The 2017 UseIT Internship: Navigating "Wicked Problems"

Image of the 2017 UseIT Class, provided by USC News

The southern segment of the San Andreas fault has not generated a large earthquake since the 1857 M7.9 Fort Tejon earthquake, a 360 km (225 miles) rupture of the fault from Parkfield down to Wrightwood. The "recurrence interval" of a large earthquake along this segment is 150 years, and now here we are: 160 years later. The next “big one” could shake the region at any time. However, recurrence intervals simply refer to the average amount of time between events, they don’t determine exactly when an event will occur. With no way of knowing if it will hit this week or 20 years from now, fear hardly seems like a reasonable response to this ever-present threat.

Scientists have been working toward a better understanding of the earthquake problem: the Southern California Earthquake Center's (SCEC) Undergraduate Studies in Earthquake Information Technology (USEIT) interns have spent the summer investigating how certain earthquake scenarios could affect the probability of the next large earthquake occurring in Southern California, a “Grand Challenge” proposed by SCEC director Thomas H. Jordan. The program, which began in 2004, brings students from all different backgrounds and fields of study together to work on better understanding the seismic activity of Southern California.

The 2017 UseIT Grand Challenge

1. Develop a computational system for the probabilistic forecasting of earthquake sequences in Southern California using long seismicity catalogs generated on the Blue Waters supercomputer by the RSQSim rupture simulator.

2. Apply the system to three initial-event scenarios: M6.1 Parkfield, M7.0 Mojave, M6.0 Bombay Beach. Compare the simulator-based probabilities for large aftershocks (M ≥ 7) with the values given by the official Uniform California Earthquake Rupture Forecast, Version 3.

3. Select multi-event scenarios that could threaten the Los Angeles region, and illustrate their hazard and risk with sequence-specific maps of expected ground motions, economic losses, and human casualties.

Animation of a 35,000-year catalog of simulated earthquakes in California, with a magnitude range between 5.0 and 6.0.. The catalog was developed by SCEC UseIT interns with RSQSim and visualized with SCEC-VDO.

"My first impression of SCEC and of the USEIT internship was definitely kind of intimidating. I don’t come from a background in earthquake science so I was concerned that I was just going to be completely underprepared.” said intern and computer engineering major Olivia Dorencz (Boston University).  

All the interns overcame hurdles like these by spending the first few weeks of the program learning about earthquakes and the various technologies scientists use to study them. This culminated in a two-day field trip to faults around the Los Angeles area and beyond, led by USC Earth Sciences Professor James Dolan and geologist Scott Lindvall of Lettis Consultants International, Inc. 

With their newfound knowledge, the USEIT interns broke ground on their efforts to solve this year’s "Grand Challenge" by dividing into four different teams based on their interests and experience:

The High-Performance Computing (HPC) Team takes the first step in these efforts. Responsible for generating hundreds of thousands of years of simulated earthquake catalogs using the NCSA's Blue Waters supercomputer, they have created the raw data that makes everyone else’s work possible. “High-performance computing can be done to generate data for all sorts of different groups and causes, and it just opens a lot of different doors.” remarked HPC team member Sarah Troise (Washington and Lee University), acknowledging the value of supercomputing beyond just the scope of earthquake science.


These simulated earthquake catalogs contain a dense amount of information, and it has been the job of the Probability Team to sort through this data with scripts written in Java to extract information about earthquake event sequences and their probabilities. By looking at initial scenarios near Parkfield, Mojave, and Bombay Beach, they researched if and how moderate earthquakes could increase the probability of large earthquakes happening in the weeks following these initial events.

This type of earthquake forecasting continues to be a developing field in which USEIT interns have been integral. "Knowing a range of possibilities in California's seismic future allows earthquake scientists to prioritize their research and further develop tools useful to governments, engineers, and  other stakeholders." -Amelia Midgley (University of Massachusetts, Amherst).

The Hazard and Risk Visualization Team (HaRVi) Team used the analyses of the Probability team to generate ShakeMaps and HAZUS reports displaying the impact that large earthquake sequences could have on Los Angeles and other urban areas. They compiled their results into an interactive ESRI StoryMap created using ArcGIS which describes in detail the initial event scenarios they were investigating.

The HaRVi team took the first steps in making the work of the USEIT interns public, a job that was continued by the Communications Work Group. Consisting of members from every other team, the Communications Work Group has developed social media content, videos, and other means of helping the SCEC community and the public in general understand the work of the USEIT interns. 

“We serve as a bridge from actual research that’s going on to the public, and research like this can help people better prepare what’s to come so that these earthquakes will not be as harmful as they can be,” said Communications and HaRVi team member Resherle Verna (University of Southern California).

Though using different means and tools to accomplish different goals, each team consistently found themselves using SCEC-VDO, a visualization program for displaying geographic and geodetic data like fault systems, earthquakes, and political boundaries. SCEC-VDO is a unique program as it has been developed entirely by previous USEIT interns, a job that this year’s SCEC-VDO Development Team continued to tackle. They have spent the past weeks sifting through Java code, fixing bugs, developing new functionalities, and ultimately creating the first major release of SCEC-VDO.

Accompanied with a video tutorial and manual, the program can now easily be used by anyone, making it easier for scientists and members of the SCEC community to use in their research.

The USEIT interns have not only learned about the SCEC community this summer, but have become an important part of it. They will be attending the SCEC Annual Meeting this September and will get the opportunity to present their research among hundreds of earthquake scientists. For many of the interns this will be their first time presenting research, though as potential future leaders of the earthquake science community, it will likely not be their last.

Further Resources:

USEIT is supported by the NSF Research Experience for Undergraduates Site program (REU), the USC Department of Earth Sciences, the USC Undergraduate Research Program, SCEC, and other funders which vary from year to year. Significant funding for USEIT has been provided by Mr. George Joseph the founder and Chairperson of the Board, Mercury General Corporation.