Automatic Time Picking for Weak Seismic Phase in the Strong Noise and Interference Environment: An Hybrid Method Based on Array Similarity

Abstract

The extraction of travel-time curve of seismic phase is very important for the subsequent inference of the structural properties of underground media in seismology. In recent years, with the increase in the amount of data, manual processing is facing significant challenges, and automatic signal processing has gradually become the mainstream. According to the similarity of array signals and considering the elimination of outliers, we propose an improved multi-channel cross-correlation method using the L1 norm measure to obtain preliminary results, which builds on a new controllable measurement mode. Then, the post-correction step is carried out in combination with the signal gain property of beamforming technique. Based on these two methods, this paper proposes a new scheme of automatic arrival time picking. We apply the scheme to actual data to verify the effects of the two methods step by step. The entire scheme achieves fine results: direct water waves, seismic waves refracted by the crust and seismic waves reflected by the upper mantle are automatically detected. In addition, compared with the two traditional methods, the scheme proposed in this paper has a better overall effect and a reasonable computation cost.

Using Fuzzy C-Means Clustering to Determine First Arrival of Microseismic Recordings

“…Over the past few decades, researchers have proposed numerous well-established methods for first-arrival picking, such as the short/long time average ratio (STA/LTA) [ 7 ], Akaike information criterion (AIC) [ 8 ] and various improved approaches [ 9 , 10 ], wavelet denoising and its improved forms [ 11 , 12 ], PAI-S/K based on skewness and kurtosis [ 13 ], STK/LTK based on the long- and short-time window kurtosis ratio [ 14 ], the Markov optimal decision process [ 15 ], waveform similarity-based approaches [ 16 ], and improved multi-channel cross-correlation [ 17 ]. The STA/LTA offers the advantages of simplicity and computational efficiency in first-arrival picking.…”

Calculation of Theoretical Travel Time and Automatic Picking of Actual Travel Time in Seismic Data

“…da Silva and Corso proposed a robust method based on instantaneous-spectral Shannon entropy for capturing microseismic events in noisy environments without data preprocessing [22]. Wu et al proposed an improved multi-channel cross-correlation method [23]. Work on automatic picking through machine learning based technology is reported in [24,25].…”