TY - JOUR A1 - Steinberg, Andreas A1 - Vasyura-Bathke, Hannes A1 - Gaebler, Peter Jost A1 - Ohrnberger, Matthias A1 - Ceranna, Lars T1 - Estimation of seismic moment tensors using variational inference machine learning JF - Journal of geophysical research : Solid earth N2 - We present an approach for rapidly estimating full moment tensors of earthquakes and their parameter uncertainties based on short time windows of recorded seismic waveform data by considering deep learning of Bayesian Neural Networks (BNNs). The individual neural networks are trained on synthetic seismic waveform data and corresponding known earthquake moment-tensor parameters. A monitoring volume has been predefined to form a three-dimensional grid of locations and to train a BNN for each grid point. Variational inference on several of these networks allows us to consider several sources of error and how they affect the estimated full moment-tensor parameters and their uncertainties. In particular, we demonstrate how estimated parameter distributions are affected by uncertainties in the earthquake centroid location in space and time as well as in the assumed Earth structure model. We apply our approach as a proof of concept on seismic waveform recordings of aftershocks of the Ridgecrest 2019 earthquake with moment magnitudes ranging from Mw 2.7 to Mw 5.5. Overall, good agreement has been achieved between inferred parameter ensembles and independently estimated parameters using classical methods. Our developed approach is fast and robust, and therefore, suitable for down-stream analyses that need rapid estimates of the source mechanism for a large number of earthquakes. KW - seismology KW - machine learning KW - earthquake source KW - moment tensor KW - full KW - waveform Y1 - 2021 U6 - https://doi.org/10.1029/2021JB022685 SN - 2169-9313 SN - 2169-9356 VL - 126 IS - 10 PB - American Geophysical Union CY - Washington ER - TY - JOUR A1 - López-Comino, José Ángel A1 - Cesca, Simone A1 - Niemz, Peter A1 - Dahm, Torsten A1 - Zang, Arno T1 - Rupture directivity in 3D inferred from acoustic emissions events in a mine-scale hydraulic fracturing experiment JF - Frontiers in Earth Science N2 - Rupture directivity, implying a predominant earthquake rupture propagation direction, is typically inferred upon the identification of 2D azimuthal patterns of seismic observations for weak to large earthquakes using surface-monitoring networks. However, the recent increase of 3D monitoring networks deployed in the shallow subsurface and underground laboratories toward the monitoring of microseismicity allows to extend the directivity analysis to 3D modeling, beyond the usual range of magnitudes. The high-quality full waveforms recorded for the largest, decimeter-scale acoustic emission (AE) events during a meter-scale hydraulic fracturing experiment in granites at similar to 410 m depth allow us to resolve the apparent durations observed at each AE sensor to analyze 3D-directivity effects. Unilateral and (asymmetric) bilateral ruptures are then characterized by the introduction of a parameter kappa, representing the angle between the directivity vector and the station vector. While the cloud of AE activity indicates the planes of the hydrofractures, the resolved directivity vectors show off-plane orientations, indicating that rupture planes of microfractures on a scale of centimeters have different geometries. Our results reveal a general alignment of the rupture directivity with the orientation of the minimum horizontal stress, implying that not only the slip direction but also the fracture growth produced by the fluid injections is controlled by the local stress conditions. KW - directivity KW - earthquake source KW - induced seismicity KW - hydraulic KW - fracturing KW - acoustic emissions Y1 - 2021 U6 - https://doi.org/10.3389/feart.2021.670757 SN - 2296-6463 VL - 9 PB - Frontiers Media CY - Lausanne ER -