Searcharxiv⌕ Search

arXiv subjects

Sergio Vinciguerra

Publications and source records attributed to Sergio Vinciguerra.

6 recordsLinked to original sources
arXiv2025

Generalised Time-Series Analysis of Fault Mechanics Using Explainable AI

Understanding how faults nucleate and grow is a critical problem in earthquake science and hazard assessment. This study examines fault development in Alzo granite under triaxial pressures ranging from 5 to 40 MPa by applying a Time Delay Neural Network (TDNN) to multi-parameter acoustic emission (AE) data. The TDNN integrates waveform-derived attributes, including peak delay and scattering attenuation, with occurrence-based metrics such as time distributions, Gutenberg-Richter b-values, and spatial fractal dimensions, to characterize the transition from distributed microcracking to localised faulting. Optimised via genetic algorithms, the TDNN dynamically weights these parameters, enabling accurate characterisation of fault growth stages. Our results delineate three distinct phases of fault evolution: nucleation of random microcracks indicated by changes in elastic wave scattering, initiation of fault growth reflected in evolving AE spatial and magnitude distributions, and fault coalescence marked by exponential increases in peak delay and b-value shifts. The model predicts the timing and magnitude of stress drops across varying pressures and failure mechanisms, from axial splitting to shear localisation, providing deeper insights into fault mechanics through explainable AI models.

physics.geo-ph↗
arXiv2007

Analysis of Self-Organized Criticality in the Olami-Feder-Christensen model and in real earthquakes

We perform a new analysis on the dissipative Olami-Feder-Christensen model on a small world topology considering avalanche size differences. We show that when criticality appears the Probability Density Functions (PDFs) for the avalanche size differences at different times have fat tails with a q-Gaussian shape. This behaviour does not depend on the time interval adopted and is found also when considering energy differences between real earthquakes. Such a result can be analytically understood if the sizes (released energies) of the avalanches (earthquakes) have no correlations. Our findings support the hypothesis that a self-organized criticality mechanism with long-range interactions is at the origin of seismic events and indicate that it is not possible to predict the magnitude of the next earthquake knowing those of the previous ones.

cond-mat.stat-mech↗
arXiv2006

Multifractal analysis of Mt. St. Helens seismicity as a tool for identifying eruptive activity

We present a multifractal analysis of Mount St. Helens seismic activity during 1980-2002. The seismic time distribution is studied in relation to the eruptive activity, mainly marked by the 1980 major explosive eruptions and by the 1980-1986 dome building eruptions. The spectrum of the generalized fractal dimensions, i.e. Dq vs q, extracted from the data, allows us to identify two main earthquake time-distribution patterns. The first one exhibits a multifractal clustering correlated to the intense seismic swarms of the dome building activity. The second one is characterized by an almost constant value of Dq 1, as for a random uniform distribution. The time evolution of Dq (for q=0,2) calculated on a fixed number of events window and at different depths, shows that the brittle mechanical response of the shallow layers to rapid magma intrusions, during the eruptive periods, is revealed by sharp changes, acting at a short time scale (order of days), and by the lowest values of Dq (0.3). Conversely, for deeper earthquakes, characterized by intense seismic swarms, Dq do not show obvious changes during the whole analyzed period, suggesting that the earthquakes, related to the deep magma supply system, are characterized by a minor degree of clustering, which is independent of the eruptive activity.

physics.geo-ph↗
arXiv2003

Detection of invisible and crucial events: from seismic fluctuations to the war against terrorism

We prove the efficiency of a new method for the detection of crucial events that might have useful applications to the war against terrorism. This has to do with the search for rare but significant events, a theme of research that has been made of extreme importance by the tragedy of September 11. This method is applied here to defining the statistics of seismic main-shocks, as done in cond-mat/0212529. The emphasis here is on the conceptual issues behind the results obtained in cond-mat/0212529 than on geophysics. This discussion suggests that the method has a wider range of validity. We support this general discussion with a dynamic model originally proposed in cond-mat/0107597 for purposes different from geophysical applications. However, it is a case where the crucial events to detect are under our control, thereby making it possible for us to check the accuracy of the method of detection of invisible and crucial events that we propose here for a general purpose, including the war against terrorism. For this model an analytical treatment has been recently found [cond-mat/0209038], supporting the claims that we make in this paper for the accuracy of the method of detection. For the reader's convenience, the results on the seismic fluctuations are suitably reviewed, and discussed in the light of the more general perspective of this paper. We also review the model for seismic fluctuations, proposed in the earlier work of cond-mat/0212529. This model shares with the model of cond-mat/0107597 the property that the crucial events are imbedded in a sea of secondary events, but it allows us to reveal with accuracy the statistics of the crucial events for different mathematical reasons.

cond-mat.stat-mech↗
arXiv2003

Power-Law Time Distribution of Large Earthquakes

We study the statistical properties of time distribution of seimicity in California by means of a new method of analysis, the Diffusion Entropy. We find that the distribution of time intervals between a large earthquake (the main shock of a given seismic sequence) and the next one does not obey Poisson statistics, as assumed by the current models. We prove that this distribution is an inverse power law with an exponent $μ=2.06 \pm 0.01$. We propose the Long-Range model, reproducing the main properties of the diffusion entropy and describing the seismic triggering mechanisms induced by large earthquakes.

cond-mat.stat-mech↗