EVALUATING OF THE EARTHQUAKE HAZARD PARAMETERS WITHBAYESIAN METHOD FOR THE DIFFERENT SEISMIC SOURCEREGIONS OF THE NORTHERN IRAN

Bayesian method is straightforward procedure of estimating of hazard parameters. We present the results of Bayesianstatistics technique for northern Iran to estimate the earthquake hazard parameters (maximum regional magnitude Mmax,b value of GR relationship and seismic activity rate or intensity λ) and quantiles of its probabilistic distribution on somefuture time interval. Therefore, we using a reliable earthquake catalogue which is homogeneous for 3.1 ≤ mb ≤ 6.2 andduring the period before 1900 to August 2018 is compiled. The northern Iran has been further divided into six seismiczones. In this study, quantiles of functions of distributions of true and apparent magnitudes for future time intervals of 50,100 and 475 years are calculated with confidence limits for probability levels of 50%, 70% and 90 % in all seismogenicsource zones. The maximum magnitude is for the regions 1 and 6 with the range of greater than 7.6 and the lowest for theregions 2 and 5. Note that in this method, we used a priori uncertainty domain of values of parameters. This approachallows the uncertainty of earthquake magnitude to be accounted for seismic hazard parameters like the β-value which isthe slope of the magnitude-frequency law (where, b = loge β) and the intensity (rate) λ of seismic activity and theiruncertainties are also estimated. Indeed to calculate the seismic hazard, we are looking for the possibility of a majorearthquake in a long time period (Lyubushin and Parvez, 2010). We used a statistical approach. The advantage ofstatistical approach is based on its generality. It does not need to identify values of different parameters, which we mustknow if we try to solve differential equations, the most of which could not be defined with sufficient accuracy. Seismichazard analysis in different regions of Iran has been carried out in different studies, for example: Khodaverdian et al.(2016) using an up-to-date earthquake catalog for all grid points estimated seismicity parameters. Among statisticalmethods, Bayesian approach has a special interest that comes from its ability to take into consideration uncertainty ofparameters in fitted probabilistic laws and a priory given information (Campbell, 1983). The advantages of this approachare given by the ability to consider the uncertainty of the parameters in probabilistic calculations and initial information(Lyubushin, 2010). Since there is no need for the mid-steps to estimate the maximum ground seismic acceleration andother parameters, like the maximum likelihood magnitude estimation occurring in a particular zone or source has made itsimpler (Lamarre et al., 1992). Jafari (2007) using Bayesian probability approach calculated the probability of occurrenceof earthquakes with more than a special magnitude in different time periods for the Alborz region. Jafari (2010) using theprobabilistic analysis of the prediction of the next earthquake, was carried out maximizing the conditional probabilitydensity of earthquake occurrence in Tehran. The maximum likelihood magnitude of a certain source or region and itsassociated uncertainty, the Gothenburg-Richter slope, and the seismic activity rate (intensity) are the most probableearthquake hazard parameters used in the estimation of probable seismic hazard which be estimated for the study area.After the division of northern Iran using seismic information, geological, focal mechanism and fault trend into sixseismic sources, the seismic parameters were calculated using the Bayesian static method. The estimated values changingbetween 7.21 and 7.82 are listed in Table 1 and mapped in Figure 1 by dividing into third groups (7.21 ≤ Mmax < 7.40;7.40 ≤ Mmax < 7.60; Mmax ≥ 7.70).

Keywords:

Bayesian method , Earthquake hazard parameters , Maximum regional magnitude , Northern Iran