Deep Neural Networks for Creating Reliable PmP Database With a Case Study in Southern California. Issue 4 (18th April 2022)

• Record Type: Journal Article
• Authors: Ding, Wen; Li, Tianjue; Yang, Xu; Ren, Kui; Tong, Ping
• Is Part Of: Journal of geophysical research. Volume 127:Issue 4(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-04-18
• Subjects: PmP database -- deep neural networks -- unbalanced data set -- ResNet -- autoencoder -- southern California; Geomagnetism -- Periodicals; Geochemistry -- Periodicals; Geophysics -- Periodicals; Earth sciences -- Periodicals; 551.1
• DOI: 10.1029/2021JB023830
• Languages: English
• ISSNs: 2169-9313

Abstract

Abstract: Recent progresses in artificial intelligence and machine learning make it possible to automatically identify seismic phases from exponentially growing seismic data. Despite some exciting successes in automatic picking of the first P‐ and S‐wave arrivals, auto‐identification of later seismic phases such as the Moho‐reflected PmP waves remains a significant challenge in matching the performance of experienced analysts. The main difficulty of machine‐identifying PmP waves is that the identifiable PmP waves are rare, making the problem of identifying the PmP waves from a massive seismic database inherently unbalanced. In this work, by utilizing a high‐quality PmP data set (10, 192 manual picks) in southern California, we develop PmPNet, a deep‐neural‐network‐based algorithm to automatically identify PmP waves efficiently; by doing so, we accelerate the process of identifying the PmP waves. PmPNet applies similar techniques in the machine learning community to address the unbalancement of PmP datasets. The architecture of PmPNet is a residual neural network (ResNet)‐autoencoder with additional predictor block, where encoder, decoder, and predictor are equipped with ResNet connection. We conduct systematic research with field data, concluding that PmPNet can efficiently achieve high precision and high recall simultaneously to automatically identify PmP waves from a massive seismic database. Applying the pre‐trained PmPNet to the seismic database from January 1990 to December 1999 in southern California, we obtain nearly twice more PmP picks than the original PmP data set, providing valuable data for other studies such as mapping the topography of the Moho discontinuity and imaging the lower crust structures of southern California. Plain Language Summary: The success of machine learning in computer sciences, medical sciences, and many other fields has accelerated the implementation and development of machine learning techniques in seismology, making it possible to automatically identify seismic phases from the exponentially growing seismic data. At present, the auto‐identification of later seismic phases, such as the Moho‐reflected PmP waves, remains a significant challenge in matching the performance of experienced analysts. The main difficulty lies in the rare identifiable PmP waves, which makes the identification problem inherently unbalanced. In this work, by utilizing a high‐quality PmP data set in southern California, we develop a deep‐neural‐network‐based algorithm, PmPNet, to accelerate the process of identifying the PmP waves. We conduct systematic research with field data, and conclude that the PmPNet can efficiently achieve high precision and high recall simultaneously for automatically identifying the PmP waves from a massive seismic database. Applying the pre‐trained PmPNet to the seismic database from January 1990 to December 1999, we have tripled the PmP data set in southern California. Key Points: A deep‐neural‐network‐based algorithm, PmPNet, is developed to automatically pick the Moho‐reflected seismic waves PmP PmPNet efficiently achieves high precision and recall simultaneously to automatically identify PmP waves from a massive seismic database PmPNet creates a large PmP database with a total of 28, 093 high‐quality PmP picks in southern California