Abstract.The large amount of digital data recorded by permanent and temporary seismic networks makes automatic analysis of seismograms and automatic wave onset time picking schemes of great importance for timely and accurate event locations. We propose a fast and efficient P-and S-wave onset time, automatic detection method based on neural networks. The neural networks adopted here are particular neural trees, called IUANT2, characterized by a high generalization capability. Comparison between neural network automatic onset picking and standard, manual methods, shows that the technique presented here is generally robust and that it is capable to correctly identify phase-types while providing estimates of their accuracies. In addition, the automatic post processing method applied here can remove the ambiguity deriving from the incorrect association of events occurring closely in time. We have tested the methodology against standard STA/LTA phase picks and found that this neural approach performs better especially for low signal-to-noise ratios. We adopt the recall, precision and accuracy estimators to appraise objectively the results and compare them with those obtained with other methodologies.1 This paper has not been submitted elsewhere in identical or similar form, nor will it be during the first three months after its submission to Journal of Seismology. § Previously at OGS-CRS, Trieste, Italy. 2Tests of the proposed method are presented for 342 earthquakes recorded by 23 different stations (about 5000 traces). Our results show that the distribution of the differences between manual and automatic picking has a standard deviation of 0.064 s and 0.11 s for the P and the S waves, respectively. Our results show also that the number of false alarms deriving from incorrect detection is small and, thus, that the method is inherently robust.
In this paper, a vision-based system for underwater object detection is presented. The system is able to detect automatically a pipeline placed on the sea bottom, and some objects, e.g. trestles and anodes, placed in its neighborhoods. A color compensation procedure has been introduced in order to reduce problems connected with the light attenuation in the water. Artificial neural networks are then applied in order to classify in real-time the pixels of the input image into different classes, corresponding e.g. to different objects present in the observed scene. Geometric reasoning is applied to reduce the detection of false objects and to improve the accuracy of true detected objects. The results on real underwater images representing a pipeline structure in different scenarios are shown. The presence of seaweed and sand, different illumination conditions and water depth, different pipeline diameter and small variations of the camera tilt angle are considered to evaluate the algorithm performances.
S U M M A R YWe investigated the high frequency attenuation of S waves in the southeastern Alps and northern External Dinarides using waveforms from 331 earthquakes (3.0 < M w < 6.5). The spectral decay parameter, k, was computed using 1345 three component high quality records, collected by the Italian Strong Motion Network (RAN) and by the Short-Period Seismometric Network of northeastern Italy (NEI) in the period 1976-2007. Weak motion data from 11 stations of the NEI network and strong motion data collected by five accelerometers of the RAN were analysed. The k parameter was estimated in the 0-250 km distance range, in a frequency band extending from the corner frequency of the event up to 25 or 45 Hz, using the amplitude acceleration Fourier spectra of S waves. The observed record-to-record variability of k was modelled by applying a generalized inversion procedure, using both parametric and nonparametric approaches. Our results evidence that k is independent on earthquake size, while it shows both site and distance dependence. Stations of the NEI network present the same increase of k with epicentral distance, R E , and show values of the zero-distance k parameter, k 0 (S), between 0.017 and 0.053 s. For the whole region, the k increase with distance can be described through a linear model with slope dk/dR E = (1.4 ± 0.1) × 10 −4 s km -1 . Assuming an average S-wave velocity, β = 3.34 km s -1 between 5 and 15 km depth, we estimate an average frequency independent quality factor, Q I = 2140, for the corresponding crustal layer. The non-parametric approach evidences a weak positive concavity of the curve that describes the k increase with R E at about 90 km distance. This result can be approximated through a piecewise linear function with slopes of 1.0 × 10 −4 and 1.7 × 10 −4 s km -1 , in accordance with a three layers model where moving from the intermediate to the bottom layer both Q I and β decrease. Two regional dependences were found: data from earthquakes located westward to the NEI network evidence weaker attenuation properties, probably because of S-wave reflections from different parts of the Moho discontinuity under the eastern Po Plain, at about 25-30 km depth, while earthquakes located eastward (in western Slovenia), where the Moho deepens up to 45-50 km, evidence a higher attenuation. Moreover, the k estimates obtained with data from earthquakes located in the area of the 1998 (M w = 5.7) and 2004 (M w = 5.2) Kobarid events are 0.017 s higher than the values predicted for the whole region, probably because of the high level of fracturing that characterizes fault zones. The comparison between measured and theoretical values of k, computed at a few stations with available S-wave velocity profiles, reveals that the major contribution to the total k 0 (S) is due to the sedimentary column (from surface to 800 m depth). The hard rock section contribution is limited to 0.005 s, in accordance with a maximum contribution of 0.010 s predicted by the non-parametric inversion.
. The partitioning of radiated energy and the largest aftershock of seismic sequences occurred in the northeastern Italy and western Slovenia.
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.