Probability Modeling of Carbon Dioxide Emissions in Tanzania

Abstract

Carbon dioxide emissions are a significant driver of climate change, impacting ecosystems, human health, and economies. In Tanzania, increasing carbon dioxide emissions from fossil fuel use, deforestation, and industrial growth contribute to environmental challenges such as air pollution, agricultural disruption, and changing weather patterns. This study models Tanzania’s carbon dioxide emissions using the Generalized Log-Logistic distribution and compares it to other models, including Burr XII, log-logistic, Weibull, and log-normal distributions. The results show that the Generalized Log-Logistic distribution provides the best fit to the data, outperforming other models in goodness-of-fit, log-likelihood values, and information criteria. Three estimation methods such as maximum likelihood, least squares, and weighted least squares were applied, with maximum likelihood yielding the lowest mean squared error, making it the most effective for parameter estimation. The likelihood ratio test further confirmed that the Generalized Log-Logistic model offers a superior fit compared to its sub-models, demonstrating its robustness in representing carbon dioxide emissions data. These findings establish the Generalized Log-Logistic model as a valuable tool for monitoring carbon dioxide emission patterns in Tanzania, essential for managing rising emissions. This study underscores the importance of reliable probability models in addressing environmental challenges and informing strategies to mitigate carbon dioxide emissions.