GNGTS 2018 - 37° Convegno Nazionale

GNGTS 2018 S essione 2.1 357 Residual analysis. Bindi et al. (2018) performed a consistency check over the flatfile by analyzing different residual distributions obtained from ad-hoc ground motion prediction equations of the absolute spectral acceleration (SA), displacement and Fourier amplitude spectra (FAS) for shallow crustal earthquakes (focal depth < 40km). They identified those earthquakes, stations, and recordings showing the largest deviations from the GMPE median predictions, and also evaluated the statistical uncertainty on the median model to get insights on the applicable magnitude–distance ranges and the usable period (or frequency) range. Robust median predictions are obtained up to 8 s for SA and up to 20 Hz for FAS, although median predictions for Mw ≥ 7 show significantly larger uncertainties with ‘bumps’ starting Fig. 1 - a) Number of recordings in the time interval 1975–2016 for different distance ranges. In the box, histogram of the number of records per event; b) Magnitude vs distance (R JB if available otherwise R EPI ) scatter plot of recordings in the flatfile grouped by style of faulting. U: undefined; SS: strike-slip; TF: thrust; NF: normal. The records at zero distance (R JB = 0) are plotted at 0.1 km. Figures are taken from Lanzano et al. (2018a). Fig. 2 - a) Period dependence of the asymptotic standard error on the median (σ μ ) for SA considering different magnitudes and a fixed distance of 30 km; b) standard deviation of different residual distributions for SA. White circle: standard deviation of the between-station distribution; gray square: standard deviation of the between-event distribution; white triangle: standard deviation of the between-country distribution; black diamond: standard deviation of the residual distribution. Figures are taken from Bindi et al. (2018).