Testing for Nonlinearity in Time Series of Monthly Precipitation in Panxi Region

Based upon the introduction of the phase space reconstruction, the nonlinearity of monthly precipitation time series in Panxi region is investigated by using the data of monthly precipitation time series for past 47 years.First the use of quantitative G-P correlation dimension method is described in detail.What is more, this theory is applied in investigating the method of searching for the main quantitative indexes.These parameters include the saturation correlation dimension D2 and Kolmogorov entropy.The results indicated that there is a nonlinear chaotic characteristic in monthly precipitation time series.Then the power spectrum of monthly rainfall time series analysis method is used to validate their nonlinear chaotic characteristic.Besides, the Cao method is applied to exclude the possibility that the precipitation time series are random.The research on nonlinear chaotic characteristics of monthly precipitation will provide the theory basis for further studying the complexity and evolutionary law of monthly precipitation time series.