Carbon dioxide (CO2) emissions pose a significant environmental challenge globally, deteriorating climate change and its associated impacts. In Bangladesh, rapid industrialization and urban expansion have intensified CO2 emissions, requiring timely and accurate forecasting for effective mitigation strategies. This study proposes a systematic approach to near-real-time CO2 emission prediction, utilizing advanced statistical (ARIMA, SARIMAX), machine learning (RF, LSTM) models and LSTM integrated SARIMAX. By evaluating prediction accuracy across different energy sectors, including renewables, bioenergy, solar, wind, hydro, nuclear, gas, and coal, we identify the most effective forecasting model. Our findings provide crucial insights for policymakers, enabling informed decision-making and proactive emission reduction measures. This research contributes to addressing critical gaps in CO2 emission prediction and providing improved methodologies for forecasting in Bangladesh.
Analyzing Carbon Dioxide Emissions and Various Energy Sources in Bangladesh Using Statistical and Machine Learning Models to Forecast Future Trends: An Exploration of Performance of Different Forecasting Models
13 Views
7 Downloads