Methods of Time Series Data Mining: análisis de datos sísmicos para construir un mapa de riesgo sísmico del Ecuador.

Abstract

The seismic network of the Instituto Geofisico of the Escuela Politécnica Nacional has allowed to locate 71461 events between January of 2013 and May of 2022 with magnitudes between 2.0 and 7.6. Using this data, we computed and mapped the b-value of the Gutenberg-Richter distribution in Ecuador using the maximum likelihood estimation. We removed events with a very high uncertainty in their location and a magnitude of completeness Mc = 2.20 was found for the whole catalogue, reducing the number of events to 51397. We also removed events that we considered as aftershocks and a final catalogue that contains 14034 independent events was obtained. Then, we performed 3 methods for dividing the study region in different zones and computing b along the country. In each zone the b-value was calculated using Utsu equation based on MLE. We found b-values between 0.5 and 2.6 along the all the maps. The lower b-values were located in front of the coast of the country, the higher b-values were located between the western and the center region of the country and for the eastern region of the country the b-value couldn't be adequately mapped because it did not have enough events for the analysis. We also performed the same computation and mapping for different completeness magnitudes of Mc= 2.5, 3.0, 3.5 and 4.0.