← Back to Sessions

Advancing USGS Scientific Modeling Through Cloud Computing

Description: 

The U.S. Federal Government released an Executive Order to carry out a Cloud Smart strategy aimed at adopting cloud-based solutions for modernizing cyberinfrastructure, enhancing security, and providing the American people with higher quality services and products (https://cloud.cio.gov/strategy/). Software developers at the U.S. Geological Survey (USGS) Geologic Hazards Science Center (GHSC) are adopting this strategy by moving IT infrastructure from on-premises servers to the cloud. Here, we give an overview of how the GHSC is running various scientific applications on the cloud. We will discuss the hybrid cloud systems (a cloud architecture that lies between lift-and-shift and purely cloud native approaches) used for running 1) the USGS earthquake-triggered global ShakeMap and down-stream near-real-time Ground Failure product, 2) user-initiated high-throughput computing simulations of seismic site response using the finite element platform OpenSees, 3) a semi-automatic approach for running finite fault inversions, and 4) USGS slab models via a web form request. Adopting the Cloud Smart strategy has not only educated GHSC developers on the capabilities of cloud computing but has prompted the development of new USGS earthquake hazard products (e.g., the Sequence Product, which links USGS event pages for foreshocks, mainshocks, and aftershocks), and has shifted seismic workflows (e.g., aftershock relocations) to follow cloud-native serverless approaches. We will also address examples of how USGS scientists interact with these cloud systems, benefits and drawbacks of migrating to the cloud, associated costs, and how one can get started with cloud computing.

Session: Leveraging Cutting-Edge Cyberinfrastructure for Large Scale Data Analysis and Education - I

Type: Oral

Date: 5/2/2024

Presentation Time: 04:30 PM (local time)

Presenting Author: Kirstie

Student Presenter: No

Invited Presentation: Yes

Authors