Human Vulnerability to Dam Break Floods

Abstract eng:
Dam breaks caused catastrophic consequences to human lives. The making of a risk mitigation plan requires the evaluation of potential human casualty in a particular dam break event. The assessment of human vulnerability to dam-break floods is at the core in the exercise. The mechanisms of casualties in dam-break floods are complex. McClelland and Bowles (1999) listed 41 means by which people may die in a flood. There are at least four categories of factors influencing the fatality in a flood: hydraulic factors, area and building factors, time and weather factors, and population factors. At the time being the problem of human vulnerability in floods cannot be addressed solely by physically based methods. Rather comprised methods integrating physical mechanisms, observed data and judgments are often used. This raises an interesting Bayesian interference problem: Given observed casualties, what are the chances of the many factors that influence fatality in a flood and the inter-relationships among these factors? Or given the chances of some influence factors in a flood, what is the likely fatality rate in the flood? This paper describes a human risk analysis model (HURAM) for assessing the human vulnerability to dam-break floods. A Bayesian network is constructed considering 15 nodes (parameters) and 23 arcs (inter-relationships) that influence the loss of life according to a logic structure of loss-of-life mechanisms. Four components can be distinguished in the network: evacuation, sheltering, flood severity and loss of life. The nodes and arcs are quantified with historical data, existing models, and physical analyses. The prior probabilities of the medium nodes and the final node (i.e. fatality rate) can be calculated with program Hugin Lite. The suggested fatality rates in Graham’s model and the calculated fatality rates using the proposed model are generally in good agreement. The new model is able to take into account a large number of important parameters and their inter-relationships in a systematic structure, include the uncertainties of these parameters and their inter-relationships, incorporate information derived from physical analysis, empirical models and historical data, and update the predictions when information in specific cases is available. Professor Wilson Tang enthusiastically applied Bayesian statistics to integration of judgmental and scientific knowledge, geotechnical model calibration and landslide risk analysis. The work presented in this paper is along the direction Professor Tang had explored.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket