A continuous point process model for daily rainfall (Q1113562)

A probabilistic model for the temporal description of daily rainfall at a fixed point is presented. The model is a member of the family of point process models. Model development is based on statistics estimated from rainfall data in Lebanon. Scale considerations for Markovian models and a theory of projection are used to determine the continuous process of alternation between dry and wet periods. The wet spells are defined by a number of storms each of which is associated with a storm depth \(Y_ i\) and an interstorm time interval \(T_ i\). Computational results are presented for data from Lebanon. The model is successful in preserving the structure of the occurrence process, as well as the distributional properties of the rainfall amount.