SCEC Award Number 25328 View PDF
Proposal Category Individual Research Project (Single Investigator / Institution)
Proposal Title Analysis and Possible Implications of Recent California Earthquakes using Earthquake Nowcasting and QuakeGPT
Investigator(s)
Name Organization
John Rundle University of California, Davis
SCEC Milestones C1-1, D2-1, D2-2, D3-1, D3-2 SCEC Groups RC, EFP, ASI
Report Due Date 03/15/2026 Date Report Submitted 03/25/2026
Project Abstract
 It is of critical importance to understand whether these events are foreshocks that indicate that much larger magnitude earthquake is imminent. To analyze these events, we will apply our recently developed expertise in earthquake nowcasting. Earthquake nowcasting is a relatively new method that employs a simple 2-parameter filter on the observed monthly seismic rate of small earthquakes. The filter uses machine learning to improve the nowcast filter. The methods are being actively developed in a variety of ongoing research projects. Rundle et al. also discusses a new model "QuakeGPT" that uses stochastic earthquake simulations to train a science transformer model, an attention-based technology similar to that which underlies ChatGPT and similar Large Language Models.

During the coming year, we propose to closely monitor the incoming data with these newly developed methods and continue to evaluate the results for their importance and significance.

Summary of Previous Research Supported in Part by SCEC Funds:
Machine Learning (ML) Nowcasting: Receiver Operating Characteristic
Nowcast Curves
Intellectual Merit We extend the natural time forecast to calendar time forecasts using an ensemble approach. The only assumption in our method is that the statistics of the local region are the same as in the larger surrounding regions. The method has significant skill, as defined by the Receiver Operating Characteristic (ROC) test, which improves as time since the last major earthquake increases. We apply the method to the same local region as in our first paper around Los Angeles, California, following the January 17, 1994 magnitude M6.7 Northridge earthquake.
Broader Impacts SCEC was founded by Kei Aki and collaborators with the explicity goal of developing a "Master Model" to forecast and predict earthquakes in California and elsewhere. To date, that goal has not been achieved,in fact, one could argue that the problem is so difficult that it has essentially been abandoned in favor of more prosaic results. In this paper, we return to this basic goal develop the means to achieve Aki's original goal.
Project Participants John B Rundle1,2,3, Ian Baughman1, Andrea Donnellan4,3, Lisa Grant5, Geoffrey Fox6, Kazuyoshi Nanjo7,
1 University of California, Davis, CA
2 Santa Fe Institute, Santa Fe, NM
3 Jet Propulsion Laboratory, Pasadena, CA
4 Purdue University, West Lafayette, IN
5 University of California, Irvine, CA
6 University of Virginia, Charlottesville, VA
7Shizuoka University, Japan
Exemplary Figure Figure 4. Final slide of a movie that can be produced with the python code, showing the current state of the target earthquake potential in the circle with nowcast curve (left), progress through the earthquake cycle thermometer (left center), calendar time forecast (right center) and map of small earthquakes (right)
Linked Publications

Add missing publication or edit citation shown. Enter the SCEC project ID to link publication.