Spatiotemporal dynamics and key drivers of carbon emissions in regional construction sectors: Insights from a Random Forest Model

Abstract

Carbon emissions are a substantial global issue, and the construction sector is a significant contributor to this rising trend. This research utilizes the Random Forest Model, a sophisticated machine learning method, to examine the determinants of carbon emissions in China's construction sector at the regional scale. The study highlights environmental impacts by identifying the primary drivers of emissions and promoting cleaner, low-carbon solutions. The model integrates provincial data from China to capture the complex, non-linear relationships between variables such as energy usage, economic activity, and policy initiatives with carbon emission intensity. The findings reveal significant regional disparities, with higher emissions and intensities in southeastern and northeastern provinces due to rapid urbanization and industrial dependency, while central and northwestern regions exhibit lower emissions. Furthermore, the study identifies those characteristics such as urbanization rate, labor productivity, and Gross Domestic Product (GDP) per capita initially contribute to increased emissions but later facilitate reductions as energy efficiency improves. Industrial upgrades, technological innovation, and cleaner energy transitions are essential for reducing emissions in the construction industry while supporting long-term growth. These findings underscore the importance of region-specific carbon reduction policies aligned with China's national targets.