Analyzing a 2000 multi-CMT solution ensemble and comparing to alternative rapid source models, we highlight the importance of incorporating complex source effects for realistic tsunami simulations. 

Large earthquakes rupture faults over hundreds of kilometers withinminutes. Finite-fault models elucidate these processes and provideobservational constraints for understanding earthquake physics. However,finite-fault inversions are subject to non-uniqueness and substantialuncertainties. The diverse range of published models for thewell-recorded 2011 M_w 9.0 Tohoku-Oki earthquake aptly illustrates thisissue, and details of its rupture process remain under debate. Here, wecomprehensively compare 32 finite-fault models of the Tohoku-Okiearthquake and analyze the sensitivity of three commonly-usedobservational data types (geodetic, seismic, and tsunami) to the slipfeatures identified. We first project all models to a realisticmegathrust geometry and a 1-km subfault size. At this scale, we observepoor correlation among the models, irrespective of the data type.However, model agreement improves significantly when subfault sizes areincreased, implying that their differences primarily stem fromsmall-scale features. We then forward-compute geodetic and teleseismicsynthetics and compare them with observations. We find that seismicobservations are sensitive to rupture propagation, such as thepeak-slip-rise time. However, neither teleseismic nor geodeticobservations are sensitive to spatial slip features smaller than 64 km.In distinction, the synthesized seafloor deformation of all modelsexhibits poor correlation, indicating sensitivity to small-scale slipfeatures. Our findings suggest that fine-scale slip features cannot beunambiguously resolved by remote or sparse observations, such as thethree data types tested in this study. However, better resolution maybecome achievable from uniformly gridded dense offshore instrumentation.