Building a National High-Resolution Geophysics Reference Collection for 2030 Computation
- 1. Australian National University, nigel.rees@anu.edu.au
- 2. Australian National University, lesley.wyborn@anu.edu.au
- 3. Australian National University, ben.evans@anu.edu.au
- 4. University of Melbourne, rebecca@auscope.org.au
- 5. University of Melbourne, tim@auscope.org.au
- 6. Australian National University, rui.yang@anu.edu.au
- 7. Australian National University, Yue.s@anu.edu.au
Description
Large volumes of geophysical data have been acquired by universities, industry, federal/state government agencies since the 1950s. However, in many geophysical disciplines the valuable raw time series data has not been made publicly accessible and research geophysicists often have to go through a myriad of hurdles to gain access to raw and time series datasets of interest. In order to increase online collaboration, reduce time for analysis, and enable reproducibility and integrity of scientific discoveries, geophysical datasets will need to evolve to adopt the FAIR (Findable, Accessible, Interoperable, Reusable) data principles for both human and machine-to-machine interactions. The ARDC/AuScope/NCI/TERN funded National High-resolution Geophysics Reference Collections for 2030 Computation Project is working towards making accessible online the rawer, high-resolution versions of AuScope coinvested magnetotelluric and passive seismic datasets, ensuring that they comply with the FAIR data principles and can be integrated with existing government datasets on the NCI HPC platform. Targeted raw geophysical datasets will be ingested and organised at NCI so that they can be (re)processed with computational tools available within the NCI compute systems, and derivative versions of processed data can be linked back to the source datasets. Geophysical data releases will be discoverable through the NCI data catalogue and Research Data Australia: metadata will be structured to enable `vertical' integration between repositories that have higher level products, but still reference the rawer data at NCI. The project will make high-resolution geophysical datasets suitable for programmatic access in HPC environments at NCI with the intention of more easily enabling inter-geophysical disciplinary science. Managed geophysical software environments will be created that allow users the ability to fluently scale their Jupyter analysis notebooks to NCI's HPC Gadi system using CPUs and GPUs. This will ultimately lay the foundations for more rapid data processing by 2030 next-generation scalable, data-intensive computation including artificial intelligence, machine learning and data assimilation.
Notes
Files
AEGC_2023_ID111.pdf
Files
(322.5 kB)
Name | Size | Download all |
---|---|---|
md5:e17476dd03de6ac550270bb32215fb2c
|
322.5 kB | Preview Download |