Abstract
The accuracy of P-wave arrival picking is essential for seismic analysis. The improvement in the accuracy of P-wave arrival picking is generally achieved through improved algorithms and the processing of waveforms. Therefore, we propose a method that uses deep learning to detect local windows to enhance the accuracy of P-wave arrival picking. The local window is defined as a short time window containing the main components of the signal. The faster-RCNN model is trained on the dataset with the calibrated local window. The trained faster-RCNN model is used for the local window detection of new records, and the existing algorithm is going to work in the local window. As a validation, four kinds of automatic P-wave arrival picking algorithms (wavelet-transform-based approach, PphasePicker algorithm, STAFD/LTAFD algorithm, and deep learning method) are used to conduct experiments in synthetic seismic records and field seismic records, respectively. The field experimental results show that the method proposed in this article can improve the picking capacity of the four methods by 17.5%, 37.6%, 62.4%, and 46.8%, respectively. No matter which algorithm is used, the accuracy of P-wave arrival picking in the local window is generally enhanced. The method presented in this article has a positive effect on improving the accuracy of seismic records.
Highlights
P-wave arrival picking is a crucial step in seismic analysis, and it is the premise of the event location, source mechanisms calculation, origin time determination, and subsurface velocity inversion
It was demonstrated that the local window detection via faster-RCNN could enhance the ability of the P-wave arrival picking
This article proposes a method to enhance the accuracy of P-wave arrival picking by detecting local window from the waveform through deep learning
Summary
P-wave arrival picking is a crucial step in seismic analysis, and it is the premise of the event location, source mechanisms calculation, origin time determination, and subsurface velocity inversion. The P-wave picking has always been a drag in the whole seismic analysis process. Manual P-wave arrival picking has become a very exhausting task for analysts. Finding a way to pick the P-wave arrival automatically and accurately is suitable for earthquake early warning and allows analysts to shift their energy to more meaningful tasks. Many methods have been proposed to automate the P-wave picking and have achieved excellent results. These methods can be roughly divided into traditional
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
