Modeling and forecasting of metrological factors using arch process under different errors distribution specification

Various weather phenomenon are difficult to model and forecast with high precision. This study has modelled and forecasted the various parameter namely maximum and minimum temperature, morning and evening relative humidity using parametric models namely Autoregressive Integrated Moving Average (ARIMA) and Generalized Autoregressive conditional heteroskedasticity (GARCH)) models. The data consisted of daily time series data for Hoshangabad district of Madhya Pradesh from January, 1996 to November, 2019. The AIC and BIC criterion were used to select among competing models. Present investigation has revealed that ARIMA-GARCH models are more suitable for forecasting of minimum temperature, maximum temperature and relative humidity.