Insights into earthquake hazard map performance from shaking history simulations

Area of Science:

• Earthquake seismology
• Geophysics
• Hazard assessment

Background:

• Recent large earthquakes have exhibited ground shaking exceeding predictions from standard earthquake hazard maps for typical return periods.
• Discrepancies raise questions about the accuracy of current probabilistic seismic hazard analysis (PSHA) methodologies and map parameters.

Purpose of the Study:

• To investigate the discrepancy between predicted and observed earthquake shaking.
• To assess the verification (map implementation) and validation (forecasting accuracy) of seismic hazard maps.

Main Methods:

• Simulated an area's seismic history to generate shaking distributions.
• Compared simulated "observed" shaking with predictions from hazard maps created using identical parameters.
• Evaluated map performance by assessing internal consistency and forecasting capability.

Main Results:

• Simulated shaking distributions showed consistency with the hazard map's mean predictions.
• Individual simulated earthquake histories exhibited significant scatter, with infrequent large events causing shaking stronger than mapped.
• Probabilistic seismic hazard analysis (PSHA) demonstrated internal consistency, suggesting maps are "built right" in terms of methodology.

Conclusions:

• The study suggests that probabilistic seismic hazard analysis (PSHA) is internally consistent and verified.
• Validating seismic hazard maps against actual earthquake histories is complex due to inherent variability and the difficulty in distinguishing chance from parameter bias.