AGOS surface and borehole array
As part of the Australian Geophysical Observing System (AGOS) the University of Melbourne has established a seismic monitoring network in Victoria's South Gippsland region, in order to better understand natural earthquake occurrence in one of Australia's most seismically active regions.
To the end of 2014, the network comprised 8 surface seismic stations and 4 borehole instruments (Figure 1). In addition, the AuScope Seismometer in Schools (AuSIS) program has a stations at Sale and Rosebud.
The first deployment of surface instruments occurred in the month prior to the M 5.2 Moe earthquake (2012-06-19:10:56). Following the main shock three additional surface stations (Fish Creek, Willow Grove, Loch) were deployed in 2012. In 2014, Somers was deployed. There are currently 8 surface instruments deployed in Gippsland.
In October 2014, 3 borehole instruments were deployed along the Strzelecki hills, at depths of ~130 m. These instruments increase the spatial resolution of the Gippsland array, and offer a chance to monitor microseismicity that would otherwise go undetected and/or poorly resolved. in this note we highlight some of the features resolved by having the subsurface capability. An additional instrument (A346 on map) installed by a mining company at a depth of 1.3 km was integrated into our network, with this data also telemetered in real time.
Since their deployment a number of earthquakes have been detected with the new borehole instruments (Figure 1), including a sequence of low magnitude aftershock events following the M3.2 Moe earthquake of October 17, 2014. As our processing and analysis methods develop, we envisage that much smaller events will emerge from the data.
A small event near Korumburra (M 1.5, 2014-10-10UTC10:20) provided an opportunity to compare waveforms on surface and subsurface instruments. While P-wave arrivals can be easily detected on borehole stations, they are difficult to detect on surface stations.
Figure 2 shows this event at the nearest borehole station, and Figure 3 shows the corresponding STA/LTA filter response, which is commonly used for automatic event detection. Figure 3 shows the P and S-wave picks from the borehole station furthest from the event.
Figure 4 shows the event at the nearest telemetered surface staion. Whilst the event is clearly present on the surface record, the signal to noise level is much smaller. Also, no clear S-wave arrival can be determined. The corresponding STA/LTA filter shows multiple spikes of similar amplitude, meaning that any automatic detection routine using a STA/LTA filter is ambiguous.
A magnitude 3.2 earthquake felt locally in the Moe earthquake occurred within 10 Kms of a borehole instrument on 17 October, 2014. This allowed us to trial a Non-linear earthquake location method, which we are currently developing as the network -standard. The results of the location are shown in Figure 8.
We are currently developing tools for the automatic detection and location of microseismicity, as well as the real-time noise cross-correlation of ambient noise from all stations. In combination with other AGOS geophysical data, we are building a state-of-the art seismic monitoring laboritory in a region of significant (intraplate) earthquake activity.
All software developed as part of the project is available at https://github.com/unimelb-geophys. Ongoing data collection will contribute to future versions of the Victorian Earthquake Hazard Map (VEHM).