OKSP: a novel deep learning automatic event detection pipeline for seismic monitoring in Costa Rica

Small magnitude earthquakes are the most abundant but the most difficult to locate robustly and well due to their low amplitudes and high frequencies usually obscured by heterogeneous noise sources. They highlight crucial information about the stress state and the spatio-temporal behavior of fault s...

Full description

Main Authors: Van der Laat, Leonardo, Baldares, Ronald, Chaves, Esteban, Meneses, Esteban
Format: Otro
Language: Inglés
Published: Instituto de Ingenieros Eléctricos y Electrónicos (IEEE) 2024
Subjects:
Online Access: http://hdl.handle.net/11056/27678
https://doi.org/10.48550/arXiv.2109.02723
id RepoUNACR27678
recordtype dspace
spelling RepoUNACR276782024-04-17T18:36:15Z OKSP: a novel deep learning automatic event detection pipeline for seismic monitoring in Costa Rica Van der Laat, Leonardo Baldares, Ronald Chaves, Esteban Meneses, Esteban ALGORITMOS ALGORITHMS DETECCIÓN SÍSMICA SEISMIC DETECTION PRUEBAS TESTS TERREMOTOS EARTHQUAKES SISMOLOGÍA SEISMOLOGY Small magnitude earthquakes are the most abundant but the most difficult to locate robustly and well due to their low amplitudes and high frequencies usually obscured by heterogeneous noise sources. They highlight crucial information about the stress state and the spatio-temporal behavior of fault systems during the earthquake cycle, therefore, its full characterization is then crucial for improving earthquake hazard assessment. Modern deep learning algorithms along with the increasing computational power and efficiency are exploiting the continuously growing seismological databases, worldwide, allowing scientists to improve the completeness for earthquake catalogs, systematically detecting and locating smaller magnitude earthquakes and reducing the errors introduced mainly by human intervention through traditional approaches in seismological observatories. In this work, we introduce OKSP, a novel deep learning automatic earthquake detection pipeline for seismic monitoring in Costa Rica. Using Kabre supercomputer from the Costa Rica High Technology Center, we applied OKSP to the day before and the first 5 days following the Puerto Armuelles, M6.5, earthquake that occurred on 26 June, 2019, along the Costa Rica-Panama border and found 1100 more earthquakes previously unidentified by the Volcanological and Seismological Observatory of Costa Rica. From these events, a total of 23 earthquakes with magnitudes below 1.0 occurred a day to hours prior to the mainshock, shedding light about the rupture initiation and earthquake interaction leading to the occurrence of this productive seismic sequence. Our observations show that for the study period, the model was 100% exhaustive and 82% precise, resulting in an F1 score of 0.90. This effort represents the very first attempt for automatically detecting earthquakes in Costa Rica using deep learning methods and demonstrates that, in the near future, earthquake monitoring routines will be carried out entirely by AI algorithms. Los terremotos de pequeña magnitud son los más abundantes, pero los más difíciles de localizar bien y con solidez debido a sus bajas amplitudes y altas frecuencias, normalmente oscurecidas por fuentes de ruido heterogéneas. Destacan información crucial sobre el estado de tensión y el comportamiento espacio-temporal de los sistemas de fallas durante el ciclo sísmico, por lo tanto, su caracterización completa es entonces crucial para mejorar la evaluación del peligro sísmico. Los algoritmos modernos de aprendizaje profundo junto con la creciente potencia y eficiencia computacional están explotando las bases de datos sismológicas en continuo crecimiento, en todo el mundo, permitiendo a los científicos mejorar la exhaustividad para los catálogos de terremotos, detectando y localizando sistemáticamente terremotos de menor magnitud y reduciendo los errores introducidos principalmente por la intervención humana a través de enfoques tradicionales en los observatorios sismológicos. En este trabajo, presentamos OKSP, una novedosa tubería de detección automática de terremotos de aprendizaje profundo para el monitoreo sísmico en Costa Rica. Usando la supercomputadora Kabre del Centro de Alta Tecnología de Costa Rica, aplicamos OKSP al día anterior y a los primeros 5 días posteriores al terremoto de Puerto Armuelles, M6,5, que ocurrió el 26 de junio de 2019, a lo largo de la frontera entre Costa Rica y Panamá, y encontramos 1100 terremotos más previamente no identificados por el Observatorio Vulcanológico y Sismológico de Costa Rica. De estos eventos, un total de 23 terremotos con magnitudes inferiores a 1,0 se produjeron un día u horas antes de la sacudida principal, arrojando luz sobre el inicio de la ruptura y la interacción sísmica que condujo a la ocurrencia de esta productiva secuencia sísmica. Nuestras observaciones muestran que, para el periodo estudiado, el modelo fue 100% exhaustivo y 82% preciso, lo que dio como resultado una puntuación F1 de 0,90. Este esfuerzo representa el primer intento de detección automática de terremotos en Costa Rica utilizando métodos de aprendizaje profundo y demuestra que, en un futuro próximo, las rutinas de monitoreo de terremotos serán llevadas a cabo enteramente por algoritmos de IA. Universidad Nacional, Costa Rica Instituto Tecnológico de Costa Rica Centro Nacional de Alta Tecnología, Costa Rica Observatorio Vulcanológico y Sismológico de Costa Rica 2024-04-17T18:33:35Z 2024-04-17T18:33:35Z 2021 http://purl.org/coar/resource_type/c_8544 http://hdl.handle.net/11056/27678 https://doi.org/10.48550/arXiv.2109.02723 eng Acceso abierto Atribución-NoComercial-CompartirIgual 4.0 Internacional http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Instituto de Ingenieros Eléctricos y Electrónicos (IEEE) IEEE International Conference on BioInspired Processing (BIP) (3ra: 2021: noviembre 4-5: Costa Rica)
institution Universidad Nacional de Costa Rica
collection Repositorio UNA-Costa Rica
language Inglés
topic ALGORITMOS
ALGORITHMS
DETECCIÓN SÍSMICA
SEISMIC DETECTION
PRUEBAS
TESTS
TERREMOTOS
EARTHQUAKES
SISMOLOGÍA
SEISMOLOGY
spellingShingle ALGORITMOS
ALGORITHMS
DETECCIÓN SÍSMICA
SEISMIC DETECTION
PRUEBAS
TESTS
TERREMOTOS
EARTHQUAKES
SISMOLOGÍA
SEISMOLOGY
Van der Laat, Leonardo
Baldares, Ronald
Chaves, Esteban
Meneses, Esteban
OKSP: a novel deep learning automatic event detection pipeline for seismic monitoring in Costa Rica
description Small magnitude earthquakes are the most abundant but the most difficult to locate robustly and well due to their low amplitudes and high frequencies usually obscured by heterogeneous noise sources. They highlight crucial information about the stress state and the spatio-temporal behavior of fault systems during the earthquake cycle, therefore, its full characterization is then crucial for improving earthquake hazard assessment. Modern deep learning algorithms along with the increasing computational power and efficiency are exploiting the continuously growing seismological databases, worldwide, allowing scientists to improve the completeness for earthquake catalogs, systematically detecting and locating smaller magnitude earthquakes and reducing the errors introduced mainly by human intervention through traditional approaches in seismological observatories. In this work, we introduce OKSP, a novel deep learning automatic earthquake detection pipeline for seismic monitoring in Costa Rica. Using Kabre supercomputer from the Costa Rica High Technology Center, we applied OKSP to the day before and the first 5 days following the Puerto Armuelles, M6.5, earthquake that occurred on 26 June, 2019, along the Costa Rica-Panama border and found 1100 more earthquakes previously unidentified by the Volcanological and Seismological Observatory of Costa Rica. From these events, a total of 23 earthquakes with magnitudes below 1.0 occurred a day to hours prior to the mainshock, shedding light about the rupture initiation and earthquake interaction leading to the occurrence of this productive seismic sequence. Our observations show that for the study period, the model was 100% exhaustive and 82% precise, resulting in an F1 score of 0.90. This effort represents the very first attempt for automatically detecting earthquakes in Costa Rica using deep learning methods and demonstrates that, in the near future, earthquake monitoring routines will be carried out entirely by AI algorithms.
format Otro
author Van der Laat, Leonardo
Baldares, Ronald
Chaves, Esteban
Meneses, Esteban
author_sort Van der Laat, Leonardo
title OKSP: a novel deep learning automatic event detection pipeline for seismic monitoring in Costa Rica
title_short OKSP: a novel deep learning automatic event detection pipeline for seismic monitoring in Costa Rica
title_full OKSP: a novel deep learning automatic event detection pipeline for seismic monitoring in Costa Rica
title_fullStr OKSP: a novel deep learning automatic event detection pipeline for seismic monitoring in Costa Rica
title_full_unstemmed OKSP: a novel deep learning automatic event detection pipeline for seismic monitoring in Costa Rica
title_sort oksp: a novel deep learning automatic event detection pipeline for seismic monitoring in costa rica
publisher Instituto de Ingenieros Eléctricos y Electrónicos (IEEE)
publishDate 2024
url http://hdl.handle.net/11056/27678
https://doi.org/10.48550/arXiv.2109.02723
_version_ 1796728482073411584
score 12.248849