Probabilistic simple climate model projections in support of ScenarioMIP scenario development

Pre-print (really, pre-pre-print) of the forthcoming article. This version only contains MAGICC output (this can be double checked against the published CMIP7 ScenarioMIP climate assessment, 10.5281/zenodo.19825038) and a draft of a first figure. To track progress on the manuscript, please see github.com/znicholls/CMIP7-ScenarioMIP-Simple-Climate-Model-Projections. This has not been reviewed by the core author team or a scientific journal yet so use with caution. Notes on the figure: a, b) Median (line) and 5th to 95th percentile range (shading) from MAGICC from 2000 to 2100 and from 2100 to 2500 (note the different y-axes). a) also shows the observed and human-induced warming assessed in Indicators of Global Climate Change 2025 (IGCC; Forster et al., 2026). c, d) Year of peak warming (to 2500), on the same time axes as a) and b). Median is shown in white. Inner box is 33rd to 67th percentile, next box is 17th to 83rd, outer box is 5th to 95th percentile. e) Rate of warming (linear trend over ten years, as in IGCC) for each scenario from MAGICC over 2016–2025, 2026–2035, 2031–2040, 2036–2045 and 2041–2050. In 2016–2025, IGCC’s assessed rate of human-induced warming is also shown (grey; white line: best estimate, box: likely range, 5th to 95th percentile after rounding in IGCC). The mistmatch between MAGICC and IGCC should be noted and will be discussed in more detail in the forthcoming paper. The boxes are the same as in c), d) in terms of percentiles they represent. f–h) Peak warming (to 2500), warming in 2100 and warming in 2500. (Repeating) In the scenarios’ boxes in c–h), the outer box shows the 5th to 95th percentile range, the inner boxes the 17th to 83rd and 33rd to 67th percentile ranges and the white line the median. In each of f–h), M and H are shown in their own panel, on their own scale. IGCC’s values are global mean surface temperature, ours are global surface air temperature (the distinction will be discussed in more detail in the forthcoming paper).