Publication Date:
2021-07-03
Description:
Subduction zones are monitored using space geodesy with increasing resolution, with the aim of better capturing the deformation accompanying the seismic cycle. Here, we investigate data characteristics that maximize the performance of a machine learning binary classifier predicting slip‐event imminence. We overcome the scarcity of recorded instances from real subduction zones using data from a seismotectonic analog model monitored with a spatially dense, continuously recording onshore geodetic network. We show that a 70–85 km‐wide coastal swath recording interseismic deformation gives the most important information on slip imminence. Prediction performances are mainly influenced by the alarm duration (amount of time that we consider an event as imminent), with density of stations and record length playing a secondary role. The techniques developed in this study are most likely applicable in regions of slow earthquakes, where stick‐slip‐like failures occur at time intervals of months to years.
Description:
Plain Language Summary:
Machine learning, a group of algorithms that produce predictions based on past “experience,” has been successfully used to predict various aspects of the earthquake process, including slip imminence. The accuracy of those algorithms depends on a variety of data characteristics, for example, the amount of data used for building the “experience” of the model. We focus on this point using a scaled representation of a seismic subduction zone and a monitoring technique similar to Global Navigation Satellite System. We identify the most useful surface regions to be monitored and the parameter that most strongly influences prediction accuracy for the timing of upcoming laboratory earthquakes. The routine implemented in this study could be used to predict the onset and extent of slow earthquakes.
Description:
Key Points:
We investigate the performances of a binary classifier predicting slip‐event imminence in analog models of megathrust seismic cycling.
A 70–85 km‐wide coastal swath is the region producing the most important information for the imminence classification.
Length of time that we consider an event imminent plays a primary role in tuning the performances of a binary classifier predicting the imminence of analog earthquakes.
Description:
DAAD‐Prime
Keywords:
551.22
;
megathrust earthquakes
;
machine learning
;
analog modeling
Type:
article
Permalink
