Timeseries data for model building plays a primary role in the analysis, prediction, and forecasting of feature values and explains the structure of the dataset, which is helpful for classification and forecasting problems. Usually, in time series modelling, the main objective is to collect data carefully and study rigorously the past data observations to develop an inherent structured model for prediction and forecasting. This model, which is used to generate future and forecasted time series data, The main focus is on seasonal time series data collected from the SCATSAT-1 Scatterometer data set. The present article shows that to develop first- and second order auto-regressive models, a fuzzy seasonal auto-regressive integrated moving averages model with heteroscedastic disturbances was developed for the forecasting and predic-tion of future datasets in a fuzzy environment. A fuzzy membership function was developed for the classification of the dataset using the inflexion point methodology. For model adequacy checking, we adopted measures such as mean absolute percent error, mean forecast error, mean squared error, and root mean squared error.