RISK ANALYSIS OF EARTHQUAKE ON CONSTRUCTION PROJECTS FOR DIFFERENT REGIONS OF IRAN, USING SECOND ORDER AND GLS LINEAR REGRESSION, STATISTICAL TESTS AND POISSON STATISTICAL MODEL

Document Type : Article

Authors

1 University of South Australia

2 Faculty of Engineering Tehran University

3 Dept. of Civil Engineering Sharif University of technology

Abstract

In this paper, for every construction project in one of the twenty seismic zones of Iran, whether it is in the design or construction phase or even as a finished project, a number is obtained representing the project risk coeficient or the probability of a specific earthquake in the utilization interval of a specific project. To find this number, for every abovementioned zone, the Gutenberg-Richter" linear relation is used to demonstrate the relationship between earthquake magnitude and the number of earthquakes greater than, or equal to, a given magnitude. To check the existence of Heteroscedasticity and autocorrelation - which are two ignored terms in classic linear regression - for recorded data, some tests are applied. If each of the mentioned terms is found, treating techniques relating to that term, which is named GLS regression, will be applied to calculate the best regression line for the
Gutenberg-Richter" relationship. The second - degree relationship of the earthquake magnitude and the number of the earthquake is also assessed. Using Poisson statistical distribution and having the project design earthquake and project utilization interval
as two concepts of earthquake magnitude and earthquake period, the third factor, which is design earthquake risk, will be obtained for a specfic project in a specific location and for the indicated project utilization interval.

Keywords