A non-parametric method for automatic determination ofP-wave andS-wave arrival times: application to local micro earthquakes

Crustal Fault Connectivity of the Mw 7.8 2016 Kaikōura Earthquake Constrained by Aftershock Relocations

“…Accurate identification of phase arrival times is essential to assuring high‐precision earthquake locations in standard workflows. We introduce here (see Text S1) a hybrid method for automated detection and onset estimation, called REST, which combines a modified version of the nearest‐neighbor similarity scheme of Rawles and Thurber () with the autoregressive approach of Kushnir et al (). We obtain accuracy information on the REST algorithm by comparing autopicks with picks made by analysts for a subset of earthquakes and to picks obtained with the nonparametric autopicker method of Rawles and Thurber (), kpick (Text S1).…”

Comparison of recent S-wave indicating methods

“…The method [4] was presented in 2015 and uses a nearest neighbours-based approach of Nikolov [14]. A method does not need to use parameters estimated from data, but uses data itself to build a model.…”

“…As the process of manual picking is time-consuming, methods of automatic detection are recommended (these however may be less accurate). In this paper four recently developed methods estimating S-wave arrival are compared: the method operating on empirical mode decomposition and Teager-Kaiser operator [2], the modification of STA/LTA algorithm [3], the method using a nearest neighbour-based approach [4] and the algorithm operating on characteristic of signals' second moments. The methods will be also compared to wellknown algorithm based on the autoregressive model [5].…”

“…In Rawles-Thurber method the thresholds th P and th S are appointed from the mesh of mean absolute errors between picks and the length of window was taken as in Rawles-Thurber paper [4]. The range of distance between P-wave and S-wave picks is determined on the base of reference picks in data sets.…”

An Automatic Phase Picker for Local Earthquakes with Predetermined Locations: Combining a Signal‐to‐Noise Ratio Detector with 1D Velocity Model Inversion

“…Beginning with the short-term average/long term average (STA/LTA) method (Allen 1978(Allen , 1982Earle and Shearer, 1994), a few higher order statistical metrics have been proposed, including envelope functions (e.g., Baer and Kradolfer, 1987) and skewness and kurtosis functions (e.g., Saragiotis et al, 2002;Kuperkoch et al, 2010;Baillard et al, 2014). Algorithms from the datamining community are also introduced for phase picking purpose, such as neural networks (Dai and MacBeth, 1997;Wang and Teng, 1997;Zhao and Takano, 1999;Gentili and Michelini, 2006), autoregressive methods (Takanami and Kitagawa, 1988;Leonard and Kennett, 1999;Sleeman and van Eck, 1999;Zhang et al, 2003), and nearest neighbors (Rawles and Thurber, 2015). Some studies combine these approaches to improve the robustness of the picking procedure (Sleeman and van Eck, 1999;Patane et al, 2003;Diehl et al, 2009;Ross and Ben-Zion, 2014).…”