← Back to Sessions

Sensitivity Analysis of Seismic Hazard Parameters for the Understanding of Its Uncertainties: A Study Case for Central America

Description: 

The parameters involved in seismic hazard calculations introduce both epistemic and random uncertainties that must be controlled to minimize errors and obtain an accurate seismic model representing expected accelerations in a region. In this regard, it is essential to apply a sensitivity analysis of the results provided by seismic hazard calculations to understand the behavior of uncertainties associated with the parameters. Statistical methods (e.g., RMSE, MSE, and CV) are employed for such analysis, relating observed variables to those predicted by the models used, enabling both analytical and spatial examination of interacting parameters in seismic hazard.

Consequently, we conducted a case study for Central America, analyzing the sensitivity of parameters involved in seismic hazard calculations. In this case, we examined the fitting method for the recurrence model (least squares and maximum likelihood), the use of scaling laws, and the application of ground motion models (GMM). Ultimately, we obtained coefficient of variation (CV) maps illustrating the variability of results and highlighting regions in Central America where these trends are evident. This variability and sensitivity analysis helped refine the weights and parameters of the logical tree used to generate new seismic hazard maps for Central America for PGA and Sa(1.0s) at return periods of 475, 975, and 2475 years. Additionally, we compared our results with seismic hazard maps created by other authors in the region to assess the significance of acceleration variability.

Session: Understanding and Quantifying the Variability in Earthquake Source Parameter Measurements [Poster Session]

Type: Poster

Date: 5/3/2024

Presentation Time: 08:00 AM (local time)

Presenting Author: Carlos

Student Presenter: Yes

Invited Presentation: 

Authors