Spatial Correlation Analysis Using Iranian Strong Motion Dataset

Ground motions recorded from a specific earthquake at nearby sites are spatially correlated. This correlation is as a result of common earthquake source properties, similar propagation path and local soil conditions. The spatial correlation decreases by the increase in separation distance of sites. The necessity of spatial correlation modeling is exposed especially in seismic risk assessment of infrastructure. Due to the fact that the complicated tectonic setting of Iran has resulted in the numerous destructive earthquakes, Investigation of spatial correlation would be an effective step towards more accurate evaluation of seismic risk in Iran. For this propose, correlation analysis is carried out using Iranian strong motion dataset and functional forms of Gaussian and exponential are fitted on the experimental semivariograms. The obtained results show that Gaussian model can better fit the values at shorter separation distances. The proposed spatial correlation decay evaluated by exponential model is comparable with correlation values presented by previous studies.