Identification of regional parameters of a stochastic model for rainfall disaggregation

This paper demonstrates how the Gyasi-Agyei-Willgoose hybrid model for poin t processes could be regionalised for daily rainfall disaggregation using l imited high resolution data within a region of interest. Their model is a p roduct of the binary non-randomised Bartlett-Lewis rectangular pulse model and a lognormal autoregressive model used as a jitter. The computationally efficient multinormal approximation to parameter uncertainty is used to gro up the monthly parameter values of the binary model. For central Queensland , Australia, it has been established that the parameters of cell origins an d the duration of the rectangular pulse of the binary model could have cons tant values for all months. Second harmonic Fourier series is used to repre sent the seasonal variation of the parameter governing the storm lifetime. The storm arrival rate is a function of the daily dry probability and the o ther parameters. Additive properties of random variables with finite varian ces were used to scale down the daily mean and variance of the historical d ata to the simulation timescale, values required by the jitter model. The r esults of using observed daily rainfall statistics to capture sub-daily sta tistics by the regionalised model are very encouraging. The model is theref ore a valuable tool for disaggregating daily rainfall data. (C) 1999 Elsevi er Science B.V. All rights reserved.