Deriving vulnerability curves using Italian earthquake damage data

Abstract

The concerted effort to collect earthquake damage data in Italy over the past 30 years has led to the development of an extensive database from which vulnerability predictions for the Italian building stock can be derived. A methodology to derive empirical vulnerability curves with the aforementioned data is presented herein and the resulting curves have been directly compared with mechanics-based vulnerability curves. However, it has been found that a valid comparison between the empirical and analytical vulnerability curves is not possible mainly due to a number of shortcomings in the database of surveyed buildings. A detailed discussion of the difficulties in deriving vulnerability curves from the current observed damage database is thus also presented.

A comparison of seismic risk maps for Italy

“…Many of the shortcomings of empirical vulnerability curves are related to the fact that the vibration characteristics of the buildings are not taken into account (for example, due to the use of peak ground acceleration or velocity) or that the macroseismic intensity is used to define the ground shaking, but this parameter is directly obtained from observed damage data and thus the damage and ground shaking intensity are not independent. A recent study was carried out by Colombi et al (2008) to remove these shortcomings by dividing the damaged buildings into classes as a function of the number of storeys and estimating the displacement demand from a ground-motion prediction equation using an estimated mean period of vibration for each building type. Nevertheless, further disadvantages with the use of the observed damage database of Italian buildings for the generation of empirical curves were identified by Colombi et al (2008).…”

“…A recent study was carried out by Colombi et al (2008) to remove these shortcomings by dividing the damaged buildings into classes as a function of the number of storeys and estimating the displacement demand from a ground-motion prediction equation using an estimated mean period of vibration for each building type. Nevertheless, further disadvantages with the use of the observed damage database of Italian buildings for the generation of empirical curves were identified by Colombi et al (2008). In particular, problems related to the inadequate compilation of the post-earthquake assessment forms meant that much of the data had to be removed from the database and so the proportion of damaged buildings in each damage band was very likely underestimated.…”

The Global Earthquake Model Physical Vulnerability Database

“…In the empirical approach, statistical regression analyses are applied to earthquake damage data to derive sets of fragility curves (e.g. Rossetto et al 2003;Colombi et al 2008). Fragility curves can also be derived based on the elicitation and pooling of the subjective opinion of a large group of experts (e.g.…”

İstanbul’daki 2000 Yılı Sonrası Binalar İçin Tasarım Depremi Altında Karşılaştırmalı Yapısal Hasar ve Mali Kayıp Tahminleri

“…Geçmiş deprem gözlemlerinden elde edilen hasar verilerine dayanan makro-sismik yöntem ile bölgesel ölçekte hasarlı bina dağılımlarının tahmini gerçekleştirilebilir. Bu iş için şiddet bazlı ampirik hasar-görebilirlik ilişkilerinden yararlanılır [8][9][10][11][12][13]. Kentsel alanlarda mühendislik servisi alarak depreme dayanıklı tasarım ilkelerine göre tasarlanıp inşa edilen binalardan oluşan envanterlerin deprem riskinin değerlendirilmesi için, ki bu yeni binalar için geçmiş deprem hasar verisi bulunmadığı da göz önüne alındığında, analitik yöntemler kullanılması uygun olacaktır.…”