مطالعه‌ي تاثير لحاظ عدم قطعيت هاي عمق رخدادها در نتايج تحليل خطر لرزه ‌يي احتمالاتي

Influence of Depth Uncertainty Treatment on Results of Probabilistic Seismic Hazard Analysis

در تحليل خطر لرزه‌يي احتمالاتي مرسوم، از لحاظ اثر عدم قطعيت عمق رخدادها چشم‌پوشي شده و يك توزيع يكنواخت براي آن منظور مي‌شود؛ در حالي كه عمق رخدادها در طبيعت از يك توزيع يكنواخت تبعيت نمي‌كند و حتماً متغيري تصادفي با پراكندگي قابل ملاحظه است. در اين نوشتار، با آگاهي از محدوديت‌هاي روش كلاسيك و به‌منظور ارزيابي اثر لحاظ اين عدم قطعيت در نتايج نهايي تحليل خطر، از روش شبيه‌سازي مونت‌كارلو در مطالعه‌ي موردي تحليل خطر شهرستان ازنا استفاده شده است. ابتدا نتايج حاصل از روش مونت‌كارلو در مقابل روش كلاسيك صحت‌سنجي شده و در ادامه، با اعمال فرض توزيع غيريكنواخت حاصل از تحليل داده‌هاي كاتالوگ، فرايند تحليل خطر تكميل شده است. نتايج نشان داده است كه درنظرگرفتن توزيع واقع‌گرايانه‌ي عمق مي‌تواند تاثير قابل ملاحظه‌يي در نتايج نهايي تحليل خطر در خلاف جهت اطمينان داشته باشد.

In the conventional framework of probabilistic seismic hazard analysis, depth uncertainties are not recognized and a uniform distribution is assumed for the depth of the earthquake. However, the non-uniform distribution of depth and its remarkable deviation is obvious. The main objective of this paper is to evaluate the influence of the treatment of depth uncertainties on the entire results of seismic hazard analysis. Due to the limitations of the classical approach of hazard analysis, the Monte Carlo method was applied to capture the depth uncertainties. In this method, the random variables can be entered as a probabilistic density function with observed means and standard deviation, or in tabular formats. In this paper, seismic hazard assessment of Azna City, Iran, was undertaken using the Monte Carlo approach. First, the results of the Monte Carlo approach were validated against the classic approach, and, then, a nonuniform depth distribution, based on the observed depth of the earthquake, was assumed in a tabular format; the process of hazard assessment was then completed. Finally, the results, including hazard curves, in different periods, were obtained. In conclusion, the pseudo spectral accelerations derived from nonuniform depth distribution were higher in comparison with uniform distribution, equal to the observed mean of the depth of the earthquake. The differences between pseudo spectral accelerations in two different depth distributions were increased at higher hazard levels. The results of comparison showed that by considering the observed mean minus half of the standard deviation as uniform depth distribution, would have relatively equal results to those of realistic depth distribution. In this study, using the observed mean minus half of the standard deviation instead of the observed mean as a uniform depth distribution was proposed for considering depth uncertainties in a conventional framework of probabilistic seismic hazard analysis.