Hidden Markov Models Incorporating Additional Dependence in Regional Rainfall Modeling

• Activity Number: 587
• Type: Contributed
• Date/Time: Wednesday, August 3, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Nadarajah Iyer Ramesh*+
• Companies: University of Greenwich
• Keywords: Hidden Markov models ; Rainfall modelling ; Maximum likelihood ; Regional precipitation series ; Daily rainfall

Abstract

Hidden Markov models can be modified in several ways to form a rich class of flexible models that are useful in many environmental applications. One of the issues that come up very often when basic hidden Markov models are used to model environmental data is about their ability to accommodate sufficient dependence between observations. We consider a class of hidden Markov models that incorporate additional dependence among observations to model daily rainfall time series. The focus of the study is on models that introduce additional dependence between the state level and the observation level of the process and also on models that incorporate dependence at observation level. Construction of the likelihood function of the models is described along with the usual second order properties of the process. Maximum likelihood method is used to estimate the parameters of the models. Application of the proposed class of models is illustrated in an analysis of regional daily rainfall time series from South East England for the winter season during 1931 to 2010.