Abstract

Chile is a country affected continuously by large-scale seismic events, including those that occurred during the 2014 (Iquique, M_w=8,2) and 2015(Illapel, M_w=8,3) earthquakes. Estimating the impact of earthquakes on structure inventories can support the development of risk reduction strategies. Previous events are helpful for calibrating risk models and contribute to understanding the consequences of earthquakes. It should be noted that empirical data are available to estimate future damage in Chile. The objective of this study is to estimate the empirical fragility curves of reinforced concrete, reinforced masonry, adobe, and timber houses using damage information from the 2014 and 2015 earthquakes. The fragility curves were estimated using three Intensity Measures (IM) to identify which IM was better correlated with the observed damage and which fragility curves better adjusted the observed data using the Akaike Index Criteria value. These IMs were obtained from shake maps by the United States Geological Survey (USGS) and were used as the Intensity Measure Type (IMT) to express ground shaking. The methodology used has three steps: (1) organize the damage observed for the 2014 and 2015 earthquakes according to the Hazus damage scale (slight, moderate, extensive, and collapse); (2) consider IM are: the Peak Ground Acceleration (PGA), Peak Ground Velocity (PGV), and Spectral Acceleration at a period of 0.3 seconds (S_a(0.3)); and (3) estimate the empirical fragility curves using a Generalized Linear Model with Probit as Link function. The obtained results indicate that, for the analyzed houses, S_a(0.3) is the IM that correlates better with the observed damage. A better adjustment of fragility curves for observed damage was obtained using PGA for reinforced concrete and reinforced masonry houses, and for adobe and timber houses, using PGV.

Keywords:Fragility curves; Earthquakes; Data survey; Intensity measure; PGA; PGV; S_a(0.3)