Developing the Seismic Network Design App

Interactive visualisation of a minimum detectable earthquake magnitude modelled for a specific seismic network design. Image: Dr Pavel Golodoniuc

Interactive visualisation of a minimum detectable earthquake magnitude modelled for a specific seismic network design. Image: Dr Pavel Golodoniuc


In order to detect earthquakes effectively, seismologists must design spatial arrays of seismometers that can capture even the quietest sounds from under the ground. To help with this task, Dr Pavel Golodoniuc and AuScope’s AVRE team have worked with researchers from The University of Melbourne to develop the Seismic Network Design App, a free tool that allows seismologists to design and optimise seismic arrays, and a shining example of new analytics and computational possibilities via AuScope’s Engage Program, which is now open for new proposals.


The challenge

A major challenge when deploying an array of seismometers (seismic array) is predicting the smallest earthquake that could be detected and located by that network. Changing the spacing and number of seismometers dramatically affects sensitivity and precision and is very important when researchers need to identify very weak earthquakes. 

For cost reasons, it is important to optimise network design prior to deploying seismometers in the field. In doing so, seismologists must accurately account for parameters such as station locations, site-specific noise levels, source parameters, seismic velocity and attenuation in the wave propagation medium, the expected signal-to-noise ratio, and the minimum number of stations required to compute high quality locations.

Our approach

First, we worked with Dr Januka Attanayake and Abraham Jones and the seismology team at The University of Melbourne to better understand their solution for optimising seismic array design to date: an analytical method called SENSI that has been developed by Tramelli et al. (2013) to design seismic networks, including the GipNet array deployed to monitor seismicity in the Gippsland region in Victoria, Australia.

The method, originally implemented as a Fortran code, can flexibly handle source (source depth and stress drop), medium (density, P and S velocity, attenuation), and station (number of stations, geometry, station-specific noise) parameters. The underlying physics and mechanics of the code are straightforward, and when applied sensibly, it can be used as a basis for the design of seismic networks anywhere in the world.

Next, we imported this stand-alone modelling code into a scalable Cloud environment to allow for parallelisation, and built the Seismic Network Design App, a web application interface to allow interactive visualisations to facilitate human-computer interaction and experimentation. The app allows users to run simulations for a spread of input parameters and provides a variety of data visualisation options for comparative analysis.

Visualising seismic network sensitivity in Seismic Network Design App. Image: supplied

Visualising seismic network sensitivity in Seismic Network Design App. Image: supplied

Comparing modelling of a seismic network sensitivity for earthquakes at different depths. Image: supplied

Comparing modelling of a seismic network sensitivity for earthquakes at different depths. Image: supplied

The new Seismic Network Design App is now undergoing ‘field testing’ and will soon be released publicly for the benefit of national and, indeed, global audiences. Future development may include solutions for the inverse problem — finding the best seismic network design within the set boundary conditions.

Enabled by our AVRE Engage Program

Started in 2019, the AVRE Engage program assists the AuScope community in developing data access and analytics workflows and widgets to address real-world scientific problems through the provision of seed funding and computational expertise across a three months period. Engage projects vary in scope but all benefit from application of information technology in a wide range of fields: HPC/Cloud computing, code optimisation and parallelisation, scientific data visualisation, automation of data management and processing, data management and beyond.

Calling for proposals

We are currently seeking collaboration opportunities with Australian universities, government institutions and the industry on real-world scientific problems that would benefit from application of new technology, computing and storage infrastructures, data managements, etc. We are committing a large portion of the funds to engaging in activities with other AuScope components.

The AVRE Engage will work with representatives from other AuScope components and communities to develop proof of concept or complete workflows and widgets to assist in data delivery, web presence or meeting FAIR data requirements. The team is keen to hear from you about your project ideas and find a way to turn them into a reality.

 

 
 

AUTHORS
Written by
Dr Pavel Golodoniuc;
and edited by
Jo CondonDr Januka Attanayake and Abraham Jones

CALL FOR PROJECT PROPOSALS

MORE INFORMATION
If you would like to know more about the
AVRE Engage program, please contact Dr Pavel Golodoniuc or Dr Jens Klump

Questions about the Seismic Network Design App (SENSI) can be addressed by Dr Januka Attanayake, Research Fellow at the University of Melbourne