Evaluating the time of occurrence of selected global warming levels for CMIP5 and CMIP6 models

This technical note explains the rationale for choosing to work with changes in global mean temperature (also known as "Global Warming Levels"; GWL) rather than defined periods in the future. It documents the methodology used to estimate changes in global mean temperature to enable others to reproduce these results for other models. Finally, it presents the estimates of when specific thresholds will be crossed by different models/ensemble members/emissions scenarios in the North American Coordinated Downscaling Experiment (NA-CORDEX) for reference by American Society of Civil Engineers (ASCE) design code developers. The scenarios considered do not include “low emissions” pathways that do not have a potential to reach high GWL. Additional models associated with the Coupled Model Intercomparison Projects v5 and v6 (CMIP5, CMIP6) are included in the Appendices. ASCE has decided to consider a future GWL of 3°C for structures whose design life will end in 2080-2100, referred to throughout as 3GWL. Given the non-linear relationship between temperature and variables such as snow or freezing rain, this technical note also considers a 2°C, or 2GWL. The warmest models in NA-CORDEX cross 2GWL at the start of the simulation period with the earliest ensemble members crossing between 2017-2036, compared to the ensemble mean of 2022-2041. There is one model-emission combination that does not cross the 2GWL threshold prior to 2100. Several model and emission combinations do not cross the 3GWL threshold prior to 2100, although the warmest models cross the threshold between 2036-2055.