A Multisite Stochastic Downscaling Model of Daily Rainfall Occurrences with Long Term Persistence

Direct use of outputs from the General Circulation Models (GCMs) for climate change impact assessment is often limited by their incapability at representing local features and dynamics at spatial scales finer than the in-built GCM grid scale. This has led to the development of downscaling techniques for transfer of coarse GCM simulated weather output to finer spatial resolutions. However, the downscaling models are often suspected for their validity in the future climate conditions and their inability to simulate the hydrologic extremes. This paper presents a stochastic downscaling model for simulation of multi-site daily rainfall occurrences with the aim of proper simulation of rainfall extremes. At-site rainfall occurrences are modelled using a Modified Markov model (MMM) as described in Mehrotra and Sharma (2007) that defines the temporal persistence in the rainfall occurrence by updating at each time step the Markovian transition probabilities on the basis of recent past rainfall behaviour and the selected atmospheric variables. The spatial dependence across the rainfall occurrence field is maintained through spatially correlated random numbers and atmospheric and other variables defining the history of rainfall in the recent past, common across the stations. The proposed model is applied for downscaling of rainfall occurrences at a network of 45 raingauge stations around Sydney in Australia and its performance evaluated. The analyses of the results show that the scheme of updating the transition probabilities of the at-site rainfall occurrence model and the logic of providing spatial treatment separately, imparts considerable accuracy and flexibility in the representation of characteristics of interest in hydrologic studies. These characteristics include representation of spell patterns, spatial distribution, and low and higher time scale persistence characteristics and as generic indicators of water balance and variability that are of importance in a catchment scale water balance simulation. [GRAPHICS] .