Industrial clusters with substantial carbon-reduction potential

To successfully reduce environmental emissions, companies need to expand the scope of their emissions accounting to include entire supply chains. A clustering approach has been used to find emission-intensive industry clusters. However, this approach did not include entire direct and indirect supply chains when forming high emission industry clusters. We propose a new method based on a modified normalized cut function with Leontief’s input–output model and basic clustering algorithms to find industry clusters with high levels of embodied within-cluster emissions that are well separated in the supply chain network. We use this method to identify 58 carbon-intensive clusters of Japanese industries and visualize the within-cluster supply chains in terms of embodied carbon flows. We recommend that companies collaborate within clusters to reduce environmental emissions. Our results provide new insights on where to target emissions reduction actions and technology development within industrial supply chains.