STOCHASTIC CHARACTERISTICS OF DAILY RAINFALL AT PURAJAYA REGION
Ahmad Zakaria · 2011
Aim of this research is to study stochastic charact eristics of daily rainfall series. The study was un dertaken using 25 years (1977-2001) data of Purajaya region. The s eries of the daily rainfall data assumed was trend free. The periodic components of daily rainfall series were represente d by using 253 harmonic expressions and stochastic components of daily rainfall series were presented using second o rder autoregressive parameters. Validation of gener ated daily rainfall series was done by comparing between generated with measured daily rainfall series. For periodic model ing, mean of the correlation coefficient between generated and measu red daily rainfall series with the number of the da ta N is equal to 512 days for 25 years was found to be 0,9576. For perio dic and stochastic modeling, mean of the correlatio n coefficient was found to be 0.9999. Therefore, developed periodic a nd stochastic model could be used for future predic tion of daily rainfall time series. Keyword: daily rainfall, fast fourier transform, fourier an alysis, autoregressive model, least squares method.