Seismic Hazard Assessment and Its Uncertainty for the Central Part of Northern Algeria

This study presents a probabilistic seismic hazard assessment for the central part of northern Algeria using two complementary seismic models: a fault-based model and a gridded seismicity model. Two ground-motion attenuation equations were chosen using the Pacific Earthquake Engineering Research Center Next-Generation models, as well as local and regional ones. The ranking method was used to assess their ability to gather accurate data. To account for epistemic uncertainty in both components of the assessment, the seismic hazard was computed using a logic tree approach. Expert judgment and data testing were used to evaluate the weights assigned to individual ground-motion prediction equations. The seismic hazard maps depicted the obtained results in terms of spectral accelerations at oscillation periods of 0.0, 0.2, and 1.0 s, with 10% and 5% probabilities of exceedance in 50 years, and for soil types B, B/C, C, and C/D, as defined by the National Earthquake Hazards Reduction Program. From the analysis, the uncertainty is expressed as both a 95% confidence band and the coefficient of variation (COV). Annual frequencies of exceedance and hazard curves were estimated for the selected cities, as well as uniform hazard spectra for the previously quoted probabilities of exceedance and the soil types considered. Peak ground acceleration values of 0.44±0.17\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$0.44\, \pm 0.17$$\end{document} g and 0.38±0.06\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$0.38\, \pm 0.06$$\end{document} g were reported for the B/C soil type in the cities of Algiers and Blida, respectively, for a return period of 475 years. Seismic maps for the selected return periods depicting the classification of the estimated values are also displayed in terms of very high, high, medium, low and very low degrees of reliability. Furthermore, a seismic hazard disaggregation analysis in terms of magnitude, distance, and azimuth was carried out. The primary goal of such analyses is to determine the relative contribution of different seismic foci and sources to seismic hazard at specific locations. Thus, for each studied city, for the considered return periods and for the soil type B/C, the so-called control or modal earthquake was estimated. At Algiers, events with magnitudes Mw 5.0–5.5 and distances of less than 10 km contribute the most to the mean seismic hazard over a 475-year period. However, for the same return period, those events with Mw 7.0–7.5 and located between 10 and 20 km away contribute the most to the seismic hazard at Tipaza.