Estimation of Carbon Dioxide Concentration for Baghdad City Based on
Artificial Neural Networks
Saif Hassan, Ali M. Al-Salihi Hasoon, Jasim M. Rajab
The study developed a model for predicting CO2 concentration in the center of Iraq's city of Baghdad using artificial neural networks that show the quantities of such emissions that are mainly global warming and can accurately predict greenhouse gas emissions for up to 20 years, supporting environmental management strategies and policies. And using three functions and neural networks. An artificial neural network (ANN) model was used to estimate the ultimate CO2 concentration. Depending on the values of the error and the coefficient of correlation, the ANN model indicated that the suggested model was better appropriate to characterize CO2 in any part of the world. It was shown that the Scaled Conjugate Gradient method (SCG) had the lowest height value of (R) and mean-squared error (MSE). The ideal neuronal count for the SCG hidden layers was 10, with an MSE of 17.24 and R = 0.931. Consequently, the ANN demonstrated outstanding performance in the CO2 prediction value. The results of the study emphasize how important it is to use CO2 prediction as the primary parameter for assessing the impacts of global warming.
Keywords: Carbon dioxide, Artificial neural networks, Global warming, Greenhouse gas, Baghdad city
