Prediksi Emisi Karbon di Asia Tenggara Menggunakan Machine Learning: Implikasi Terhadap Perubahan Iklim dan Kebijakan Mitigasi

Abstract

Southeast Asia plays a strategic role in climate change mitigation; however, the region is highly vulnerable to the impacts of the climate crisis, exacerbated by population density and coastal economic activities. Carbon dioxide (CO₂) emissions, primarily driven by deforestation and industrial sectors, make a significant contribution to climate change, with Indonesia being the largest emitter. Countries in the region have ratified international agreements to reduce greenhouse gas emissions, including Indonesia, which targets net-zero emissions by 2060. This study aims to enhance carbon emission analysis using a machine learning approach, specifically the Random Forest Regressor, to predict carbon emission trends while considering economic and demographic variables. The model is expected to improve prediction accuracy and provide guidance for more effective mitigation policies. Carbon emission projections in Indonesia indicate a significant increase from 2022 to 2032, with emissions expected to exceed 177.42% MtCO₂, driven by urbanization, industrial sector growth, and dependence on fossil fuels. This rise in emissions could exacerbate the accumulation of greenhouse gases, trigger extreme weather events, and threaten economic stability and public health, particularly in the form of respiratory and cardiovascular diseases. It is anticipated that this research will provide valuable insights for designing more sustainable and responsive policies to address the environmental and health challenges facing the region.