Subsurface Imaging of the Lake George Basin, Eastern Australia through Seismic Interferometry of Traffic Noise

Traditional seismic imaging through use of an active source is not feasible in many areas, causing limitations in the ability to develop subsurface models. This prompts for innovative ways to circumnavigate this issue, one of which being the use of ambient noise as a source. An area that provides a suitable test case is the Lake George basin, covering an area of 150 km2 in south-eastern New South Wales, Australia. The lake is flanked to the west by the Lake George Fault and is part of the broader Great Dividing Range, a north-south chain of mountain ranges that extends almost 3500 km from the north-eastern tip of Queensland through to the western extents of Victoria. The fault is a west-dipping reverse fault, with the basin thought to have formed between ca. 3.93 Ma and the present due to fault propagation and folding. Little is known about the subsurface structure of the vast basin, including its relationship to the Lake George Range west of the fault. In late 2020, a dense array of 3-component 100 nodal seismometers were deployed for approximately one month to collect data in the north-western segment. The nodes were deployed in an elongated grid pattern with 30 50 m spacing. Given the proximity of these sensors to the Federal Highway, connecting Canberra to Sydney, an excellent opportunity presents itself to utilise the vehicle noise to assist in the 3-D rendering of the subsurface structure of the basin. To do this, we firstly conduct spectrum analysis to characterise the traffic signals amongst other noise present in the collected data. Beamforming analysis is performed to access spatial influence of traffic noise on wave excitation across the array. Then, we apply the cross-coherence method to the seismic interferometry of traffic noise and use array seismology to extract the fundamental and first high-mode surface waves (at 1-10 Hz), which allow us to constrain the shear-wave velocities in the top 100 meter of the basin for the first time. Whilst successes here will vastly improve the understanding of the Lake George region, it will also provide scalable opportunities to apply this method to a whole range of applications where conventional active source imaging is not feasible.