An extended generalized conditional intensity measure method for aftershock ground motion selection

Abstract

AbstractThis study proposes an extended generalized conditional intensity measure (EGCIM)‐based approach to selecting aftershock ground motions that match the target aftershock EGCIM distributions. The main purposes of the proposed methodology are threefold: (a) to consider the multiple characteristics of aftershock ground motions (e.g., amplitude, frequency contents, cumulative effects, and duration), (b) to account for the correlations between mainshocks (MS) and aftershocks (AS) as well as cross‐correlations between AS IMs, and (c) to associate with the conventional mainshock GCIM approach to enforce mainshock‐consistency for practical applications. For these purposes, three components of the EGCIM methodology for aftershock ground motion selection are utilized: (1) the multivariate aftershock EGCIM distributions of any vector of aftershock intensity measures (IMs) are constructed conditioned on two MS‐AS IMs, (2) the correlations between mainshocks and aftershocks as well as the cross‐correlations between AS IMs are considered, and (3) the random realizations of aftershock IMs are generated as a target using the aftershock EGCIM distributions to select aftershock ground motions from the specific database. The proposed methodology is applied to a case study in the LPCC station in New Zealand, whose aftershock EGCIM distributions of the considered 18 IMs are constructed for the case scenario. Among them, the correlations between mainshocks (MS) and aftershocks (AS) as well as the cross‐correlations between AS IMs are determined using the total residuals (epsilons) based on 662 real MS–AS ground‐motion pairs. Several examples are then discussed to illustrate the application of the proposed aftershock EGCIM approach, including the comparison of the aftershock acceleration spectrum with the ASK14 model and the effects of different components of the aftershock IM weight vector. Then, the proposed approach is compared with the conventional aftershock synthesis methods. It is observed that the conventional methods perform biased representations of different aftershock characteristics, of which the limitations are improved in the proposed approach. Moreover, a sensitivity study is performed for a typical midrise structure to investigate the effects of aftershock ground motion selection on the aftershock fragility analysis. The results indicate that the selection of aftershock records considering different AS characteristics has a notable impact on the aftershock fragility assessment. The uncertainties associated with record‐to‐record variations in aftershock fragility analysis can be effectively reduced by using the proposed selection approach that incorporates the comprehensive ground motion characteristics. The findings in this study promote the extension of the existing methods for aftershock ground motion selection, which can be applied to structural design or seismic risk analysis considering the aftershock effects.