The Development and Validation of Regression Models to Predict Energy-related CO₂ Emissions in Turkey

Abstract

In this study, the main indicators effecting energy-related CO₂ emissions (ERCDE) in Turkey is determined by correlation analysis and the ERCDE is modeled by multiple linear regression analysis as a function of the various indicators. Five indicator over the period 1971–2010, namely country population (CP), gross domestic product (GDP), alternative and nuclear energy consumption (ANEC), combustible renewables and waste energy consumption (CRWC), fossil fuel consumption (FFC) are used in the analyses. The derived models are validated through statistical tests such as the determination coefficient, t-test, F-test, and residual analysis. Projections are also made for future planning using trend analysis (TA). As a result of the study, CP is determined as the most significant variable explaining the ERCDE in Turkey which is followed by FFC, CRWC, GDP and ANEC. The study reveals that the ERCDE can be effectively predicted as a function of CP, GDP and FFC. Using the TA, the ERCDE is estimated as 335.41Mt for the year 2025, 47.75% higher than 2010.

Keywords:

• CO₂ emissions
• forecasting
• modeling
• Turkey