Data-driven probabilistic post-earthquake fire ignition model for a community

• Published in: Fire safety journal Vol. 94; pp. 33 - 44
• Main Authors: Elhami Khorasani, Negar, Gernay, Thomas, Garlock, Maria
• Format: Journal Article Web Resource
• Language: English
• Published: Lausanne Elsevier Ltd 01.12.2017; Elsevier BV; Elsevier Science
• Subjects: Civil engineering; Communities; Community; Data processing; Earthquakes; Engineering, computing & technology; Fire following earthquake; Fires; Geographic information systems; Hazus; Ignition; Ingénierie civile; Ingénierie, informatique & technologie; MAEViz/Ergo; Platforms; Probabilistic; probabilistic model; Probability; Remote sensing; Satellite navigation systems; Seismic activity; Statistical analysis; Hazus; Fire following earthquake; MAEViz/Ergo; Probabilistic; Community; Ignition
• ISSN: 0379-7112; 1873-7226; 1873-7226
• DOI: 10.1016/j.firesaf.2017.09.005

Fire following earthquake (FFE), a cascading multi-hazard event, can cause major social and economical losses in a community. In this paper, two existing post-earthquake fire ignition models that are implemented in Geographic Information System (GIS) based platforms, Hazus and MAEViz/Ergo, are reviewed. The two platforms and their FFE modules have been studied for suitability in community resiliency evaluations. Based on the shortcomings in the existing literature, a new post-earthquake fire ignition model is proposed using historical FFE data and a probabilistic formulation. The procedure to create the database for the model using GIS-based tools is explained. The proposed model provides the probability of ignition at both census tract scale and individual buildings, and can be used to identify areas of a community with high risk of fire ignitions after an earthquake. The model also provides a breakdown of ignitions in different building types. Finally, the model is implemented in MAEViz/Ergo to demonstrate its application in a GIS-based software. •Fire Following Earthquake modules in Hazus and MAEViz/Ergo are investigated.•A probabilistic post-earthquake fire ignition model is proposed based on historical data.•The proposed model provides results at census tract and individual building levels.•The model and a GIS-based platform can be used to evaluate resiliency of a community.