A Comparative Study on Methods for Loss of Life Estimation - Applications to Case Studies in the United States (PAP014528)
Sebastaiaan N. Jonkman, B. Maaskant, B. Kolen, M. Zethof, W. P. Lehman, J. T. Needham
Flood risk management in mega cities
Both in the USA and in the Netherlands extensive studies on approaches for identifying flood risks- the combination of levee failure probabilities and consequences are ongoing. Important elements of these studies are methods to analyze life loss for floods and the effectiveness of evacuation and emergency management. Both in the Netherlands and the USA, the risks to life associated with flooding are also taken into account in the national policy. It is therefore important that the methods that are utilized to analyze risks to life are well-validated. The main objective of this study is to compare Dutch and American methods for the analysis of loss of life and evacuation for a number of case studies in the US (i.e. a comparison effort). In the first part an overview and comparison of methods for loss of life estimation and evacuation analysis that have been developed in the Netherlands and the United States. In the United States HEC FIA and Lifesim are used, which are more process-based models that take into account the dynamics of flooding and people movement in the area. The Netherlands used empirical curves to relate mortality to flood conditions. The potential for evacuation in the Netherlands is analyzed using a combination of macro-scale traffic models and response time. Similarities and differences between the modeling approaches and input and output types are discussed. In the second part of the studies a number of case studies have been selected as part of the comparison effort. The Natomas Basin in California is a low-lying area with a population of about 100,000 and is threatened by flooding from the Sacramento river. The various modeling approaches give very different outcomes for the life loss and these are dependent on the different responses of the various models to the large flood depths (up to 5m) that can occur. Also, the potential for evacuation of the Natomas Basin is analyzed, and depending on evacuation efficiency the required time for evacuation for the basin ranges between 7 hours and more than 24 hours. A second case study concerns the Herbert Hoover Dike in Florida, a large dike on Lake Okeechobee with two small communities in the vicinity. For this breaching scenario water depths and flow velocities are more moderate and differences between model outcomes are smaller. These cases are the first steps in a benchmark research on life loss models. It is discussed how the empirical basis of life loss models can be improved, e.g. by using available datasets for the life loss due to the flooding of New Orleans during hurricane Katrina (2005). In a further comparison and benchmarking effort other models and empirical datasets can be added.