Automated Detection and Characterization of

Swarms and Mainshock-Aftershock Sequences in

Nicaragua and Costa Rica

By Lina Miesse, Michael Brudzinski, Wilnelly Ventura-Valentin

Poster at a glance

Poster at a glance

Authors

Lina Miesse

3rd year Environmental Earth Science Major, Sustainability Co-major, M.En combined student (co-author)

Michael Brudzinski

Department of Geology and Environmental Earth Science (advisor, co-author)

Wilnelly Ventura-Valentin

Department of Geology and Environmental Earth Science Ph.D candidate (co-author)

Abstract

A new methodology for the automated identification and characterization of swarms and Mainshock-Aftershock (MsAs) sequences was recently developed and applied to seismicity in the Mexico subduction zone (Ventura-Valentin et al., in review). This method uses a nearest neighbor approach to identify clusters (Zaliapin & Ben-Zion, 2013) and then integrates five quantitative characteristics derived from Omori, Båth, and Gutenberg-Richter laws to differentiate swarms from MsAs. Specifically, the algorithm calculates these attributes: magnitude difference, mainshock order, rate decay, magnitude decay, ratio of magnitude range to number of events. Intriguingly, the analysis in Mexico found twice as many swarms as MsAs. Our project is investigating whether this method produces a similar result in the nearby portion of the Central American subduction zone in Nicaragua and Costa Rica.

Introduction

Earthquakes clustered in space and time are typically grouped into two main categories, either a Mainshock-Aftershock (MsAs) or a Swarm sequence
Swarms tend to be less studied than MsAs with their origins remaining enigmatic and operational forecasting continues to be challenging
A newly developed methodology for automated identification & characterization of Swarms and Mainshock-Aftershock (MsAS) sequences was recently applied to seismicity in the Mexico subduction zone (Ventura-Valentin et al., in revision), revealing several hundred swarms that were not previously recognized
This project investigates whether this method produces similar results in the nearby portion of the Central American Subduction zone in Nicaragua & Costa Rica.

Figure 1a. Example of mainshock-aftershock event, with labeled key features

Figure 1b. Example of swarm event, with labeled key features

Data & Methods

Methodology Step-by-Step

Seismicity catalogs collected from the USGS, RSN, Georgia Tech
Applied Ventura-Valentin (in revision) technique to identify sequences
Manually rated sequences to confirm automated characterization method
Identified distribution of MsAs and swarm sequences

Automated sequence detection

The nearest neighbor approach is used to identify cluster events (Zaliapin & Ben-Zion, 2013), shown in figure 2.

Figure 2. Example of a detected sequence (blue) shown in Longitude and Latitude over time. Circles shows earthquakes and lines show neighbor relationships.

Components of Automated Discernment of Swarms and MsAs

Lack of Foreshocks

1) Percentage of the sequence yet to occur when the largest event occurs

Båth's Law

2) Magnitude differences between the largest event and the next 4 largest events

Omori’s Law

3) p-value of the power-law decay rate

3) Slope of seismicity rate over time based on dividing seismicity into 5 time bins

4) Slope of maximum magnitude over time based on dividing seismicity into 5 time bins

Ratio of magnitude to number of events

5) (Maximum – minimum magnitude) / Number of events

Conclusions

The automated detection and characterization technique fared well when deployed in a new region with automated ratings similar to manual characterization
Overall behavior of earthquake sequences along Nicaragua and Costa Rica differed from that in Mexico despite being in the same Central American Subduction Zone
Unlike Mexico where this technique produced twice as many Swarms as MsAs, we found the USGS and RSN catalogs identified twice as many MsAs as Swarms.
There does appear to be some similarity as swarms are more common at sliver faults in both regions
Mainshock-aftershock events were more common overall offshore and swarms were more common onshore
Swarm events did also appear to have some relationship to volcanoes, such Momotombo, Poas, and Arenal

Figures

Seismicity Catalogs

Figure 3a. Magnitudes over time for the Nicaragua and Costa Rica area catalog from the USGS NEIC.

Figure 3b. Magnitudes over time for the Costa Rica area catalog from Arroyo-Solózano & Linkimer, 2021 and Red Seismológica Nacional de Costa Rica, 2021.

Figure 3c. Magnitudes over time for the Nicaragua area catalog from Kyriakopoulos et al. [2015].

Manual Rating vs. Automated Rating, Visualized

Figure 4. Correlation plot between automated (y-axis) and manual (x-axis) characterization for each sequence in the USGS catalog.

While manual ratings tended to be more conservative and only the automated method could surpass 1, both methods are generally in agreement

Characterization

Figure 5a. RSN: Costa Rica

Figure 5b. USGS: Nicaragua & Costa Rica

Figure 5c. SSN: Mexico

Figure 5. Histograms illustrate the number of sequences with at least 10 events with that automated rating. a) USGS catalog, b) RSN catalog, c) results for Mexico from Ventura-Valentin for reference. The number of detected sequences was too small in the Nicaragua catalog for histograms to be appropriate. Color scale is tied to the rating to highlight swarms as blue and MsAs as red.

Figure 6. Map depicting automatically identified cluster events for a) USGS, b) RSN, and c) Nicaragua catalogs. Each symbol represents the average location of all epicenters within that sequence (not individual events). Color scale shows automated rating from technique of Ventura-Valentin (in revision).

References

Arroyo-Solózano, M., y Linkimer L. (2021). Spatial variability of the b-value and seismic potential in Costa Rica. Tectonophysics, 814, 228951. https://doi.org/10.1016/j.tecto.2021.228951
Kyriakopoulos, C., A. V. Newman, A. M. Thomas, M. Moore-Driskell, and G. T. Farmer (2015), A new seismically constrained subduction interface model for Central America, J. Geophys. Res. Solid Earth, 120, doi:10.1002/2014JB011859.
Red Sismológica Nacional de Costa Rica (2021). The Costa Rica National Seismological Network Catalog during 1975-2020. DOI: https://doi.org/10.15517/TC
Zaliapin, I., & Ben‐Zion, Y. (2013). Earthquake clusters in southern California I: Identification and stability. Journal of Geophysical Research: Solid Earth, 118(6), 2847-2864.

Acknowledgements

We thank Jens Mueller and Andrew Newman for their contributions to the data collection and analysis process. We thank Erin Szucs, Sharif Coker, James Kirchenwitz, and Mehrnaz Khalkhali for the constructive feedback. We thank the Miami University Family fund and Miami University chapter of Louis Stokes Alliance for Minority Participation for the funds and resources that made attending this conference possible.

Page updated

Google Sites

Report abuse