Spatial Correlation of Peak Ground Motions (PGA and PGV) and Spectral Accelerations Based on the Iranian Earthquake Data

Seismic risk assessment in a regional scale needs to estimate the ground motion intensity measures (GMIMs) at different sites together for a given earthquake scenario. It is shown that GMIMs and structural responses are correlated in nearby sites. So, spatial correlation models have a fundamental role in risk estimation, especially, when lifeline systems or distributed structures are of concern. The trend of these correlations depends on some variable such as magnitude, the local geology and inter-site distance and the spectral period. Considering regional dependency of earthquake source, path and site effects, it is important to develop empirical region-specific correlation models. In this study, the spatial correlation model is proposed for different intensity measures including peak ground acceleration (PGA), peak ground velocity (PGV) and spectral acceleration (SA) based on earthquake records of the Iranian plateau. The intra-event residuals are computed using local ground motion models proposed by Zafarani and Soghrat (J Earthq Eng 28(4):922-945, 2024). A correlation model is developed using semivariogram analysis of strong motion data. As a framework for correlation modeling, a kind of exponential model is employed. The results show that the range generally has an increasing trend as a function of period that seems to be in agreement with other studies on this subject.