Item

Abstract

Taking stock of climate change evidence is essential to helping cities address climate change. However, such efforts face challenges in appraising the growing scholarship in this fast-moving area. Here we use supervised and unsupervised machine learning to identify and classify over 53,000 urban climate studies, creating a dynamic, interactive and searchable evidence database for researchers and policymakers. Nearly 20,000 are city-specific case studies, revealing a rapidly growing yet unevenly distributed knowledge base. Notably, small and fast-growing cities, particularly in Africa and Asia, remain substantially underrepresented, contributing to topical, geographic and disciplinary biases in previous Intergovernmental Panel on Climate Change (IPCC) assessments. We propose three strategies to address this: (1) synthesizing case studies to support IPCC uptake, (2) identifying cross-city learning opportunities and (3) closing evidence gaps in the Global South. Thereby, our systematic stocktake helps inform adaptation and mitigation efforts in cities, guides future research and strengthens the IPCC’s ability to deliver robust, policy-relevant evidence.