Partners: Marine Denolle, Yiyu Ni, and Kuan-Fu Feng
SSEC Engineers: Carlos Garcia Jurado Suarez
Research Goals and Domain
The Earth’s ambient field contains a great deal of information about its structure. Changes in this structure occur at a vast range of temporal and spatial scales. We can record these changes thanks to the long-term deployment of seismic networks and the rapid expansion of sensing technologies. These data volumes are exponentially growing, along with the number of sensors.
Software Problem
Leveraging the data resulting from this growth of sensors is a challenge for the seismological community. Researchers need user-friendly, scalable software that can run on modern cyberinfrastructure. NoisePy is a software package used to image the Earth’s subsurface and monitor its changes through time. The computational needs of these calculations and data exceed what is possible on local machines. Furthermore, at the time of collaboration, the package existed across two repositories with differing implementations.
Software Solution
SSEC worked with researchers in UW’s Denolle Quake Lab to combine the NoisePy codebase into a single Python package that maintains performance while scaling capacity. It was adapted for the cloud, allowing users to leverage AWS Batch compute with their own data, using NoisePy to analyze large volumes of seismic readings.
Impact
In May 2023, the team introduced the software in a workshop, successfully engaging the seismology community. The package now enables scientists to calculate high-volume cross correlations and is set up to handle up to 2TB of data daily.