Global-mean surface temperature is shown as an anomaly relative to the
1850–1900 pre-industrial baseline. Everything left of the present
is one shared record, stitched from two sources: a paleoclimate reanalysis of the
deglaciation (Osman et al. 2021) running from the Last Glacial Maximum to the
mid-19th century, and the instrumental HadCRUT5 record
(1850–2025).
Right of the present are the futures, and there are two sets to choose between.
CLIMBER-X gives three long pathways run to the year 3000 with an
Earth system model of intermediate complexity: Sustainable stewardship
(P1, deep and fast emission cuts), Middle of the road (P2), and a
Hostile Earth system (P3 — the same emissions as P2 but a higher climate
sensitivity). These are the Planetary Health Check's own pathways, and the only
ones here that reach the far end of the timeline.
CMIP7 gives the seven ScenarioMIP marker scenarios — the set the
world's climate models are now being run with — from Very Low through to High.
Each is defined by the question it was built to answer rather than by a forecast
it makes: High, for instance, is a deliberate rollback of climate policy, not a
business-as-usual baseline; Medium takes the policies actually in force in 2025
and freezes them. Three are drawn by default (the coolest, the benchmark and the
hottest) and all seven on request. The temperatures are not model output but an
emulation: FaIR v2.2 driven by each scenario's harmonised
emissions across an 841-member, observation-constrained ensemble, of which the
median is drawn and the 5th–95th percentile range is shaded behind the selected
scenario. The CMIP7 Earth system model runs themselves are expected from late 2026.
The CMIP7 scenarios end at 2500 and the CLIMBER-X pathways at
3000, but the timeline never moves: it always runs to 3000, so
switching between the two changes what is drawn, never the axis. A dashed rule
marks where a set runs out. Past 2100 none of these are really scenarios any more
— the integrated assessment models stop there, and everything after is an
extension protocol that brings emissions to zero and holds them, so the last
centuries show seven states relaxing under one shared rule rather than seven
choices. Very Low is the exception, and the reason it is still cooling at the far
right: it keeps removing carbon past 2400.
The colour scale behind the curves and in the stripe panel above is the same
throughout: deep blue for the glacial cold, greens around the pre-industrial
baseline, warm reds through the hotter futures, and violet-black at the top of the
range, which the CMIP7 High marker reaches around 2450.
References
-
Osman, M. B., Tierney, J. E., Zhu, J. et al. (2021). Globally resolved surface
temperatures since the Last Glacial Maximum. Nature 599, 239–244.
doi:10.1038/s41586-021-03984-4
-
Morice, C. P., Kennedy, J. J., Rayner, N. A. et al. (2021). An updated assessment
of near-surface temperature change from 1850: the HadCRUT5 data set.
JGR Atmospheres 126, e2019JD032361.
doi:10.1029/2019JD032361
-
Willeit, M., Ganopolski, A., Robinson, A. & Edwards, N. R. (2022). The Earth
system model CLIMBER-X v1.0. Geoscientific Model Development 15,
5905–5948.
doi:10.5194/gmd-15-5905-2022
-
Riahi, K., van Vuuren, D. P., Kriegler, E. et al. (2017). The Shared
Socioeconomic Pathways and their energy, land use, and greenhouse gas emissions
implications. Global Environmental Change 42, 153–168.
doi:10.1016/j.gloenvcha.2016.05.009
-
van Vuuren, D. P., Sanderson, B. M., O'Neill, B. C. et al. (2026). The Scenario
Model Intercomparison Project for CMIP7. Geoscientific Model Development
19, 2627–2656.
doi:10.5194/gmd-19-2627-2026
-
Leach, N. J., Jenkins, S., Nicholls, Z. et al. (2021). FaIRv2.0.0: a generalized
impulse response model for climate uncertainty and future scenario exploration.
Geoscientific Model Development 14, 3007–3036.
doi:10.5194/gmd-14-3007-2021
-
Rockström, J. et al. (2026). CLIMBER-X long-term climate pathways to the year
3000 [Data set]. Zenodo.
zenodo.org/records/19696770