Ramesh, Nadarajah ORCID: https://orcid.org/0000-0001-6373-2557 and Garthwaite, Andrew (2017) Some stochastic models for seasonal rainfall at fine time-scales. In: 39th Conference on Stochastic Processes and their Applications, July 24 - 28, 2017, Moscow, Russia. (Unpublished)

Preview

PDF (Author Accepted Abstract) 17864 RAMESH_Stochastic_Models_for_Seasonal_Rainfall_2017.pdf - Presentation Download (71kB) | Preview

Abstract

Stochastic point process models have been widely used to model rainfall time series. Doubly stochastic Poisson processes provide a rich class of models for analysing fine time-scale rainfall data. Models of this type have been used by several authors to describe fine-scale rainfall characteristics at a single site as well as at multiple sites. Ramesh et al. (2013) developed a class of multisite models for analysing tipping-bucket rainfall data recorded over a number of stations in a catchment area. In this paper, we extend the univariate class of models for fine time-scale rainfall to accommodate seasonality and study a number of seasonal doubly stochastic Poisson process models. This includes models incorporating atmospheric covariates in the analysis. The application of these models is illustrated in the modelling of sub-hourly rain gauge data from England. One of the advantages of this class of models, when compared with similar models, is that their likelihood function can be calculated in a tractable form suitable for numerical optimisation. This allows us to use the maximum likelihood approach to estimate the parameters of the proposed stochastic models. We use some of the second-order properties of the fine-scale rainfall aggregations in discrete time intervals for model assessment.

Item Type:	Conference or Conference Paper (Paper)
Uncontrolled Keywords:	doubly stochastic; rainfall modelling; seasonal rainfall; fine-scale rainfall
Subjects:	Q Science > QA Mathematics
Faculty / School / Research Centre / Research Group:	Faculty of Engineering & Science > School of Computing & Mathematical Sciences (CMS) Faculty of Engineering & Science
Last Modified:	04 Mar 2022 13:07
URI:	http://gala.gre.ac.uk/id/eprint/17864

Actions (login required)

View Item

Downloads