Spatiotemporal clustering of GHGs emissions in Europe: Exploring the role of spatial component

In this study, we propose a novel application of spatiotemporal clustering in the socio-economic and environmental sciences, with a particular focus on regionalized time series of greenhouse gases (GHGs) emissions from a range of economic sectors. Utilizing a hierarchical spatiotemporal clustering methodology, we analyze yearly time series of emissions by gases and sectors from 1990 to 2022 for European regions at the NUTS-2 level. While the clustering algorithm inherently incorporates spatial information based on geographical distance, the extent to which space contributes to the definition of groups still requires further exploration. To address this gap in the literature, we propose a novel indicator, namely the Joint Inertia, which quantifies the contribution of spatial distances when integrated with other features. Through a simulation experiment, we explore the relationship between the Joint Inertia and the relevance of geography in exploiting the groups structure under several configurations of spatial and features patterns, providing insights into the behavior and potential of the proposed indicator. The empirical findings demonstrate the relevance of the spatial component in identifying emission patterns and dynamics, and the results reveal significant heterogeneity across clusters in trends and dynamics by gases and sectors. This reflects the heterogeneous economic and industrial characteristics of European regions. The study highlights the importance of the spatial and temporal dimensions in understanding GHGs emissions, offering baseline insights for future spatiotemporal modeling and supporting more targeted and regionally informed environmental policies.