S017-09
Next-generation model of the Australian crust from synchronous and asynchronous ambient noise imaging
Wednesday, 9 December 2020: 04:34
Virtual
Yunfeng Chen1,2, Erdinc Saygin3,4, Brian L N Kennett5, Mehdi Tork Qashqai2 and Juerg R Hauser2, (1)University of Alberta, Edmonton, AB, Canada, (2)Deep Earth Imaging, CSIRO, Perth, WA, Australia, (3)Deep Earth Imaging Future Science Platform, CSIRO, Perth, WA, Australia, (4)University of Western Australia, Department of Physics, School of Physics, Mathematics and Computing,Faculty of Engineering and Mathematical Sciences, Perth, WA, Australia, (5)Australian National University, Canberra, ACT, Australia
Abstract:
The proliferation of seismic networks in Australia has laid the groundwork for improved probing the subsurface.
Despite ever-growing seismic instrumentation across the country, the last major effort of mapping continental-scale structures with seismic ambient noise was conducted more than a decade ago, thereby demanding a new appraisal of its crustal structure.
In this study, we develop a new crustal model of the Australian continent using a large dataset that consists of nearly 30 years (1992-2019) of continuous seismic recordings from over 2200 stations.
This unprecedented dataset is further exploited with the recently developed ambient noise imaging workflow of Chen & Saygin (2020) that integrates results from temporary seismic arrays deployed at different times. We compute two sets of noise correlation functions (NCFs) between 1) synchronous stations with the conventional ambient noise correlation (i.e., C
1) and 2) asynchronous pairs with the high-order correlation technique (i.e., C
2) based on correlational and convolutional types of source-receiver interferometry.
The C
2 approach enables extracting 1-3 times more NCFs than available from using C
1 alone, and the combined dataset leads to over 200,000 high-quality NCFs to image the crustal structures, significantly improved upon the most recent model constructed from 7500 measurements.
We invert the Rayleigh wave arrival times for group velocities between 4-40 sec using a trans-dimensional inversion. This non-linear inversion approach adopts an adaptive parameterization and Bayesian inference to account for unbalanced data sampling and allows assessment of the model uncertainties. The final 3D shear velocity model reveals fine-scale structure in the Australian crust. The low velocities at shallow depths (<10 km) are in excellent agreement with the distribution of known sedimentary basins and also hint at the presence of unreported basins/sub-basins in previously poorly explored areas, for example, northern Australia. At lower crustal depths (30-40 km), our model clearly delineates the boundaries of major Archean blocks such as the Yilgarn and Pilbara cratons in Western Australia. High-velocity structures also characterize the lower crust/uppermost mantle of the Phanerozoic New England Orogen near the eastern continent margin. In conclusion, this study provides significantly improved constraints on the shear velocity structures and builds a new basis for the next-generation crustal model of the Australian continent.
