← Back to Sessions

Investigating the Seismicity of Yucatan, Mexico, Using Machine Learning Techniques

Description: 

The Yucatan Peninsula is the subaerial portion of a carbonate platform deposited on the Yucatan Block unit, a continental microplate that covers southeastern Mexico, northern Guatemala and Belize. It has generally been considered tectonically stable, while on a broader time scale it has been interpreted to have been subsiding since the early/middle Eocene. In this study, we use machine learning techniques to investigate seismicity in Yucatán, southwestern Mexico. In this region, seismicity appears to be extremely low compared to the rest of the country. The coverage of seismic stations is also very poor, which could be one of the reasons for the unusual lack of reported earthquakes in this region. We employed the ConvNetQuake and EQTransformer neural network algorithms to identify and locate small earthquakes, not recorded in the catalogs, but hidden in the seismograms. We identified an average of 637 seismic events per year occurring in the studied region between 2007 and 2021. In addition, we investigate the possible relationship between some of these newly detected earthquakes and some of the 23 oil wells drilled in the Yucatan Peninsula since 2003. Our preliminary results show that the Yucatan Peninsula has a low seismicity, but contrary to national catalogs there are seismic events that have not been reported due to a significant lack of instrumentation. These events reflect important characteristics of the stress state of this region and we can use this information to study the orientation of the principal stresses and induced seismicity associated with oil and gas exploitation.

Session: Tectonics and Seismicity of Stable Continental Interiors [Poster]

Type: Poster

Date: 4/19/2023

Presentation Time: 08:00 AM (local time)

Presenting Author: Jorge Castro

Student Presenter: Yes

Invited Presentation: 

Authors