GIS based methodologies of seismic Hazard and risk analysis for Bangladesh

Abstract

A geographic information system (GIS) provides the ideal environment for conducting a comprehensive regional seismic hazard and risk analysis. GIS has the ability to store, manipulate, analyse and display the large amount of required spatial and tabular data. The system can typically be linked to external computational programs, high level database management systems, and knowledge-based expert systems. The objective of this research is the development of a methodology using geographic information system technology to conduct a regional multi-hazard seismic risk analysis. The term multi-hazard refers to the consideration of ground shaking and secondary site effects of soil amplification, liquefaction, landslide, and surface fault rupture. The methodology involves a modular framework that allows new models and data base information to be included as the technology advances. This thesis describes in detail the current GIS and the various steps in a regional seismic hazard and risk analysis. An overview of the different models for estimating the effects of local site conditions is presented. This research includes the development and example illustration of a GIS-based methodology for quantifying and combining the hazards associated with these secondary site effects. The methodology to combine the various hazards is based on a weighted average approach that utilises the knowledge of local experts. This thesis also covers the estimation of regional earthquake damage and loss, including the development of a methodology for compiling a comprehensive inventory of structures in a large region. An attempt is also made to investigate and distinguish the PGA/intensity values of a region due to liquefaction, land slide and fault rupture. To generate artificial earthquake a locally built SHAKE table with two chambers has been used. The left chamber and the right chamber have been used to investigate the primary effect and the secondary effects of an earthquake respectively. By analyzing the data of ground motion obtained from the two chambers, quantification of secondary site effects of fault rupture, liquefaction, and landslide have been made possible. This thesis also covers the development of acceleration based attenuation relation for Bangladesh. Due to lack of sufficient recorded strong ground motion data, intensity data have been used to develop an attenuation relation by regressing 40 past earthquake records from eight events. This relation shows that Joyner and Boore’s acceleration attenuation model may be adequate to predict the dissipation of acceleration with distance for Bangladesh and its surrounding region. A substantial part of this thesis is devoted to a case study that illustrates the ideas and methodologies developed in this research. The case study shows possible damage and loss within Sylhet City corporation area for a scenario event having the same magnitude and location of 1918 Srimangal earthquake with magnitude 7.6. The various seismic hazards such as liquefaction, landslide and amplification have been quantified and integrated, and a structural inventory of nearly 3040 buildings, 106 km gas lines and about 118 km of water supply lines have been compiled. Economic loss estimation has been estimated using the damages expected to be suffered due to the scenario event. Among total buildings, 59% is expected to be damaged. In case of lifeline, total number of damage points is 204 for water pipes and total number of damage points is 981 for gas pipes. A GIS model has been developed which can be used for estimating regional seismic hazard including secondary site effects for a large region. Numerous maps and tables of inventory data and results have been included in this thesis to help prove the effectiveness of the GIS.