Group B, Poster #152, Earthquake Forecasting and Predictability (EFP)

A Decade of Prospective Evaluations of One-Day Seismicity Forecasts for California: First Results

José A. Bayona, Marcus Herrmann, William H. Savran, Fabio Silva, Philip J. Maechling, Warner Marzocchi, & Maximilian J. Werner
Poster Image: 

Poster Presentation

2023 SCEC Annual Meeting, Poster #152, SCEC Contribution #13163 VIEW PDF
Since 2007, the Collaboratory for the Study of Earthquake Predictability (CSEP) has been prospectively evaluating 24-hour seismicity forecasting models for California to address seismological questions with important implications for time-dependent earthquake hazards, such as: How are earthquakes clustered and triggered? Are small earthquakes useful for forecasting larger events? Do seismicity models exhibit better performance with time?. The pool of 24 earthquake forecast models under evaluation includes various flavours of ETAS, STEP, non-parametric models, and ensemble models. Using statistical methods developed by CSEP, we assess the consistency of these models against observed M3.95+ ea...rthquakes and compare their long-term performance with that of a time-invariant smoothed seismicity model. Our prospective dataset contains nearly 600 target events, including the 2010 M6.5 Ferndale, 2010 M7.2 El Mayor-Cucapah, and 2014 M6 South Napa earthquakes. Preliminary test results show that most models tend to overestimate the number of earthquakes, especially in times of seismic quiescence, but consistently forecast the spatial distribution of observed epicenters, suggesting that model productivity parameters may need regular updating and that spatial smoothing parameters are relatively more stable over time. These results may be helpful in improving our ability to forecast earthquake clustering, informing the next generation of time-dependent earthquake forecasting models, and advancing Operational Earthquake Forecasting in California.