On Some Uses of Non-normal Coefficients of Variation

Abstract

Coefficients of variation larger than 1 are frequently encountered in arid-land climatology where rainfall distributions are non-normal. The coefficient need not be considered in the context of Gaussian statistics. In spite of their apparent incongruity, large, non-normal coefficients may be useful, both in characterizing and regions ecologically and in stating precipitation probabilities for extreme cases where normal probability integrals cannot be applied.