Application of the Vertical‑to‑Horizontal (VH) Ratio of the Peak Ground Velocity (PGV) in the Bay of Algiers using Data from Weak to Moderate Earthquakes

Abstract

This research aims to identify the most appropriate predictor for the Vertical‑to‑Horizontal (VH) ratio of the Peak Ground Velocity (PGV) from the five existing models in the literature useful for regional or site‑specific probabilistic seismic hazard assessment, and practical applications in the Bay of Algiers. Firstly, dataset of 285 observed VH ratios of PGV was compiled from nine seismic stations within the study area, installed on different flat and irregular surfaces. This dataset was derived from earthquakes with moment magnitude (Mw) ranging from 3.3 to 5.3. Next, the dataset was categorized into two groups based on the soil type at the station locations: rock (S1) (Vs30 above 800 m/s) and stiff (S2) (Vs30 360‑800 m/s) soils. After that, a preliminary linear regression analysis was performed for each group of observed VH ratios of PGV as a function of the Joyner‑Boore distance (RJB) of near field and compared with three selected candidature predictors: Akkar et al. (2014), Bozorgnia and Campbell (2016), and Ramadan et al. (2021) (RA2021). For an extensive evaluation, the Euclidean Distance‑based ranking method (EDR) was applied on the three mentioned candidature predictors. For rocky soil, the results indicate that all models closely align with the linear regression fit, around a VH ratio of PGV of 0.6. However, RA2021 appears to provide a reasonable fit, with a VH ratio of PGV of 0.4 for stiff soil, despite the significant site‑effects at the respective stations. The EDR showed that RA2021 gives the lowest sigma with the observed ratios for S1 and S2 soil classes. For the far field, estimates of the VH ratio of PGV are provided for three strong earthquake magnitudes (6.5, 7.0, and 7.5) and different soil classes, using the existing models.