Regionally adjusted stochastic earthquake ground motion models, associated variabilities and epistemic uncertainties
Jaleena Sunny,Marco de Angelis,Benjamin Edwards
DOI: https://doi.org/10.1007/s10950-024-10195-7
2024-02-10
Journal of Seismology
Abstract:An optimisation-based calibration technique, using the area metric, is applied to determine the input parameters of a stochastic earthquake-waveform simulation method. The calibration algorithm updates a model prior, specifically an estimate of a region's seismological (source, path and site) parameters, typically developed using waveform data, or existing models, from a wide range of sources. In the absence of calibration, this can result in overestimates of a target region's ground motion variability, and in some cases, introduce biases. The proposed method simultaneously attains optimum estimates of median, range and distribution (uncertainty) of these seismological parameters, and resultant ground motions, for a specific target region, through calibration of physically constrained parametric models to local ground motion data. We apply the method to Italy, a region of moderate seismicity, using response spectra recorded in the European Strong Motion (ESM) dataset. As a prior, we utilise independent seismological models developed using strong motion data across a wider European context. The calibration obtains values of each seismological parameter considered (such as, but not limited to, quality factor, geometrical spreading and stress drop), to develop a suite of optimal parameters for locally adjusted stochastic ground motion simulation. We consider both the epistemic and aleatory variability associated with the calibration process. We were able to reduce the area metric (misfit) value by up to 88% for the simulations using updated parameters, compared to the initial prior. This framework for the calibration and updating of seismological models can help achieve robust and transparent regionally adjusted estimates of ground motion, and to consider epistemic uncertainty through correlated parameters.
geochemistry & geophysics