Scholarlink

Introduction <p>Carbon dioxide (CO₂) emissions are a major contributor to global climate change. In Ethiopia, CO₂ emissions have increased steadily in recent decades due to industrialization, rising energy consumption, expanding transportation systems, continued dependence on biomass fuels, and rapid population growth. Although Ethiopia has historically contributed only a small share to global greenhouse gas emissions, the recent upward trend poses challenges for sustainable development and environmental management. Accurate forecasting of CO₂ emissions is essential for designing effective climate mitigation strategies and evidence-based policy decisions. Therefore, this study aimed to analyze and forecast CO₂ emissions in Ethiopia using Bayesian Autoregressive Integrated Moving Average (ARIMA) and Bayesian Structural Time Series (BSTS) models.</p> Methods <p>This study analyzed and forecasted carbon dioxide (CO₂) emissions in Ethiopia using an 82-year time series dataset covering the period from 1941 to 2022. Two Bayesian time series models were employed: Bayesian ARIMA and Bayesian Structural Time Series (BSTS). The Bayesian ARIMA model captured temporal dependencies through autoregressive and moving average components, whereas the BSTS model decomposed the time series into trend, seasonality, and regression components, allowing the incorporation of external predictors. Model parameters were estimated using Markov Chain Monte Carlo (MCMC) simulation techniques. Model performance was assessed using the Watanabe-Akaike Information Criterion (WAIC) and Leave-One-Out Information Criterion (LOOIC), with the model having the lowest values selected as the optimal forecasting model.</p> Results <p>The findings revealed that Ethiopia’s mean annual per capita CO₂ emissions from 1941 to 2022 were approximately 0.054 metric tons. Among the candidate models evaluated, the Bayesian ARIMA (0, 1, 1) model demonstrated the best fit and forecasting performance based on WAIC and LOOIC criteria. Forecast results from the selected model indicate that Ethiopia’s per capita CO₂ emissions are projected to increase gradually, reaching approximately 0.167, 0.169, 0.171, 0.171, 0.175, 0.177, and 0.179 metric tons in 2024, 2025, 2026, 2027, 2028, 2029, and 2030, respectively.</p> Conclusion <p>The study indicates a persistent upward trend in Ethiopia’s annual per capita CO₂ emissions through 2030, which may intensify climate-related and environmental challenges. These findings underscore the need for strengthened environmental policies and sustainable energy interventions, including carbon taxation, cap-and-trade mechanisms, and the promotion of clean and energy-efficient technologies to reduce future emissions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A comparative approach to analyzing and forecasting carbon dioxide emissions in ethiopia using Bayesian autoregressive integrated moving average (ARIMA) and Bayesian structural time series (BSTS) models

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月7日

查看全文
  • Abdisalan Ahmed Osman,
  • Dagne Tesfaye Mengistie,
  • Daud Hussein Adawe,
  • Rediat Takele Figa,
  • Buzuneh Tasfa Marine

摘要

Introduction

Carbon dioxide (CO₂) emissions are a major contributor to global climate change. In Ethiopia, CO₂ emissions have increased steadily in recent decades due to industrialization, rising energy consumption, expanding transportation systems, continued dependence on biomass fuels, and rapid population growth. Although Ethiopia has historically contributed only a small share to global greenhouse gas emissions, the recent upward trend poses challenges for sustainable development and environmental management. Accurate forecasting of CO₂ emissions is essential for designing effective climate mitigation strategies and evidence-based policy decisions. Therefore, this study aimed to analyze and forecast CO₂ emissions in Ethiopia using Bayesian Autoregressive Integrated Moving Average (ARIMA) and Bayesian Structural Time Series (BSTS) models.

Methods

This study analyzed and forecasted carbon dioxide (CO₂) emissions in Ethiopia using an 82-year time series dataset covering the period from 1941 to 2022. Two Bayesian time series models were employed: Bayesian ARIMA and Bayesian Structural Time Series (BSTS). The Bayesian ARIMA model captured temporal dependencies through autoregressive and moving average components, whereas the BSTS model decomposed the time series into trend, seasonality, and regression components, allowing the incorporation of external predictors. Model parameters were estimated using Markov Chain Monte Carlo (MCMC) simulation techniques. Model performance was assessed using the Watanabe-Akaike Information Criterion (WAIC) and Leave-One-Out Information Criterion (LOOIC), with the model having the lowest values selected as the optimal forecasting model.

Results

The findings revealed that Ethiopia’s mean annual per capita CO₂ emissions from 1941 to 2022 were approximately 0.054 metric tons. Among the candidate models evaluated, the Bayesian ARIMA (0, 1, 1) model demonstrated the best fit and forecasting performance based on WAIC and LOOIC criteria. Forecast results from the selected model indicate that Ethiopia’s per capita CO₂ emissions are projected to increase gradually, reaching approximately 0.167, 0.169, 0.171, 0.171, 0.175, 0.177, and 0.179 metric tons in 2024, 2025, 2026, 2027, 2028, 2029, and 2030, respectively.

Conclusion

The study indicates a persistent upward trend in Ethiopia’s annual per capita CO₂ emissions through 2030, which may intensify climate-related and environmental challenges. These findings underscore the need for strengthened environmental policies and sustainable energy interventions, including carbon taxation, cap-and-trade mechanisms, and the promotion of clean and energy-efficient technologies to reduce future emissions.