Statistical characterization of the seismic action in terms of PGA and PGV: Comparison of two methods of calculation

In a country like Italy, directly affected by seismic phenomena, the assessment of vulnerability and seismic hazard throughout the territory plays a very important role, in order to minimize seismic effects on the population. In this paper two methods for determining the hazard of the Italian territory, i.e. the probability functions (CDF: cumulative distributive function and PDF: probability density function) of both PGA (Peak Ground Acceleration) and PGV (Peak Ground Velocity) at a specific site, over a given observation time, are analytically developed. Both methods are developed according to Cornell's widely upheld method (1968) and are specifically updated and adapted to the statistical and geophysical situation of the Italian territory. The first procedure is characterized by the treatment of the distance R from the epicenter to the site as a discrete variable, while the other treats R as a continuous random variable. The paper also presents illustrative examples of application of the two methodologies for the determination of the probability functions of PGA and PGV for the cities of Florence and Sturno (site of the 1980 Irpinia Earthquake). Analysis of the results obtained indicates that the approach that treats R as a continuous random variable provides the most reliable results.