Search

The increasing volume of seismic data from long-term continuous monitoring motivates the development of algorithms based on convolutional neural network (CNN) for faster and more reliable phase detection and picking. However, many less studied regions lack a significant amount of labeled events needed for traditional CNN approaches. In this paper, we present a CNN-based Phase-Identification Classifier (CPIC) designed for phase detection and picking on small to medium sized training datasets. When trained on 30,146 labeled phases and applied to one-month of continuous recordings during the aftershock sequences of the 2008 M W 7.9 Wenchuan Earthquake in Sichuan, China, CPIC detects 97.5% of the manually picked phases in the standard catalog and predicts their arrival times with a five-times improvement over the ObsPy AR picker. In addition, unlike other CNN-based approaches that require millions of training samples, when the off-line training set size of CPIC is reduced to only a few thousand training samples the accuracy stays above 95%. The deployment of CPIC takes less than 12 h to pick arrivals in 31-day recordings on 14 stations. In addition to the catalog phases manually picked by analysts, CPIC finds more phases for existing events and new events missed in the catalog. Among those additional detections, some are confirmed by a matched filter method while others require further investigation. Finally, when tested on a small dataset from a different region (Oklahoma, US), CPIC achieves 97% accuracy after fine tuning only the fully connected layer of the model. This result suggests that the CPIC developed in this study can be used to identify and pick P/S arrivals in other regions with no or minimum labeled phases. • Design and test a feasible CNN scheme for small to medium size training sets which is accurate and efficient compared to manual picking. • The aftershock dataset used in the test is known to be challenging for phase picking due to the existence of a large number of weak events and low signal-to-noise ratio. • The CNN structure we propose can be trained on the small training set (< 10, 000 labels) with 95% classiffication accuracy. • Deployment of the trained CNN is efficient: detection takes two hours and picking takes 12 hours to process 31 days of data on 14 stations. • Many weak events detected by the CNN but missed by manual pickings can be verified by a more complete catalog created via the matched filter approach. [ABSTRACT FROM AUTHOR]