<![CDATA[Carbon dioxide (CO₂) emissions are one of the primary causes of climate change, which has a significant impact on the environment and human health. As the second-largest emitter of carbon dioxide after China, the United States requires an effective forecasting system to monitor and control these emissions. This study aims to develop a time series model with independent variables using a combined SARIMA-QR method and compare its accuracy with the SARIMA model. The independent variables used include total industrial energy consumption, total electricity consumption, and the forecast values from the SARIMA model. The comparison of model accuracy is based on the MAPE values from the testing data between the SARIMA model and the SARIMA-QR model at the 0.50 quantile. The analysis results show that the SARIMA model achieves a MAPE value of 5.78%, while the SARIMA-QR model at the 0.50 quantile has a lower MAPE value compared to the SARIMA model. The improvement in accuracy in the SARIMA-QR model is due to the integration of independent variables, which provide additional relevant information, such as total industrial and electricity consumption, as well as the forecast values from the SARIMA model. This demonstrates that the use of independent variables can improve the accuracy of CO₂ emission predictions. The comparison of the accuracy of these two models is expected to serve as an important reference for the United States government in formulating more effective policies to manage carbon dioxide emissions more optimally.]]>
Copyrights © 2025
