Climate Data Forecasting Using ARIMA and Holt Winter Methods

The Artificial Neural Networks (ANNs) models have provided mixed results to solve linear problems, so it is not widely used for any data types. The autoregressive integrated moving average (ARIMA) model is a massive, very popular, and connivance linear model for long time series forecasting. Mainly used to construct different hybrid models for forecasting other time series data in the last decade. The study analyzes ARIMA and Holt-Winters (HW) forecasting procedure analysis. The machine learning and deep learning model have good frameworks that can apply to multiple forecasting processes with good precision and accuracy. Autoregressive Integrated moving average (ARIMA) and Holt-Winters (HW) approaches have been used to forecasting time series climatic variables. The two different forecasting methods are generated, and the output of both methods is compared by the climate variables rainfall and temperature for monthly data of 2001-2021.