1. An Adaptable Random Forest Model for the Declustering of Earthquake Catalogs.
- Author
-
Aden‐Antoniów, F., Frank, W. B., and Seydoux, L.
- Subjects
- *
EARTHQUAKES , *CATALOGS , *FAULT zones , *GEOLOGIC faults , *SEISMOLOGY - Abstract
Earthquake catalogs are essential to analyze the evolution of active fault systems. The background seismicity rate, or rate of earthquakes that are not directly triggered by other earthquakes, directly relates to the stressing rate, a crucial quantity for understanding the seismic hazards. Determining the background seismicity rate is challenging because aftershock sequences may dominate the seismicity rate. Classifying these events in earthquake catalogs—known as catalog declustering—is a common practice and most declustering solutions rely on spatiotemporal distances between events, such as the nearest‐neighbor‐distance algorithm, widely used in various contexts. This algorithm assumes that the nearest‐neighbor distance (NND) follows a bimodal distribution related to the background seismicity and to the aftershocks. Constraining these two distributions is crucial to distinguish the aftershocks from the background events accurately. Recent work often uses linear splitting based on the NND, ignoring the potential overlap between the two populations and resulting in a biased identification of background earthquakes and aftershock sequences. We revisit this problem with machine‐learning algorithms. After testing several popular algorithms, we show that a random forest trained with various synthetic catalogs generated by an Epidemic‐Type Aftershock Sequence model outperforms approaches such as k‐means, Gaussian‐mixture models, and support vector machine classification. We apply our model to two different earthquake catalogs: the relocated Southern California earthquake center catalog and the GeoNet catalog of New Zealand. Our model capably adapts to these two different tectonic contexts, highlighting the differences in aftershock productivity between crustal and intermediate‐depth seismicity. Plain Language Summary: Earthquakes rupture seismic faults when the fault can no longer support the stress built up by tectonic motion. Earthquake catalogs are thus a window into the tectonic processes that occur at depth. Background earthquakes occur spontaneously from tectonic stresses, creating a stress perturbation that generally triggers aftershocks. These aftershocks can also trigger other aftershocks, creating aftershock sequences that may dominate earthquakes catalogs. The accurate determination of the background earthquakes' occurrence is crucial, as it directly relates to the fault's stress state and consequently favors a better understanding of the seismic hazard. To solve this problem, we develop and train a machine‐learning model to classify background earthquakes and aftershocks in earthquake catalogs. After training our model on synthetic earthquakes catalogs, we apply it to several real cases from New Zealand and Southern California to show the effectiveness of our approach. Our results suggest that our method is adaptable to any region, independent of the style of seismicity or the catalog duration. Key Points: The nearest‐neighbor distance metric is a suitable representation of an earthquake catalog for machine‐learning modelsBoth supervised and unsupervised learning algorithms can separate aftershocks from background seismicity in synthetic catalogsA random forest classifier most efficiently declusters real earthquake catalogs in different tectonic contexts [ABSTRACT FROM AUTHOR]
- Published
- 2022
- Full Text
- View/download PDF