Rapid growth in urban climate science masks structural transferability gaps

Abstract

Urban climate research has expanded rapidly as cities become central to climate mitigation, adaptation, and risk management. However, the transferability of this accumulating knowledge across diverse urban contexts remains unclear. Here, we analyze the global evidence base of urban climate science through a synthesis of ∼8,000 published studies using a hybrid expert–large language model workflow. We examine where research is conducted, the types of contributions it provides, and how these patterns influence the transferability of findings. We find that despite increasing methodological diversity and empirical coverage, the literature exhibits profound structural biases. Crucially, mechanism-oriented research is overwhelmingly limited to high-income regions and specific climates, leaving rapidly urbanizing and climate-vulnerable regions comparatively underrepresented. Consequently, many state-of-the-art frameworks guiding urban climate adaptation are derived from environments fundamentally different from those where future urbanization and risks will be prominent.