B0452
Title: Markov modulated Poisson processes to analyse rainfall time series
Authors: Nadarajah Ramesh - University of Greenwich (United Kingdom) [presenting]
Abstract: Researchers in modelling rainfall time series use several stochastic point process models. Chief among them are cluster-based point process models constructed from either Bartlett-Lewis or Neyman-Scott processes. We describe a class of point process models based on a Markov modulated Poisson process that are useful in analysing rainfall time series. In particular, we discuss recent results from an exponentially decaying rainfall pulse model developed from this class of stochastic point processes. The proposed model is utilised to model hourly and sub-hourly rainfall data. The results of our analyses suggest that the proposed class of models provides a useful addition to the existing array of stochastic models for analysing fine-scale rainfall data.
