Study of Time Series and Development of System Identification Model for Agarwada Raingauge Station

Rainfall data from the year 1965 to 2005 have been analyzed and then plotted with respect to time to analyze time series plot and autocorrelation plot.. From the autocorrelation plots, it is observed that upto lag 4 to lag 5, which is around 10% of the length of the data, the results shows better correlations and for lags beyond 10% of the record length, there exist less correlation and more variations. Further, the rainfall data have been used to develop and analyze model using soft computing technique called System Identification Model (SIM) to predict and forecast future values. An attempt has been made to develop the best fit model for Agarwada Station of Panam catchment area by trying different models like linear parametric models, process models, correlation models and non-linear models. Model is developed for 70% data and validated for remaining 30% data. Best Fit values of each model has been checked and compared for analysis purpose. In present study, the best fit value comes out to be 100 and it is concluded that process models are the best preferred models for Agarwada Station According to Akaike's theory, the most accurate model has the smallest FPE and Loss Function. The results obtained for the same are 3.46e-024 and 2.19e-024 respectively, which are almost equal to zero.