Uluslararası İklim Anlaşmaları Bağlamında Enerji Dönüşümünün Hibrit Tabanlı Dinamik Kümeleme Analizi[Hybrid Based Dynamic Clustering Analysis of Energy Transformation in the Context of International Climate Agreements]

The study examines the effects of international climate agreements on countries' energy profiles from a time series perspective. Over 200 countries were analysed across three distinct phases during the 1990–2024 period: pre-Kyoto (1990–1996), post-Kyoto (1997–2014), and post-Paris (2015–2024). This study aims to examine the dynamic effects of the Kyoto and Paris Agreements on energy transformation. In the first stage, similarities and differences between countries' energy profiles were analysed using a time series clustering method based on dynamic time warping (DTW). Subsequently, the extent to which the cluster structures obtained in different time periods showed consistency was evaluated using the adjusted Rand index (ARI). Finally, the key energy indicators distinguishing the clusters were determined using the random forest algorithm. The findings indicate that fossil intensity was decisive before Kyoto, while hydropower came to the fore during the Kyoto period. In the Paris period, low-carbon and renewable indicators became dominant, demonstrating the differentiating effects of the agreements. Furthermore, it is understood that, in contrast to Kyoto's limited effects, Paris offers a more inclusive and sustainable framework. This study contributes an original method to the literature that combines dynamic analysis, stability measurement and explainability dimensions. Furthermore, this method provides an objective data basis for policymakers to monitor countries' compliance with agreements and identify priority indicators.