Earthquake focal mechanism data provide information about the stress state at the origin of these earthquakes. The inversion methods that are commonly used to infer the stress tensor from focal mechanisms have varying complexity but always rely on a number of assumptions. We present an iterative method built upon a classic linear stress tensor inversion that allows to relax the assumption on shear stress magnitudes while preserving the computational simplicity of the linear problem. Every iteration of our method computes the least-squares solution of the problem, which makes the method fast enough to estimate the inverted parameter errors with non-parametric resampling methods such as bootstrapping. Following previous studies, this method removes the fault plane ambiguity in focal mechanism data by selecting the nodal plane that best satisfies the Mohr-Coulomb failure criterion. We first test the performance and the robustness to noise of the proposed method on synthetic data sets, and then apply it to data from Southern California and from the Geysers geothermal field. We focus the study on investigating the consequences of relaxing the assumption of constant shear stress magnitudes. Our variable shear method successfully generalizes its constant shear counterpart: it is able to perform similarly when the constant shear assumption is a good approximation and provides more accurate results when it is not. We provide the Python package ILSI to implement the proposed method.

We propose a strategy to identify seismic signal classes in continuous single-station seismograms in an unsupervised fashion. Our strategy relies on extracting meaningful waveform features based on a deep scattering network combined with an independent component analysis. We then identify signal classes from these relevant features with agglomerative clustering, which allows us to explore the data in a hierarchical way. To test our strategy, we investigate a two-day long seismogram collected in the vicinity of the North Anatolian fault in Turkey. We interpret the automatically inferred clusters by analyzing their occurrence rate, spectral characteristics, cluster size, and waveform and envelope characteristics. At a low level in the cluster hierarchy, we obtain three clusters related to anthropogenic and ambient seismic noise and one cluster related to earthquake activity. At a high level in the cluster hierarchy, we identify a seismic crisis with more than 200 repeating events and high-frequent signals with correlating envelopes and an anthropogenic origin. The application shows that the cluster hierarchy can be used to draw the focus on a certain class of signals and extract subclusters for further analysis. This is interesting, when certain types of signals such as earthquakes are under-represented in the data. The proposed method can be also used to discover new types of signals since it is entirely data-driven.

Foreshocks are among the most powerful tools to study the processes that occur before main earthquakes. However, their detection is still sparse, especially for relatively small earthquakes. We present here a detailed foreshock analysis for the 2019 Balsorano (Italy) earthquake (Mw4.4). To improve the detection before and after the mainshock, we use receivers at distances of <75 km from the targeted seismicity, through template matching. To improve the understanding of the mechanism(s) behind the earthquake initiation, we detail the evolution of the sequence associated to this earthquake, using waveform clustering and hypocenter relocation. Differences between foreshocks and aftershocks are revealed by this analysis. Moreover, distinct patterns associated to the different seismic activities are revealed. The observed behavior highlights a complex initiation process, which apparently starts on an adjacent antithetic fault. Finally, the aftershock activity comprises different clusters with distinct spatio-temporal patterns, which suggests that each cluster has distinct triggering mechanism(s).