Curtin University’s spatial sciences teams are using big data, advanced processing power and community engagement to solve social and environmental problems.
Advanced facilities and expertise at Perth’s Pawsey Supercomputing Centre support the Square Kilometre Array – a multi-array telescope due to launch in 2024 – and undertake high-end science using big data.
Individual computers at the $80 million facility have processing power in excess of a petaflop (one quadrillion floating point operations per second) – that’s 100,000 times the flops handled by your average Mac or PC.
Curtin University is a key participant in iVEC, which runs the Pawsey Centre, and a partner in the CRC for Spatial Information. As such, it is at the forefront of research efforts to use big data to solve global issues.
For instance, says the head of Curtin’s Department of Spatial Sciences Professor Bert Veenendaal, the university’s researchers are using Pawsey supercomputers to manage, compile and integrate growing volumes of data on water resources, land use, climate change and infrastructure.
“There is a rich repository of information and knowledge among the vast amounts of data captured by satellites, ground and mobile sensors, as well as the everyday usage information related to people carrying mobile devices,” he says.
“Increasing amounts of data are under-utilised because of a lack of knowhow and resources to integrate and extract useful knowledge,” he explains.
“Big data infrastructures coupled with increasing research in modelling and knowledge extraction will achieve this.”
Curtin’s projects include mapping sea-level rise and subsidence along the Western Australian coastline near Perth, generating high-resolution maps of the Earth’s gravity field and modelling climate over regional areas, such as Bhutan in South-East Asia, across multiple time scales.
Some research projects have the potential to expand and make use of big data in the future, particularly in the area of community development.
In one such project, the team worked with members of a rural community in the Kalahari Desert, Botswana, to collect information and map data using geographic information science.
This helped the local community to determine the extent of vegetation cover in their local area, water access points for animals and how far the animals travelled from the water points to food sources.
Using this data, one local woman was able to create a goat breeding business plan to develop a herd of stronger animals.
According to Veenendaal, there is potential for big data to be used for many regional and national issues.
“Projects like this have the potential to provide data acquisition, analysis and knowledge that will inform intelligent decision-making about land use and community development on local, regional and national scales,” he says.
While procuring more funding for the Botswana project, Curtin’s researchers are planning future big data projects, such as applying global climate change models to regional areas across multiple time scales, and bringing together signals from multiple global navigation satellite systems, such as the USA’s GPS, China’s BeiDou and the EU’s Galileo. – Laura Boness