Improving the Seismic Exposure Model of Lebanon: A Risk Sensitivity Analysis

Abstract

Having a representative seismic building exposure model is essential to perform risk assessments and obtain accurate risk metrics, such as the damage state of buildings, and human and financial losses. An exposure model for Lebanon was developed by the Global Earthquake Model (GEM) Foundation in 2018, and consists of information regarding the building stock, and in particular building classes which include the following attributes: material, lateral load resisting system, code level, earthquake resistance level, and height. The model also provides information related to the cost and the number of occupants, which are utilized to calculate risk metrics. Due to the lack of recent census data and limited knowledge of the Lebanese construction industry, this current exposure model exhibits several inaccuracies, potentially yielding erroneous risk metrics. This research aims to improve the building exposure model for Lebanon by relying on an extensive literature review, census data, datasets, and consultations with local experts. The effect of revising the exposure model on risk metrics is evaluated by performing a comparative risk assessment using the GEM OpenQuake engine for the same hazard input, namely the 2014 Earthquake Model of the Middle East (EMME14) hazard model, with both the current and revised exposure models for Lebanon. The comparison shows a notable increase in the expected annual losses when the revised exposure model is used. The number of displaced people rises from 8,100 to 16,000, while the number of injured increases from 180 to 370. Casualties also more than doubled, climbing from 62 to 150. Similarly, the number of buildings damaged beyond repair increases from 610 to 1,300. Certain metrics are even more affected. The total floor area lost nearly quadruples, rising from 0.22 million sqm to 0.83 million sqm. Economic losses follow a similar pattern, surging from 340 million USD to over 1.2 billion USD. These are partly explained by a lower share of seismically designed buildings, and larger average height and floor area. Non-seismically designed low-rise unreinforced masonry or mixed materials buildings, and low to mid-rise reinforced concrete buildings with infill frames were found to be the most vulnerable building classes, leading to the largest number of casualties, building collapses, and economic losses. This study thus demonstrates the importance of having a representative exposure model to obtain accurate risk metrics, which can be used to encourage officials to implement and introduce policies to reduce losses.