A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Climate sensitivity is the global temperature rise per unit radiative forcing (°C per W/m² or °C per doubling CO₂), determined by radiative feedback processes. Equilibrium climate sensitivity (ECS, temperature change at CO₂ doubling after full equilibration) is ~3°C (range 2.5–4°C); transient climate response (warming before ocean adjustment) is ~1.8°C. Radiative feedbacks—quantified as feedback parameters (W/m²/°C)—describe how the climate system responds: water vapor feedback is positive (warming increases atmospheric water vapor, trapping more heat); cloud feedback is uncertain; ice-albedo feedback is positive; lapse-rate feedback is negative. The net of all feedbacks determines sensitivity.
Run an energy balance model with different feedback strengths and observe how equilibrium temperature responds to a forcing change. Decompose climate model warming into forcing and feedback contributions using radiative kernels.
Climate sensitivity is not fixed; it can vary with forcing magnitude, background climate state, and spatial patterns of warming. Also, positive feedbacks do not lead to runaway warming if the radiative forcing is finite; the system reaches equilibrium.
When scientists say that doubling atmospheric CO₂ will warm the Earth by roughly 3°C, they are invoking the concept of climate sensitivity — a single number that summarizes how strongly the climate system responds to a radiative perturbation. Understanding where that number comes from requires understanding feedbacks: the secondary responses of the climate system that amplify or dampen the initial warming.
Start with a simple energy balance. The Sun delivers a certain amount of energy to Earth's surface; Earth must radiate the same amount back to space to stay at a stable temperature. CO₂ acts as a partial blanket, reducing the efficiency of outgoing infrared radiation. If you double CO₂ overnight, the Earth initially absorbs more energy than it emits (a radiative imbalance of roughly +3.7 W/m²). The surface warms until outgoing radiation increases enough to restore balance. Without any feedbacks, this warming would be about 1°C — the "Planck response." The actual sensitivity of ~3°C means feedbacks amplify this by a factor of roughly 3.
The most important feedbacks are: water vapor (strongly positive — warmer air holds more water vapor, which is itself a greenhouse gas), ice-albedo (positive — warming melts reflective ice, exposing darker ocean and land that absorb more sunlight), lapse-rate (negative in the tropics — the upper troposphere warms faster than the surface, increasing outgoing radiation), and clouds (uncertain — low clouds cool by reflecting sunlight, high clouds warm by trapping infrared; their net response to warming is the largest source of uncertainty in climate sensitivity estimates). Each feedback is quantified as a feedback parameter λᵢ (W/m²/°C); the net climate sensitivity parameter λ = Σλᵢ determines how much warming a given forcing produces.
Equilibrium climate sensitivity (ECS) is the warming after the entire climate system — including the deep ocean — reaches a new steady state. Because the ocean has immense heat capacity, this takes centuries. Transient climate response (TCR) is the more policy-relevant quantity: it measures warming at the moment of CO₂ doubling in a scenario where CO₂ increases 1% per year. TCR is smaller than ECS because the ocean is still absorbing heat and suppressing surface warming. The difference — ECS minus TCR — represents the "committed warming" that would occur even if emissions stopped today.
A critical misconception is that positive feedbacks imply runaway warming. They do not. Runaway would require the feedback gain to exceed 1 — meaning the feedback amplification exceeds the restoring force. Earth's current sensitivity does not meet this condition for realistic CO₂ scenarios. What positive feedbacks do is increase the equilibrium temperature for a given forcing. Knowing the sign, magnitude, and uncertainty of each feedback is therefore the central goal of climate sensitivity research, and the spread in ECS estimates (2.5–4°C) largely reflects disagreement about cloud feedbacks.