Modeling and Forecasting Electricity Consumption Amid the COVID-19 Pandemic: Machine Learning vs. Nonlinear Econometric Time Series Models

Accurately modelling and forecasting electricity consumption is a key prerequisite for strategic sustainable energy planning and development. In this study, we use four advanced econometrics time series models and four machine learning (ML) and deep learning models including an AR with seasonality, ARX, ARFIMAX, 3S-MSARX, Prophet, XGBoost, LSTM and SVR to analyze and forecast electricity consumption during COVID-19 pre-lockdown, lockdown, releasing-lockdown, and post-lockdown phases. We use monthly data on Qatar’s total electricity consumption from January 2010 to December 2021. The empirical findings demonstrate that both econometric and ML models can capture most of the important statistical features characterizing electricity consumption (e.g., seasonality, sudden changes, outliers, trend, and potential long-lasting impact of shocks). In particular, we find that climate change based factors, e.g temperature, rainfall, mean sea-level pressure and wind speed, are key determinants of electricity consumption. In terms of forecasting, the results indicate that the ARFIMAX(1,d,0) and the 3S-MSARX(1) models outperform all other models. Policy implications and energy-environmental recommendations are proposed and discussed.