TY  - JOUR
A1  - Quinteros-Cartaya, Claudia
A1  - Köhler, Jonas
A1  - Li, Wei
A1  - Faber, Johannes
A1  - Srivastava, Nishtha
T1  - Exploring a CNN model for earthquake magnitude estimation using HR-GNSS data
T2  - Journal of South American earth sciences
N2  - Highlights
• We present the first results of a deep learning model based on a convolutional neural network for earthquake magnitude estimation, using HR-GNSS displacement time series.
• The influence of different dataset configurations, such as station numbers, epicentral distances, signal duration, and earthquake size, were analyzed to figure out how the model can be adapted to various scenarios.
• The model was tested using real data from different regions and magnitudes, resulting in the best cases with 0.09 ≤ RMS ≤ 0.33.
Abstract
High-rate Global Navigation Satellite System (HR-GNSS) data can be highly useful for earthquake analysis as it provides continuous high-frequency measurements of ground motion. This data can be used to analyze diverse parameters related to the seismic source and to assess the potential of an earthquake to prompt strong motions at certain distances and even generate tsunamis. In this work, we present the first results of a deep learning model based on a convolutional neural network for earthquake magnitude estimation, using HR-GNSS displacement time series. The influence of different dataset configurations, such as station numbers, epicentral distances, signal duration, and earthquake size, were analyzed to figure out how the model can be adapted to various scenarios. We explored the potential of the model for global application and compared its performance using both synthetic and real data from different seismogenic regions. The performance of our model at this stage was satisfactory in estimating earthquake magnitude from synthetic data with 0.07 ≤ RMS ≤ 0.11. Comparable results were observed in tests using synthetic data from a different region than the training data, with RMS ≤ 0.15. Furthermore, the model was tested using real data from different regions and magnitudes, resulting in the best cases with 0.09 ≤ RMS ≤ 0.33, provided that the data from a particular group of stations had similar epicentral distance constraints to those used during the model training. The robustness of the DL model can be improved to work independently from the window size of the time series and the number of stations, enabling faster estimation by the model using only near-field data. Overall, this study provides insights for the development of future DL approaches for earthquake magnitude estimation with HR-GNSS data, emphasizing the importance of proper handling and careful data selection for further model improvements.
KW  - Earthquake magnitude
KW  - Geodetic data
KW  - Deep learning
Y1  - 2024
UR  - http://publikationen.ub.uni-frankfurt.de/frontdoor/index/index/docId/82891
UR  - https://nbn-resolving.org/urn:nbn:de:hebis:30:3-828916
SN  - 0895-9811
VL  - 136
IS  - 104815
PB  - Elsevier
CY  - Amsterdam [u.a.]
ER  -