Nowcasting Earthquakes in Southern California with Machine Learning: Correlations with Regional Tectonic Stress
Abstract
Earthquake nowcasting is a method to evaluate the current state of seismic hazard from large earthquakes. In this paper, we connect the temporal occurrence of the largest and most potentially destructive earthquakes in California since 1984 with a readily observable property of small earthquake seismicity in the region. Our method involves the calculation of the time history of the average radius (horizontal size or extent) of "bursts" of small earthquakes, in the time leading up to and following major earthquakes in the region. We observe that the radius systematically and gradually decreases leading up to major earthquakes, increasing suddenly and discontinuously following the event. This observable pattern resembles the long-hypothesized cycle of regional tectonic stress buildup and release, or elastic rebound, associated with large destructive earthquakes. We propose that the radius of these bursts might be considered to be a proxy variable for the changing state of regional stress in Southern California. We also discuss a second method based on analysis of spatial patterns of large earthquakes leading to a similar time series. Below we show the two time series, a) RG(t) and b) c(t) from 1984 through 2020. RG(t) is calculated as a filtered optimized ensemble average of radii of gyration of small earthquake bursts as a function of time. Note that the average radius of gyration of a cluster of events is often taken to be a measure of the correlation length in statistical mechanics. c(t) is the weighted correlation time series computed from principal component analysis of the gridded timeseries of small earthquake events. In both figures, vertical red dashed lines represent large earthquakes M 6.9, vertical dotted lines represent earthquakes 6 M 6.0. Note that the vertical scale is inverted so that small values of RG(t) and b) c(t) are at the top of the figure, so values increase towards the bottom. The horizontal green dash dot line in each figure represents a decision threshold (TW) for the time window TW = 3 years, that allows a determination of the amount of information contained in the time series. For the RG(t) time series at the left, (TW) = 4.15. For X(t) time series at right, (TW) = 1.216.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFM.T53B..06R