The assessment of the Local Seismic Response (LSR) to the expected seismic motion in a territory is the subject of seismic microzonation studies. For the estimation of the seismic action, it is necessary to define some fundamental elements: the seismic input representative of the seismic action of the seismic bedrock and an adequate geotechnical model.

Generally, at least seven accelerograms compatible with the site's target response spectrum have to be identify. Typically, such accelerograms are chosen from available databases, in order to provide the best approximation of the target spectrum, based on an appropriate range of magnitude, epicentral distance, and epsilon parameter (SMEE), i.e. the number of standard deviations by which the observed logarithmic spectral acceleration differs from the mean logarithmic spectral acceleration of a ground-motion prediction by the attenuation equation.

These choice is made irrespective to all the time-dependent features characterizing the real signals of the area subject to microzonation (e. g. duration, evolution of amplitude and energy frequency content, total energy and its distribution over time, strong motion duration, etc.).

In a recent paper (Colajanni, Pagnotta, Testa, 2020) four methods for generating fully non-stationary artificial accelerograms on the basis of a target spectrum and a set of accelerograms registered in the neighborhood of the construction site, with provide a mean response spectrum that can be different from the target one. Among these, two of them, namely (Cacciola, 2010 (CA), and Rofooei, Mobarake & Ahmadi, 2001(RMA)) have proven to be particularly efficient in reproducing the characteristics of the events expected at the site. More precisely, the CA method provides reliable results, but they are highly influenced by the selection criterion of the group of real records through which the non-stationary counterpart is defined, the RMA is able to clearly define the non-stationary characteristics of a site, but gives controversial results, due to the absence of a spectrum-compatible formulation.

In this paper, the set of accelerograms obtained with the two generating methods are used as input accelerograms for a microzonation procedure, obtained from one-dimensional linear equivalent method site response analyses implemented in AlgoShake2D software (Algoritmiqa 2024), and compared with results obtained by microzonation performed through the natural accelerogram selection procedure currently used in the professional field mentioned above (SMEE) , and with the results obtained assuming as a signal to the bedrock a set of accelerograms actually recorded at the site, which constitute the benchmark in analogy to the procedure used in (Colajanni Pagnotta, Testa, 2020).