Modeling Rainfall Data Using a Shot Noise Process

Abstract

In this skripsion will be explained about the modeling of rainfall data using the shot noise process. The shot noise process is an extension of the Poisson process which is useful for modeling harmful data where shot marks is a random variable. Arrivals times is  Poisson process and the shot function also called an impulse respon function. Shot noise process was applied to rainfall data at Darmaga Bogor station in 1985-2010. The shot size in this case uses the criteria of BMKG that is greater than 100 mm /day. The result of modeling shows that the frequency data of extreme rainfall at Darmaga Bogor station in 1985-2010 distributed Poisson, and shot noise process on extreme rainfall data at Darmaga Bogor station in 1985-2010 distributed Weibull 3 parameters. Using the moment method estimator , the highest probability of extreme rainfall in 2005.