Pierre Friedlingstein ^1,2* Paulo Artaxo ³ Angela V. Gallego Sala ¹ Gensuo Jia ⁴ Chris Jones ⁵ Michio Kawamiya ^6,7 Julie Loisel ⁸ Marie-France Loutre ⁹ Kira Rehfeld ¹⁰ Alessio Rovere ^11,12 Chris Smith ^13,14,5 Roland Séférian ¹⁵ Narelle Van Der Wel ¹⁶ Elisa Ziegler ¹⁰

• ¹ University of Exeter, Exeter, United Kingdom
• ² Laboratoire de Météorologie Dynamique, Institut Pierre-Simon Laplace, CNRS, École Normale Supérieure, Université PSL, Sorbonne Université, École Polytechnique, Paris, France
• ³ Center for Amazonian Sustainability, University of São Paulo, São Paulo, Brazil
• ⁴ Institute of Atmospheric Physics, Chinese Academy of Sciences (CAS), Beijing, Beijing Municipality, China
• ⁵ Met Office Hadley Centre (MOHC), Exeter, England, United Kingdom
• ⁶ Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Yokosuka, Kanagawa, Japan
• ⁷ Advanced Institute for Marine Ecosystem Change, Tohoku University, Sendai, Japan
• ⁸ Department of Geography, University of Nevada Reno, Reno, United States
• ⁹ Past Global Changes (PAGES), Bern, Switzerland
• ¹⁰ Geo- and Environmental Research Center, University of Tübingen, Tubingen, Germany
• ¹¹ Ca' Foscari University of Venice, Venice, Veneto, Italy
• ¹² Marum, Center for Marine Environmental Sciences, University of Bremen, Bremen, Germany
• ¹³ School of Earth and Environment, University of Leeds, Leeds, United Kingdom
• ¹⁴ International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria
• ¹⁵ CNRM, Université de Toulouse, Météo-France, CNRS, Toulouse, France
• ¹⁶ World Climate Research Programme Secretariat, WMO, Geneva, Switzerland

Anthropogenic carbon dioxide emissions are the main driver of climate change, with global warming increasing almost linearly with cumulative CO2 emissions. Hence, future warming will primarily result from future emissions of CO2 with contributions from other greenhouse gases and aerosols. Climate projections of the 21 st century, such as those assessed by the IPCC, are provided from comprehensive Earth System models, driven by scenarios of the evolution of emissions from those climate forcers. While it seems now inevitable that the world will reach 1.5°C of warming above preindustrial levels by the early 2030s, the extent to which we exceed this warming level depends strongly on global activity taken now to limit emissions. In this paper, we review the current understanding on Earth system changes under two highly contrasted possible future worlds. We first focus on high-end scenarios, where anthropogenic emissions continue to increase in the 21 st century, leading to large warming levels and associated impacts on all components of the Earth System. We then assess low-end scenarios, where anthropogenic emissions rapidly decline, reaching net zero and potentially becoming net negative before the end of the 21 st century. Such "overshoot" scenarios lead to a peak in global warming followed by a slow decline in global temperature, with some degree of reversibility in key Earth system components. We also review paleoclimatic information relevant to these contrasting future worlds, giving evidence that stabilizing feedbacks operate on millennial or longer timescales, whereas destabilizing feedbacks and tipping cascades also occurred on shorter timescales.