Stochastic point rainfall modeling for correlated rain cell intensity and duration

A new stochastic point rainfall model which considers the correlation structure between rain cell intensity and duration is developed. In order to consider the positive and negative correlations simultaneously, the Gumbel's Type-II bivariate distribution is applied. For cluster characteristics of rainfall events, the Neyman-Scott cluster point process is applied. The application is performed using the rainfall station of Jeonju situated on the southwest part of the Korean peninsula. Results from generating long time rainfall events show that the proposed model reproduces well the statistical characteristics of the historical rainfall time series and the model-generated data are robust with different parameter sets when the correlation parameter is appropriately taken.