Equilibrium Climate Sensitivity¶
The long-term global mean surface warming after atmospheric CO2 doubles from a preindustrial baseline and the planetary energy budget rebalances, excluding ice-sheet feedbacks.
Core Idea¶
Equilibrium climate sensitivity (ECS) is the long-term change in global mean near-surface air temperature after atmospheric carbon dioxide (CO2) is held at twice its preindustrial reference concentration and the planetary energy budget rebalances. Under the IPCC Sixth Assessment's convention, it excludes feedbacks associated with changing ice sheets. It is a temperature-difference quantity, not one universal value shared by all models or a reading of today's atmosphere.[ref-c7819d357d10][ref-cf79948abd09]
A long-run model can estimate its rebalanced doubled-CO2 response. Paleoclimate evidence can constrain Earth's ECS after translating a mixed-forcing climate contrast into this target. A short-run energy-balance regression gives effective sensitivity, which can differ from eventual equilibrium; warming at CO2 doubling during a concentration ramp is transient climate response (TCR), a separate metric.[ref-cf79948abd09][ref-7d267453a17a][ref-00336d5c6198][ref-c7819d357d10]
Scope of Application¶
Rugenstein and colleagues analyzed millennia-long coupled-model runs under sustained elevated CO2. Their CCSM3 abrupt2x experiment ran 3,000 years, with a best equilibrium-warming estimate of 2.57 K; its first-150-year regression estimated 2.35 K. These are model-specific numbers, and the difference shows why an early effective estimate cannot silently replace the long-run target.[^ref-cf79948abd09]
Schmittner and colleagues combined Last Glacial Maximum land and sea temperature proxies with climate-model simulations. The LGM was near equilibrium but included lower greenhouse gases, larger ice sheets and more dust. Their doubled-CO2 ECS inference had a conditional median 2.3 K and 66% range of 1.7–2.6 K, which the authors say can widen under alternate assumptions or data. The raw glacial cooling is not itself ECS, and this historical paper's interval is not a universal assessed value.[ref-7d267453a17a][ref-c7819d357d10]
Clarity¶
For an ECS number, state the preindustrial baseline, sustained CO2 doubling, global surface-air temperature response, equilibrium endpoint and feedback convention. Then say how the number was obtained. Rugenstein's long run estimates a model endpoint; Schmittner's glacial analysis infers the same defined target from proxy, model and forcing assumptions.[ref-cf79948abd09][ref-7d267453a17a]
Do not call a 150-year abrupt-4xCO2 regression a direct equilibrium observation. Zelinka and colleagues use top-of-atmosphere imbalance against surface warming to calculate an effective doubled-CO2 sensitivity and note that evolving feedbacks can make it smaller than eventual equilibrium warming.[^ref-00336d5c6198]
Manages Complexity¶
ECS condenses an interacting climate response into one standardized long-run temperature difference. CO2 forcing, ocean heat uptake, clouds, water vapour and albedo influence its value. The standard target makes distinct evidence sources comparable only after their estimation limits are carried along. A model result remains model-dependent; an LGM inference remains conditional on proxy and forcing assumptions.[ref-c7819d357d10][ref-cf79948abd09][^ref-7d267453a17a]
Abstract Reasoning¶
Start with the preindustrial climate. Hold CO2 at twice that concentration. Identify the global mean surface-air temperature after incoming and outgoing planetary energy fluxes have rebalanced, and subtract the reference temperature under the declared exclusion of ice-sheet feedbacks. When this endpoint is estimated rather than attained, identify the estimation method and uncertainty.[ref-c7819d357d10][ref-cf79948abd09]
If the ocean still absorbs heat at CO2 doubling during a 1%-per-year ramp, the defined quantity is TCR. If the input is raw LGM cooling, non-CO2 forcing and changed ice sheets still need treatment before it informs conventional ECS. Doubling CO2 from another baseline without translation is a state-dependent experimental sensitivity, not automatically this entry's preindustrial-reference quantity.[ref-c7819d357d10][ref-7d267453a17a]
Knowledge Transfer¶
The reusable map is preindustrial reference → sustained doubled CO2 → energy rebalance → global mean temperature difference under a declared feedback scope. A millennial model experiment and a paleoclimate inference use different evidence for this target. A short effective-ECS extrapolation is useful evidence but should retain its label and time horizon.[ref-cf79948abd09][ref-7d267453a17a][^ref-00336d5c6198]
The live Equilibrium Prime supplies a strict prerequisite: the final energy-budget balance. ECS is not itself a balanced state; it is the temperature difference that presupposes one. Measurement procedures, climate ensembles and individual feedback processes have other roles.
Example¶
Long-run model response. In the CCSM3 abrupt2x run reported by Rugenstein and colleagues: reference → model preindustrial control; perturbation → sustained doubled CO2; response → global mean surface-air warming; endpoint → long-run best estimate 2.57 K after a 3,000-year simulation; evidence boundary → a model-specific estimate rather than a directly observed Earth value. The first-150-year 2.35 K result is an effective extrapolation, not the same endpoint.[^ref-cf79948abd09]
LGM inference. Schmittner and colleagues: reference → preindustrial Earth; perturbation evidence → LGM changes in greenhouse gases, ice sheets and dust; response evidence → land/sea proxy temperatures; mapping → model-based, assumption-bound inference to a doubled-CO2 ECS distribution with median 2.3 K. Ice-sheet and non-CO2 forcing treatment distinguishes this from raw glacial cooling or broader Earth-system sensitivity.[ref-7d267453a17a][ref-c7819d357d10]
Relationships to Other Abstractions¶
Current abstraction Equilibrium Climate Sensitivity Domain-specific
Parents (1) — more general patterns this builds on
-
Equilibrium Climate Sensitivity presupposes Equilibrium Prime
The doubled-CO2 temperature endpoint presupposes a planetary state with balanced energy fluxes.
Hierarchy path (1) — routes to 1 parentless root
- Equilibrium Climate Sensitivity → Equilibrium → Fixed Point
Neighborhood in Abstraction Space¶
Equilibrium Climate Sensitivity sits in a sparse region of the domain-specific corpus (98th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Unclustered & Miscellaneous (2551 abstractions)
Nearest neighbors
- Cloud feedback — 0.78
- Brunt–Väisälä Frequency — 0.77
- Reference Evapotranspiration Estimation — 0.76
- Albedo — 0.76
- Diffusive–Thermal Instability — 0.76
Computed from structural-signature embeddings · 2026-10-08
Not to Be Confused With¶
TCR: warming at CO2 doubling during a ramp while the system is out of equilibrium. Effective ECS: a finite-run extrapolation that may differ from eventual rebalanced warming. Climate sensitivity parameter: temperature response per unit forcing, requiring normalization before comparison. Earth-system sensitivity: may include slow ice-sheet feedbacks excluded here. Raw LGM cooling: a mixed-forcing contrast, not ECS by itself. Other-baseline doubled-CO2 sensitivity: a different reference-state experiment. Equilibrium: the prerequisite balanced state, not this scalar difference. Mathematical Metric: a distance function, not this climate usage.[ref-c7819d357d10][ref-cf79948abd09][ref-7d267453a17a][ref-00336d5c6198]
References¶
[^ref-c7819d357d10]: Intergovernmental Panel on Climate Change, “The Earth's Energy Budget, Climate Feedbacks, and Climate Sensitivity”, in Climate Change 2021, The Physical Science Basis, Working Group I, Chapter 7, DOI 10.1017/9781009157896.009, Box 7.1 and §§7.5, 7.5.3 and 7.5.5. Official assessment; definitions of ECS and TCR, effective-ECS conversion and paleoclimate forcing/ice-sheet limits.
[^ref-cf79948abd09]: M. Rugenstein, J. Bloch-Johnson, J. Gregory, T. Andrews and T. Mauritsen, “Equilibrium Climate Sensitivity Estimated by Equilibrating Climate Models”, Geophysical Research Letters 47 (2020), DOI 10.1029/2019GL083898, §§1–3 and Table 1. Original full article; the quoted long-run and early-regression values belong to specified model experiments.
[^ref-7d267453a17a]: A. Schmittner, N. M. Urban, J. D. Shakun et al., “Climate Sensitivity Estimated from Temperature Reconstructions of the Last Glacial Maximum”, Science 334 (2011): 1385–1388, DOI 10.1126/science.1203513, abstract, pp. 1385–1386 and Fig. 3. Original article; its reported posterior depends on model, proxy and forcing assumptions.
[^ref-00336d5c6198]: M. D. Zelinka, T. A. Myers, D. T. McCoy et al., “Causes of Higher Climate Sensitivity in CMIP6 Models”, Geophysical Research Letters 47 (2020), DOI 10.1029/2019GL085782, §§1–2. Original full article; 150-year abrupt-4xCO2 regression diagnoses effective, rather than fully attained equilibrium, sensitivity.