Abstract
Understanding deep crustal structure can provide us with insights into tectonic processes and how they affect the geological record. The deep crustal structure can be studied using a suite of seismological techniques such as receiver function analysis, body and surface wave tomography. Using models of crustal structure derived from these methods, it is possible to delineate tectonic boundaries and regions that may have been affected by similar processes. However, often velocity models are grouped in a somewhat subjective manner, potentially meaning that some geological insight may be missed. Cluster analysis, based on unsupervised machine learning, can be used to more objectively group similar velocity profiles and, thus, put additional constraints on the deep crustal structure. In this study, we apply hierarchical agglomerative clustering to the shear wave velocity profiles obtained by Gilligan et. al. (2016) from the joint inversion of receiver functions and surface wave dispersion data at 59 sites surrounding the Hudson Bay. This location provides an ideal natural laboratory to study the Precambrian tectonic processes, including the 1.8Ga Trans-Hudson Orogen. We use Ward linkage to define the distance between clusters, as it gives the most physically realistic results, and after testing the number of clusters from 2 to 10, we find there are 5 main stable clusters of velocity models. We then compare our results with different inversion parameters, clustering schemes (K-means and GMM), as well as results obtained for profiles from receiver functions in different azimuths and find that, overall, the clustering results are consistent. The clusters that form correlate well with the surface geology, crustal thickness, regional tectonics, and previous geophysical studies concentrated on specific regions. The profiles in the Archean domains (Rae, Hearne, and Superior) are clearly distinguished from the profiles in regions influenced by Proterozoic orogenic events (Southern Baffin Island and Ungava Peninsula). Further, the crust of Melville Peninsula is found to be in the same cluster as the crust of the western coast of Ungava Peninsula, suggesting a similar crustal structure. Our study shows the promising use of unsupervised machine learning in interpreting deep crustal structures to gain new geological insights.
| Original language | English |
|---|---|
| Pages (from-to) | 359–375 |
| Number of pages | 17 |
| Journal | Geophysical Journal International |
| Volume | 233 |
| Issue number | 1 |
| Early online date | 22 Nov 2022 |
| DOIs | |
| Publication status | Published - Apr 2023 |
Bibliographical note
AcknowledgementsAll the clustering was performed using tools from Scikit-learn (Pedregosa et al. 2011), a free open source Machine Learning for python programming language. All the maps were created using PyGMT (Uieda et al. 2021). We would like to thank Dr. David Cornwell (University of Aberdeen) for helpful suggestions that resulted in the improvement of this manuscript, and the editor Ian Bastow and two anonymous reviewers for comments that have helped improve the manuscript.
Data Availability Statement
Codes reproducing our results are available in the form of properly documented Jupyter notebooks present in the github repository—https://github.com/Akashkharita/Cluster_Analysis_Hudson_Bay, All the velocity models used from Gilligan et al. (2016) are available from https://zenodo.org/record/6998279#.Yy28mHbMLIV. H–k values for the stations used in Thompson et al. (2010) and Vervaet & Darbyshire (2022) are available in their respective papers.Keywords
- Machine Learning
- Structure of the Earth
- statistical seismology