A Bayesian Approach to Seismic Hazard Mapping: Development of Stable Design Parameters

The energy released by earthquakes propagates in the earth's crust as body and surface waves. The intensity and duration of shaking of structures located in the path of these waves depends upon the inten­sity and duration of the seismic ground motion along with the char­acteristics of the structure. Structural failures resulting in con­siderable damage and loss frequently occur because of these motions and inadequate seismic resistance of the structures.

Earthquake engineers and planners often use the words risk and hazard interchangeably in their work. Seismic risk is regarded by many to be synonymous with seismic hazard. There is some danger in this ambiguity since these two words for seismic phenomenon have different meanings. Seismic hazard is defined as "expected occurrence of future adverse seismic event (earthquake)". Seismic risk is defined as "expected consequences of future seismic event". Consequences may be life loss, economic loss, function loss and damage. Loosely, it can be said that a seismic hazard involves "nature's punch" while a seismic risk involves interaction between "nature's punch" and human activity.

The intensity and duration of future earthquake ground motions are random and can therefore be known only in the probabilistic sense of the likelihood of exceeding a given level during a given time period. (Rosenblueth and Esteva, 1966; Benjamin, 1968; Cornell, 1968). If economic planning and engineering design criteria are to be formulated on a rational basis,then it is necessary to have the best available estimates of these future ground motions. The best practical representation of earthquake loadings for a given geographical region is in the form of seismic hazard maps -- where the earthquake effect is shown in terms of the most useful engineering parameters for design. Presently there is a great need for improvements in risk mapping techniques and in the description of the related engineering parameters.

Therefore the present dissertation is divided in two parts. The first part concentrates on seismic hazard mapping which can be best defined as the exposure to seismic loading at a given location. This exposure is expressed in terms of an effect and the probability of its occurrence. The second part concentrates on a study of stable design parameters. Its general purpose is to provide a statistical and probabilistic view of the response of structures to earthquake excitation. The attention is focused on response parameters which have a direct engineering value.