Estimation of extreme rainfall exceedance probabilities: nondimensional stochastic storm transposition

A new stochastic storm transposition (SST) approach is developed to obtain credible estimates of extreme rainfall frequencies up to and beyond the probable maximum precipitation (PMP). The extreme rainfall frequencies are estimated by solving a joint probability integral which combines the probability of two phenomena that affect extreme rainfall probability: 1) The probability of storms arriving in the storm transposition region; and 2) The probability of a storm centering in the catchment threshold area, a concept arising from earlier research on stochastic storm transposition. Unlike earlier attempts to implement stochastic storm transposition, this study uses nondimensionalisation to overcome the problem of statistical nonhomogeneity for which extreme storm depths with the same exceedance probability vary systematically across the transposition region. Examples based on the Inland Generalised Southeast Australia Method transposition zone illustrate the approach. Comparison of the nondimensional SST approach with Pearse (1997) joint probability method showed good agreement. Both methods estimated the annual exceedance probability of the PMP to be about 10-5. It is concluded that the nondimensional SST approach represents a significant advance in Australian practice for estimating extreme rainfall frequencies, though more work is required to assess the approximations used in the approach.