Development of pacific exposure dataset for use in catastrophe risk assessment

An exposure dataset is one of the critical components of catastrophe risk modelling. It can also be one of the most difficult to create. When small numbers of buildings are involved, less than a thousand or so, it may be practicable to view each building and to acquire all the necessary attributes with a reasonable level of confidence. For regional or national-scale projects, however, the task is next to impossible. Ideally, for each building within the modelling areas we would know the location and value, and have enough structural information to underpin the assignment of vulnerability functions for each of several hazards. Such a dataset does not exist in the Pacific region, though various exposure databases have been developed in the past couple of decades. Instead, we have developed a systematic approach to populate and maintain a somewhat less rigorous, but practicable, exposure dataset for use in catastrophe risk modelling. In this paper, a brief review of previous development of exposure databases in the Pacific region is presented, followed by an overview of each class of available data that is used to develop the exposure dataset. Next, the methodologies adopted are illustrated, and application to a test case in Tanna Island, Vanuatu is described. Finally, the proposed exposure dataset development is discussed.