Link to PDF Version.
Project Hypothesis or Objectives:
USGS ShakeAlert is an earthquake early warning (EEW) product currently live in California and being improved and developed for the entire US West Coast. Propagation of Locally Undamped Ground Motion (PLUM) is an EEW algorithm originally developed in Japan that we are adapting for use in the US. PLUM continuously monitors ground-shaking data and, if ground motion at any one station exceeds a certain level, sends an alert to nearby locations. In order to get the most timely and accurate alerts, we seek to improve on the ground-motion accuracy of the detections and alerts. One part of this improved accuracy is assessing the ground motion response of different seismic stations, dependent on local geology, velocity or topography. Ideally, station-specific adjustment terms can be applied to create more specific alerts. However, these types of adjustments are not currently available as needed for this project. The intern would work directly with the PLUM team to develop station-specific terms in ground motion intensity (Modified Mercalli Intensity) by considering existing work and methods, acquiring seismic data, testing the available adjustments, and, if necessary, developing novel MMI-based station factors. The intern would test the improved ground-motion accuracy of detections and alerts in the PLUM system across the West Coast of the US.
Duration: 2 - 6 months
Location: Moffett Field, Mountain View, CA
Area of Discipline: Earthquakes, Geology, Geophysics, Statistics
Intern Type Preference: Any Type of Intern
Applicable NSF Division: GEO (Atmospheric, Earth Sciences, Ocean Sciences, Polar Programs), ENG (Engineering)
Expected Outcome:
Ultimately, the intern will develop a set of station-specific adjustment terms that can be used in the PLUM algorithm for EEW on the West Coast of the US. This may consist of adaptation of existing terms or development of totally novel factors. Consideration of the underlying geology, physics or topography and correlations of these parameters with the observed ground shaking will allow extrapolation to new regions. The intern will also implement these new terms into the development PLUM system to assess if the improve the alert accuracy. The intern will present this work internally to the PLUM group, to the larger ShakeAlert group as necessary, at scientific meetings and ultimately produce (with guidance) a journal manuscript or report. The intern will learn valuable scientific skills, from developing a hypothesis, literature review, adaptation of existing models, presentation skills, and troubleshooting, with support from a postdoc, the PI and also the PLUM team. The USGS will gain these necessary site-specific adjustment terms critical to PLUM which are currently absent. The addition of this information will help complete the PLUM algorithm to prepare it to become part of the live USGS ShakeAlert system. The PLUM team and USGS as a whole does not currently have the person-power, time or bandwidth to take on this project, and it is specific enough that a student intern could be sucessful.
Special skills/training Required:
The intern should have experience in scientific programming, such as Python, Matlab or R, some background in an analytical field such as physics, engineering, applied math, and some basic experience with statistics. No prior experience with earthquakes, seismology or geophysics is necessary. Guidance and mentorship will be available for data acquisition, literature review, earthquake physics and scientific programming, but the intern should be willing and prepared to work independently.
Duties/Responsibilities:
The intern will work directly with a USGS postdoc as part of the existing team of researchers developing the PLUM Earthquake Early Warning (EEW) algorithm for the United States. The intern will be involved weekly with the team's development calls to get a sense of ongoing work on EEW. The intern should have experience in scientific programming, such as Python, Matlab or R and will work independently to develop the analysis scripts and software, with guidance from the supervising postdoc and PI. While the scope of the project is somewhat specific, there are many independent tasks that the intern will take on, such as assembling seismic data, writing a processing and analysis script, troubleshooting, and presenting results to the local mentor and PI as well as the broader PLUM team.
