Predictive intelligence of machine learning models for global energy perspectives and transformations towards sustainability

Climate change is affecting ecosystems, communities, and human health worldwide. These changes pose risks to global energy systems so there is a dire need to combat climate change and limit global warming to 1.5°C. This study undertake global energy systems and forecasted total energy consumption, production and greenhouse gas (GHG) emissions worldwide for the study period 2021–2050 by taking the input data from 1970 to 2020 using the four algorithm’s namely, Holt Winter (HW), Exponential Smoothing (ES), Autoregressive Integrated Moving Average (ARIMA), and Seasonal Autoregressive Integrated Moving Average (SARIMA) implemented in Python. It is found that HW and ES have same forecast results globally with energy consumption of 236,285 TWh which can easily meet by 475,980 TWh generation until 2050. Renewables and fossil fuels contributed to 250,106 TWh units and 225,874 TWh units with 48 billion metric tons of GHG emissions until 2050. The global forecast of ARIMA model suggested that 232,878 TWh energy consumption is noticed which can easily meet by 446,126 TWh generation with 213,052 TWh share of renewables and 233,074 TWh of fossil fuels with 49 billion metric tons of GHG emissions produced until 2050. SARIMA model forecast is very much valuable for limiting global mean temperature to 1.5 °C. The global energy consumption is forecasted to be 231,022 TWh which easily meet by 350,054 TWh green energy generation potential with almost zero emissions until 2050 and it is found that SARIMA model has 98% of accuracy.