Mapping seismostratigraphical amplification effects at regional scale from geological data

Abstract

It is widely recognized that the amplification of ground motion during earthquakes is attributed to the interference of seismic waves trapped between the free surface and impedance contrasts in the shallow subsoil. Seismic Microzonation (SM) studies are devoted to evaluating these site effects, but their application in wider contexts is a hard and expensive task. To estimate seismic site effects at regional scale, the most viable approach is to utilize detailed geological and geomorphological data (1:10.000-1:50.000), which are available for large part of Italy. In the frame of the national research project “SERENA”, in this study a procedure is proposed and tested to constrain entity of 1D seismostratigraphical ground motion amplification based on geological information at the most detailed scale available. In particular, amplification factors are estimated for Seismically Homogeneous Microzones (SHM) defined on the basis geological information. Each SHM is represented as a stack flat homogeneous layers each characterized in terms of engineering-geological units by following the seismic microzonation standards. Seismic properties of each layer (shear waves velocity, density, damping and G/G0 curves) and respective range of variability are determined on the basis of the most recent literature. This information feeds a linear equivalent numerical approach and the Inverse Random Vibration Theory to compute the expected seismic response at each SHM. To account for the relevant uncertainty, 100 random profiles were generated for each SHM, which were compatible with available data. Outcomes of the relevant numerical simulations were considered to assess uncertainty affecting amplification estimates at each SHM. Through this procedure, approximately 4000 SHM were identified, distributed across over 80,000 formation outcrops mapped on Geological map of Tuscany Region, selected by dedicated ArcgisPro TM/Arcpy TM scripts elaborated for this aim. The 50th percentile of the amplification factor distribution for each SHM was taken into consideration. This process aimed to create a new map of amplification factors for the entire territory of Tuscany, achieving an optimal spatial resolution of 1:10,000. To assess the reliability of the results obtained from numerical simulations, and evaluate the possible presence of biases, outcomes of the numerical procedure here considered  were compared with those from second and third levels of Seismic Microzonation studies available in Tuscany. Approximately 1500 benchmark samples were identified, revealing distinct trends among various SHM, particularly between those with outcropping sedimentary covers and those with exposed geological bedrock. In general, amplification estimates provided by the approach here proposed provide a slight overestimate of the ones provided by the detailed seismic microzonation studies (less than 10% on average). However, this overestimate is largely within the range of uncertainty affecting regional estimates and mostly concern SHMs where bedrock outcrops. It is worth to note that by no way the proposed approach should be considered as substitute of detailed local studies. Anyway it could be considered to provide ex-ante evaluations to be used as a preliminary reference for large scale risk analysis and for a preliminary assessment of expected ground motion effects where more detailed studies are not available so far.