Fourier Series in a Neyman Scott Rectangular Pulse Model

Abstract

The ability of Fourier Series to exhibit seasonal fluctuation of rainfall process is presented. The Neyman Scott Rectangular Pulse Model with mixed exponential distribution for cell intensity is selected to describe the rainfall process. The model’s parameters were estimated by employing the Shuffle Complex Evolution (SCE-UA) method. Seasonal variation is dealt with by fitting Fourier Series to the parameters. Significant harmonics for each parameter is determined using the cumulative fraction of total variance explained by significant harmonic. Results indicate seasonal fluctuations of parameters were sufficiently represented by the Fourier Series. Comparison between Fourier Series estimations and observed data of 10 years demonstrated the ability of Fourier Series in capturing the statistical characteristics of rainfall process. Keywords: Fourier series; Neyman Scott rectangular pulse model; harmonic.