The devastation of the past earthquake events such as Kocaeli (1999), Kashmir (2005) and Haiti (2010) highlights the need for an earthquake risk assessment (ERA) framework, which can be applied to developing countries to reduce future earthquake losses. Although there are several ERA frameworks used in current practice, these frameworks are not flexible or applicable to the areas with limited input data and lacking assessment of earthquake induced secondary hazards. One of them is liquefaction, which is capable of causing settlement, tilting or even collapse of the structures. As an example, numerous liquefaction manifestations in Adapazari during 1999 Kocaeli and 1999 Chi-Chi earthquakes caused damage to a significant number of buildings due to ground failures. This past experience highlighted the importance of the integration of liquefaction hazard assessment into the existing or future earthquake risk assessment frameworks.
This research presents the development of the multi-hazard ERA framework, which can be applied to the areas with limited input data available. The proposed framework is based on Monte-Carlo simulations and consists of three main modules: Probabilistic Seismic Hazard Analysis (PSHA), Probabilistic Liquefaction Hazard Analyses (PLHA) and Seismic Risk Assessment.
The developed PSHA module generates earthquake catalogues using Monte-Carlo simulations to represent future seismicity of the region of interest. In this module, both Poisson and time dependent (renewal) models are adopted to quantify the effect of temporal dependencies between seismic events, while near-field rupture directivity effects are also taken into account. The Marmara region in Turkey is selected as a case study region. The PSHA results compare well with the recent studies performed for the study region.
The PLHA module of the framework uses Monte-Carlo simulations and simplified liquefaction assessment procedures to predict earthquake induced liquefaction hazard in a probabilistic way. The module provides a relatively flexible approach to treat uncertainties of the earthquake and soil related input parameters by including such parameters with distribution functions. Liquefaction Potential Index (LPI) and Liquefaction Severity (LS) procedures are adopted for the prediction of liquefaction hazard. The PLHA module is applied to the city of Adapazari in Turkey, which is located in a liquefaction prone area. A set of probabilistic liquefaction hazard maps are derived for Adapazari as there were no PLHA studies performed for the city. The PLHA module is verified with a scenario earthquake and observed liquefaction from the 1999 Kocaeli earthquake.
Seismic risk assessment module uses readily available data such as building stock and population data from census, global shear-wave velocity maps and satellite images. Seismic risk assessment module is applied to Adapazari. An exposure model for buildings in the case study area is developed by mapping buildings footprints from aerial and satellite images and using remote street survey. The risk results are presented in terms of loss curves and probabilistic mean damage ratio (MDR) maps considering various return periods and vulnerability models. Casualty models are incorporated into the framework for estimation of fatalities associated with earthquakes in the city of Adapazari. Social and economic losses are verified with a scenario earthquake similar to 1999 Kocaeli earthquake.
The developed ERA framework can serve as an efficient tool for the assessment of liquefaction hazard and seismic risk in the areas with limited data available. Although, due to lack of reliable procedures for accounting damage due to liquefaction in large scale seismic risk studies, the risk module of the framework should be used with caution in the areas prone to liquefaction. The proposed framework and its outcomes can be used by policy and stakeholders for earthquake preparedness and developing emergency response and recovery strategies.
