Fast climate impact emulation for global temperature scenarios with the Rapid Impact Model Emulator (RIME)

Edward Byers , Michaela Werning, Mahe Perette, Niklas Schwind, Volker Krey, Keywan Riahi, Carl-Friedrich Schleussner

Climate model emulation has long been applied to assess the global climate outcomes of IAM emissions scenarios, but is typically limited to first-order climate variables like mean surface air temperatures at minimal regional resolution. Here we introduce RIME, the Rapid Impact Model Emulator, which uses global warming level interpolation approaches based on inputs of global mean air temperature pathways to calculate a range of climate impacts and exposure indicators in gridded spatial and region-aggregated formats. The emulation is fast and versatile, moving towards batches of climate impact indicators to complement integrated assessment model scenarios thereby bridging the IPCC WGII and WGIII communities. Our lightweight emulator produces both gridded and regionally-aggregated results taking us beyond the constraints of super-computational global climate and impact models. The approach allows to assess the combined outcome of a wide range of emission and socio-economic scenarios allowing for a decomposition of drivers of uncertainty for future climate risks. While climate uncertainties are the primary concern through mid-century, our results indicates that socio-economic factors such population growth may become the dominant drivers of risk by the end of the century. We demonstrate an application to IPCC scenarios to illustrate potential further use, illustrating its potential utility while acknowledging methodological constraints and delineating a comprehensive roadmap for future development. These rapid climate risk emulation frameworks exhibit significant promise for facilitating cross-disciplinary integration and enhancing scientific inclusivity across diverse research communities.