Probabilistic seismic hazard assessment for Saudi Arabia using spatially smoothed seismicity and analysis of hazard uncertainty

We analyze the results of a Probabilistic Seismic Hazard Assessment obtained for Saudi Arabia using a spatially smoothed seismicity model. The composite up-to-date earthquake catalog is used to model seismicity and to determine earthquake recurrence characteristics. Different techniques that are frequently used for the analysis of input data are applied in the study. The alternative techniques include the declustering procedures for catalog processing, statistical techniques for estimation the magnitude–frequency relationship (MFR), and the seismicity smoothing parameter. The scheme represents epistemic uncertainty that results from an incomplete knowledge of earthquake process and application of alternative mathematical models for a description of the process. Our calculations that are based on the Monte Carlo technique include also a consideration of the aleatory uncertainty related to the dimensions and depths of earthquake sources, parameters of MFR, and scatter of ground-motion parameter. The hazard maps show that the rock-site peak ground acceleration (PGA) is the highest for the seismically active areas in the north-western (Gulf of Aqaba, PGA > 300 cm/s2 for return period 2475 years) and south-western (PGA > 100 cm/s2 for return period 2475 years) parts of the country. We show that the procedures for catalog declustering have the highest influence on the results of hazard estimations, especially for high levels of hazard. The choice of smoothing parameter is a very important decision that requires proper caution. The relative uncertainty that is ratio between the 85th and 15th percentiles may reach 50–60% for the areas located near the zones of high-level seismic activity.