Reliable Real-Time Signal/Noise Discrimination with Deep and Shallow Machine Learning Classifiers

Men-Andrin Meier, Zachary E. Ross, Anshul Ramachandran, Ashwin Balakrishna, Peter Kundzicz, Suraj Nair, Zefeng Li, Egill Hauksson, & Thomas H. Heaton

Submitted August 16, 2018, SCEC Contribution #8837, 2018 SCEC Annual Meeting Poster #064

In Earthquake Early Warning (EEW), every sufficiently impulsive signal is potentially the first evidence for an unfolding large earthquake. More often than not, however, impulsive signals are mere nuisance signals, e.g. from a nearby airport or from an instrument malfunction. One of the most fundamental - and difficult - tasks in EEW is to rapidly and reliably discriminate between real local earthquake signals, and any kind of other signal. This discrimination is necessarily based on very little information, typically a few seconds worth of seismic waveforms from a small number of stations. As a result, current EEW systems struggle to avoid discrimination errors, and suffer from false and missed alerts. In this study we show how modern machine learning classifiers can strongly improve real-time signal/noise discrimination. We develop and compare a series of non-linear classifiers with variable architecture depths, including random forests, fully connected, convolutional (CNN) and recurrent neural networks, and a generative adversarial network (GAN). We train all classifiers on the same waveform data set that includes 374k 3-component local earthquake records with magnitudes M3.0-9.1, and 946k impulsive noise signals. We find that all classifiers outperform existing simple linear classifiers, and that the deep architectures significantly outperform the more simple ones. Using 3s long waveform snippets, the CNN and the GAN classifiers both reach 99.5% precision and 99.3% recall on an independent validation data set. Most misclassifications stem from impulsive teleseismic records, and from falsely labeled records in the data set. We show that, in turn, 95.5% of the false triggers on teleseismic waveforms can be avoided with a simple secondary random forest classifier. Our results suggest that machine learning classifiers can strongly improve the reliability and speed of EEW alerts.

Meier, M., Ross, Z. E., Ramachandran, A., Balakrishna, A., Kundzicz, P., Nair, S., Li, Z., Hauksson, E., & Heaton, T. H. (2018, 08). Reliable Real-Time Signal/Noise Discrimination with Deep and Shallow Machine Learning Classifiers. Poster Presentation at 2018 SCEC Annual Meeting.

Related Projects & Working Groups