Ground-motion intensity measure correlations observed in Italian data

Abstract

Ground-motion models (GMMs) are used in probabilistic seismic hazard analysis (PSHA) to estimate the probability distribution of earthquake-induced ground-motion intensity measures (IMs). Accounting for spatial correlation and cross-IM correlation in ground-motion data has important implications on seismic hazard and risk assessment outputs. The current practice estimates the spatial correlation separately from the GMM estimation process, which may result in inconsistent and inefficient estimators of parameters in the spatial correlation models and GMMs. Moreover, several correlation models between different IMs have been calibrated and validated based on the NGA-West and NGA-West2 databases and advanced GMMs. However, modeling the correlation between different IM types has not been adequately addressed by current, state-of-the-art GMMs for Italy. To address those issues, this study first develops a series of new Italian GMMs with spatial correlation for 31 amplitude-related IMs, including peak ground acceleration (PGA) and peak ground velocity (PGV) and 5% damped elastic pseudo-spectral accelerations (PSA) at 29 periods ranging from 0.01 s to 4 s. The model estimation is performed through a recently-developed one-stage non-linear regression algorithm proposed by the authors, known as the Scoring estimation approach. Based on the newly-developed GMMs, this study finally proposes a set of analytical correlation models between the selected IMs for the considered Italian dataset.