Work

Estimating VS30 structure from topographic and geological data in the southeast of Spain

Year: 2025

Type: article

Abstract: The V S30 was introduced as a representative parameter of the site's seismic amplification and is currently considered to prescribe site-dependent design response spectra. Although the determination of V S30 by geophysical methods can be time-consuming, using empirical relationships between this quantity and the slope of the terrain is a rapid and very low-cost procedure that may be sufficient for some applications. A specific correlation at regional scale is performed using 3216 measurements of V S30 obtained from SPAC, MASW and Mini-array seismic methods, located within 8 municipalities belonging to the Spanish provinces of Almería, Granada and Murcia. The measurement points have been grouped into 10 classes with different seismic behavior according to the geological units represented on the Geological Map of Spain (MAGNA_50). Two relationships have been fitted: a) Model 1 consisting of a linear dependence between the logarithm of Vs30 and the ground slope; b) a linear relationship between log(V S30 ) and the slope including the dependence on the geological units through additive constants (model M2). The best-fitted regional V s30 model is built from a DEM with a resolution of 100 m, in which the incorporation of geological units has demonstrated its convenience in reducing residual errors by up to 28 %. The mean absolute residual obtained at the V S30 sampling sites was 80 m/s for model M1 and 64 m/s for model M2. The mean estimated V S30 values obtained from model M2 range between 360 ± 130 m/s in the marshes and wetlands formation and 750 ± 69 m/s in the geological unit composed of quartzites, schists and phyllites. • A specific relationship V S30 vs. topographic gradient is fitted for SE Spain. • Including dependence on the geological units significantly improved methods' accuracy. • Using 100 m resolution in the elevation model provided the best correlation. • V S30 residuals reduced by up to 37 % compared to Wald and Allen (2007) equations.

Source: Engineering Geology

DOI: 10.1016/j.enggeo.2025.108037