Evaluation of two common vulnerability index calculation methods

Abstract The potential benefits of using a geometric mean method for computing a vulnerability index are presented using both simulated variables as well as data from a Canadian coastal geodatabase (CanCoast). The assessment of vulnerability of natural and built coastal infrastructure to sea level rise is used to demonstrate the advantages of this method for climate change adaptation planning and decision-making. As with most real world datasets the probability distribution of individual variables in CanCoast may be skewed; this can significantly impact the resulting vulnerability index depending on the calculation method employed. The primary advantage of using a geometric mean is that the index output will remain within the original range and maintain the distributional characteristics of the input variables. This can reduce the need for subjective expert opinion in the process of assessing the vulnerability index. A comparison of the resulting computation using both the Gornitz (1991) method and the geometric mean is provided for the Atlantic Canada coastline.