A TECHNIQUE FOR SELECTING THE BOX-COX TRANSFORMATION IN FLOOD FREQUENCY-ANALYSIS

In hydrology, the Box-Cox power normal transformation has been found t o perform well in a number of at-site and regional flood frequency stu dies. In this paper, an alternative to the maximum likelihood method o f finding the Box-Cox transformation parameter, lambda, is suggested. The proposed method is based on finding the lambda that maximizes the correlation coefficient in the probability plot correlation coefficien t test for normality. The proposed method of finding lambda was tested on maximum daily discharge data from all over Canada and was found to give almost identical results to that obtained by the method of maxim um likelihood. The main advantage of the proposed method is that findi ng the appropriate transformation and objectively testing whether the transformed data are normally distributed can be done in one step. In other methods, at least two steps are required. In addition, the propo sed method makes use of techniques that are familiar to the hydrologis t, namely, the probability plot and correlation analysis.