Ground-Motion Prediction for Earthquake Early Warning

Position Description: When an earthquake occurs, it is important to rapidly process and report earthquake properties and ground motion intensity for earthquake early warning (EEW) and seismic risk mitigation. These properties include earthquake magnitude, location, and focal mechanism. Seismologists have leveraged the recent advances of deep learning (DL) algorithms to accelerate the automated analysis of large seismic datasets thanks to their compact form and overall computational efficiency over conventional algorithms. Additionally, the whole ground motion intensities in space of interest could not be predicted effectively based on the current ground motion prediction equations (GMPEs). The accuracy of predicted ground motion intensities may have a great impact on the people’s lives and properties. Therefore, the quality of EEW and risk evaluation will mainly depend on the effectiveness and efficiency of determining earthquake properties and resultant shaking intensity alert. In this project, apart from looking back to the traditional EEW workflow, we are going to advance into a new EEW scheme by developing a graph-neural-network based approach for direct ground motion prediction, which will be independent of the earthquake properties and therefore realize just-in-time earthquake early warning. We will apply this method for the earthquake catalog from Southern California seismic network with over 100,000 earthquakes occurred in the past decade. The success of this study would have a valuable contribution to the EEW community. The team work will include: 1) Review and learn EEW fundamentals and advances; 2) Learn how to realize conventional ground motion prediction by using python scripts; 3) Learn basic seismology knowledge, seismic data processing, and machine learning theory; 4) Learn how to analyze and evaluate the results of conventional and new ground motion prediction approaches.

Project Dates: June 13-July 1, 2022

Number of Available Positions: 1-2

Location: In-person or remote

Pay Rate: $700 stipend at completion of program

Qualifications/Requirements:

• High school mathematics

• High school physics

• Computer-related skills (experience with code writing preferred)

• Computer and internet connection required (OS preferred)

Project host requests that the student participate in various department activities.

The Department of Earth and Planetary Sciences (EPS) welcomes everyone and aims for a diverse and inclusive community. Preference will be given to freshman and sophomore students, but we encourage all interested students to apply. Students who are already working with members of the EPS community and non-Harvard students are not eligible for this position.