SCEC Award Number 24199 View PDF
Proposal Category Community Workshop
Proposal Title CSEP Workshop: New Frontiers in Earthquake Forecasting
Investigator(s)
Name Organization
Maximilian Werner University of Bristol (United Kingdom) José Bayona Deutsches GeoForschungsZentrum Potsdam (Germany) Philip Maechling University of Southern California
SCEC Milestones D2-1, C1,2,3-1, D1-1 SCEC Groups EFP, Seismology, RC
Report Due Date 10/08/2024 Date Report Submitted 04/05/2025
Project Abstract
 The Collaboratory for the Study of Earthquake Predictability (CSEP) develops global cyberinfra-structure for the blind evaluation of earthquake forecasting models and prediction algorithms. CSEP thereby contributes to an objective and independent assessment of the predictive power of scientific hypotheses about earthquake occurrences, some of which underpin operational fore-casts issued by agencies like the USGS and others around the world. This 2024 workshop suc-cessfully brought together CSEP, USGS and SCEC scientists from around the globe to advance earthquake forecasting and testing. The first session focused on ongoing CSEP activities and evaluations in California, China, Italy and New Zealand. A major focus is the evaluation of 24-hour earthquake forecast models in California over the period 2007-2018, which was a major CSEP initiative. The second session looked further into the future, with presentations ranging from machine-learning based forecasts and insights into predictability of low frequency earth-quakes on the central San Andreas fault. Participants also discussed the required computing re-sources for future CSEP activities and software for visualizing USGS operational forecasts.
Intellectual Merit The workshop contributes to SCEC’s goal of advancing the predictability of earthquakes by rigorous testing of models and developing new and improved models. One example of innovation was the presentation of novel ma-chine-learning based forecast methods that were compared to a current benchmark model that underpins most op-erational forecasting systems. Another example of creative work was a study that investigated the predictability of low-frequency earthquakes.
Broader Impacts CSEP actively promotes open research principles amongst the community, including the publication of repro-ducibility packages and open-source community-developed software along with training workshops and online and in-person tutorials. These efforts have broadened the participant base and . CSEP’s forecast model evalua-tions provide independent assessments of the kinds of models that international and US government agencies use for public forecasting, thereby building objective confidence in their predictive skills and exposing limita-tions.
Project Participants Please see the workshop website for a full list of participants and institutions:
https://www.scec.org/events/2024-scec-csep-workshop/
Exemplary Figure Figure 1: Illustration of the CSEP next-day California earthquake forecasting experiment between 2007 and 2018. Top: Earth-quakes during the prospective target period with M3.95+ in the CSEP California testing region. Bottom: Cumulative observed seismicity (black) and model forecasts (in colors). Inset: focus on the 2012-2018 period (taken from Bayona et al., 2024, in preparation).
Linked Publications

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