← Back to Gallery

Initial Calculations of Aleatory Variability and Epistemic Uncertainty for Physics-based PSHA Using California Simulations

For physics-based PSHA to move from research to practice, aleatory variability and epistemic uncertainty need to be properly calculated and incorporated into the seismic hazard estimation. We present initial uncertainty calculations for 1D simulations for California using the Liou and Abrahamson (2025) uncertainty framework. We separate the validation sigma, which is the standard deviation of the difference between the simulation prediction and observations, into parametric epistemic uncertainty by calculating the standard deviation of the simulated ground motion from varying the velocity model and rupture generator, and the remainder of the validation sigma is method aleatory variability, which is variability due to simplifications of the simulation method.

Method epistemic uncertainty of the median is the scientific uncertainty in the simulation methods. This is the standard deviation of the medians from different 1D simulation methods. For a single method, the epistemic uncertainty is estimated by the standard error of the mean for between-event and within-event terms. There is also method epistemic uncertainty of the aleatory variability, which is scientific uncertainty in the size of the method aleatory variability from limited cases validation and uncertainty in the distribution of the ground motion. We also quantify parametric aleatory variability, which is variability of ground motion due to variable inputs to the simulation method that are not included in the standard source characterization used in the hazard calculation, representing inner-fault variability for future events. This is quantified by the standard deviation of the resulting ground motion after sampling inner fault parameters, which are not known for future events. Lastly, we quantify epistemic uncertainty in the aleatory variability, which is scientific uncertainty in the standard deviation due to alternative rupture generators, velocity models, and Q models. Together, these uncertainties capture the required aleatory variability and epistemic uncertainty for physics-based PSHA.

Session: Constraining GMMs via Physics-Based Simulations and Complementary Observations: Integration and Practice [Poster]

Type: Poster

Room: Exhibit Hall A+B

Date: 4/17/2026

Presentation Time: 08:00 AM (local time)

Presenting Author: Irene Liou

Student Presenter: No

Invited Presentation: 

Poster Number: 34

Additional Authors