Climate change will not reverse itself on any timeline that matters to people alive today. Even in the most optimistic scenarios, where humanity reaches net-zero carbon dioxide emissions, global temperatures are expected to stabilize rather than fall, and that stabilization takes roughly a century. The question is less whether the planet “gets better” and more whether we can limit how much worse things become, and the range of possible futures is still enormous depending on the speed and scale of emission cuts.
What Happens After Emissions Stop
There is a widespread assumption that once we stop pumping carbon dioxide into the atmosphere, the planet will start cooling off. The reality is more stubborn. The concept scientists use to describe this is the Zero Emissions Commitment: the amount of additional warming or cooling expected after all human CO₂ emissions cease. The best estimate from the IPCC’s most recent major assessment is that this commitment is approximately zero, meaning temperatures would roughly hold steady, not drop, after we stop emitting.
Holding steady sounds manageable until you consider what “steady” means in practice. A modeling study published in Earth’s Future found that even after reaching net-zero CO₂ emissions, global average temperature took a median of about 90 years to stabilize, with some models showing it took more than 300 years. Several models showed cooling after emissions stopped, but two of those then reversed and warmed again. One model warmed continuously through the entire simulation. At the local level, many regions continued experiencing changing conditions well beyond the point when the global average had settled.
This means the warming we have already caused, and whatever additional warming occurs before we reach net zero, is essentially permanent on any human planning horizon. The carbon dioxide already in the atmosphere will keep trapping heat for centuries. The oceans, which have absorbed enormous amounts of energy, will continue releasing that heat slowly. There is no thermostat we can turn down.
The Overshoot Trap
Most pathways that keep warming near 1.5°C assume something called “overshoot”: temperatures temporarily exceed the target, and then large-scale carbon dioxide removal pulls them back down. On paper, this looks like a plan. In practice, the evidence is increasingly troubling. A study in Nature showed that global and regional climate change after an overshoot looks fundamentally different from a world that avoids overshooting in the first place.
The damage from temporary overshoot does not neatly undo itself. Research modeling various overshoot scenarios found that high-latitude ocean surface temperatures remain substantially elevated long after temperatures are brought back down. In a scenario where warming overshoots to 3°C before returning to 1.5°C, North Atlantic surface waters stayed up to 3.1°C warmer than in a scenario that stabilized at 1.5°C without overshooting, and Southern Ocean temperatures stayed about 1.2°C warmer. Sea levels remained up to 32% higher on centennial timescales because excess heat stored in the deep ocean continued expanding the water column.
Water systems face similar lasting damage. A study in Water Resources Research found that overshoot pathways carry qualitatively different risks than gradual warming trajectories, and those risks are likely underestimated in current research and policy. Glacial meltwater contributions to rivers, groundwater systems, and vegetation-driven changes in the water cycle all have decadal-scale memory, meaning they do not snap back when temperatures come down.
The carbon removal technology needed to pull off overshoot reversal also remains largely theoretical at the scales required. A policy analysis estimated that a carbon dioxide removal enterprise large enough to matter for climate would require financial flows of billions to trillions of dollars per year and would need to operate for roughly a century before being phased out. No such enterprise exists, and building one quickly enough to rescue an overshoot scenario is an enormous gamble.
Tipping Points That Feed on Themselves
Beyond the direct effects of human emissions, several components of the Earth system can become self-reinforcing sources of warming once certain thresholds are crossed. Permafrost, the permanently frozen ground covering vast areas of the Arctic and high-altitude plateaus, contains enormous stores of carbon accumulated over thousands of years. As it thaws, microbes break down that organic material and release carbon dioxide and methane.
Research published in Nature Communications found that permafrost loses 11 to 21 billion kilograms of carbon per 100 degree-years of cumulative warming exposure, and this relationship is roughly linear: more warming means proportionally more carbon release. On the Tibetan Plateau, a separate study documented that the area susceptible to abrupt permafrost collapse is projected to expand by 17 to 19 percent by 2100 relative to 2022. In zones where the ground has already collapsed, methane release increased by 20 percent compared to intact areas. Under a middle-of-the-road emissions scenario, carbon emissions from these collapse-prone areas are projected to surge 2.7-fold by the end of the century.
Then there is the Atlantic Meridional Overturning Circulation, the massive ocean current system that carries warm water northward and influences weather patterns across Europe, Africa, and the Americas. One statistical analysis estimated that this circulation could collapse around mid-century under current emissions trajectories. A more recent study in Nature Climate Change complicated the picture by showing that the risk depends not just on how much warming occurs but on how fast. Under slow CO₂ increases, the circulation remained stable up to 5.5°C of warming. Under faster increases, it collapsed at just 2°C. The speed of emissions, not just their total, determines the danger.
The practical takeaway is that certain parts of the climate system have their own momentum. Once permafrost begins releasing carbon at scale, or once a major ocean circulation weakens past a critical point, those processes continue regardless of what humans do with their own emissions. They add warming on top of warming.
The Paradox of Cleaning the Air
Here is something counterintuitive: some of the actions we take to reduce pollution actually accelerate warming in the short term. Sulfur-containing aerosols from burning fossil fuels reflect sunlight and brighten clouds, producing a cooling effect that partially masks the full warming from greenhouse gases. When we clean up those aerosols, the mask comes off.
This played out in a real-world experiment starting in 2020, when international shipping regulations cut the sulfur content of marine fuel by about 80 percent. The regulation was a public health win, reducing the sulfur dioxide that causes acid rain and respiratory disease. But it also removed a significant source of atmospheric particles that had been reflecting solar energy. One study estimated the resulting radiative forcing at about +0.2 watts per square meter averaged over the global ocean, which could double the rate of warming during the 2020s compared to the rate since 1980. A separate model-based analysis estimated the forcing at +0.12 watts per square meter and found that Northern Hemisphere temperature anomalies in 2022 and 2023 correlated with the areas where cloud-brightening effects from shipping had been strongest. A third study placed the figure at 0.13 watts per square meter and noted this is equivalent to roughly half the net positive forcing from all other anthropogenic aerosol reductions since the late twentieth century.
The studies do not fully agree on the magnitude, but they agree on the direction and the mechanism. Cleaning up air pollution is the right thing to do for human health. It also means we lose a temporary, accidental cooling effect and see the full force of the greenhouse warming we have already committed to. This is not an argument against pollution controls. It is a reminder that the warming “in the pipeline” is larger than what thermometers have shown so far.
Extreme Weather Is Already Getting Worse
For most people, climate change is not an abstract temperature number. It is the drought that kills a harvest, the flood that destroys a neighborhood, the heat wave that fills emergency rooms. These events are intensifying. Research on Central Asia found that the probability of experiencing an extreme temperature event (a +7°C anomaly) increased by up to a factor of seven in some areas due to global warming, with observational data showing more frequent and prolonged extreme heat events than hypothetical scenarios without warming.
Increasingly, the concern is not just individual extremes but compound events: a heat wave followed immediately by a flood, or drought and wildfire striking the same region in the same season. A review in Dialogues on Climate Change found that increases in the frequency, intensity, and duration of extreme weather are raising the likelihood of these compound and cascading events, harming human health, ecosystems, and communities today, with larger impacts projected unless there are significant investments in adaptation and mitigation.
The health consequences of compound events are worse than what you would expect from adding the effects of each event separately. A scoping review of the medical literature found that across 27 studies comparing compound extremes to individual events, 20 found that the compound events carried elevated health risks. Among the small number of studies that formally tested for synergistic effects, more than half found that the combined risk was greater than the sum of its parts.
Human Heat Tolerance Has a Lower Ceiling Than We Thought
For years, the theoretical upper limit of human survival in heat was pegged at a wet-bulb temperature of 35°C. Below that threshold, the human body can theoretically cool itself through sweating. Above it, core temperature rises uncontrollably. But laboratory research on young, healthy adults performing basic activities found that no subject reached a critical wet-bulb temperature of 35°C, and all mean values were significantly lower than the theoretical threshold. In humid conditions, the actual limit at which the body can no longer regulate its temperature is well below 35°C.
This matters because conditions exceeding these lower, more realistic thresholds are already occurring. A study using a physiology-based model found that non-survivable heat stress conditions have been reached during present-day heat events, all at wet-bulb temperatures below 35°C. Older people are especially vulnerable, with regular exceedances of deadly thresholds during recent events. And extremely hot but dry conditions proved just as deadly as hot and humid ones, broadening the geography of danger beyond tropical regions.
These findings mean the safe zone for outdoor human activity is narrower than climate models have traditionally assumed. Regions in South Asia, the Persian Gulf, and parts of Africa and South America are approaching or have already touched these limits during peak summer events. Further warming compresses the margin further.
Where We Are Headed in Deep Time
One way to grasp the scale of projected change is to look at what past climates resembled when the planet was as warm as we are making it. A study comparing future climate projections to geological benchmarks found that under a high-emissions scenario, by 2030 the climate most closely resembles the mid-Pliocene, roughly 3 million years ago, when sea levels were substantially higher than today and forests grew at high latitudes. By 2150, the closest analog shifts to the early Eocene, about 50 million years ago, when there was no permanent ice on Earth and crocodile relatives lived above the Arctic Circle.
These are not exact analogs. A separate study found that the mid-Pliocene climate cannot serve as a straightforward model for future conditions in all respects, because the response of Northern Hemisphere winter weather patterns to elevated CO₂ is actually the opposite of what happened during the Pliocene due to different continental configurations and ice sheet boundaries. The comparison still has value for understanding broad temperature ranges and ecosystem responses, but it breaks down for regional weather patterns.
The point is not to predict a specific future but to give a sense of the magnitude. The climate system is heading toward states that have not existed during all of human civilization, and is doing so at a rate far faster than any natural transition in the geological record.
Temporary Carbon Removal and Buying Time
Not all the news is bleak. Research into carbon dioxide removal is turning up some pragmatic strategies, even if the grand vision of pulling billions of tons from the atmosphere remains distant. One approach targets short-lived climate pollutants like methane, which traps far more heat per molecule than CO₂ but breaks down in the atmosphere within about a decade. A study in Nature showed that temporary carbon storage can serve as compensation for these short-lived pollutants. Offsetting 1 kilogram of methane requires about 101 kilograms of CO₂ stored for 100 years in something like durable wood products, or about 498 kilograms stored for just 20 years in something like bioplastics.
This is not a solution to the full climate problem, but it offers a way to address methane’s outsized near-term warming contribution using storage that does not need to last forever. Methane comes from livestock, rice paddies, landfills, and fossil fuel extraction. Cutting it rapidly could slow the rate of warming within a couple of decades, buying time for the harder work of decarbonizing electricity, transport, and industry.
Why Believing It Can Work Actually Matters
There is a psychological dimension to whether climate change “gets better” that shapes policy and collective action. Research in risk analysis found that people who believe action to reduce climate risks is both possible and effective are more likely to support mitigation efforts, even after controlling for political ideology and beliefs about whether climate change is human-caused. Both personal confidence in one’s own ability to contribute and belief that collective and government action can make a difference independently predicted stronger support for climate policy.
This creates a feedback loop, and it can run in either direction. Despair and fatalism reduce political pressure, which slows policy, which makes outcomes worse, which deepens despair. Conversely, tangible progress builds confidence in further action. The rapid cost declines in solar and wind energy over the past decade, the adoption of electric vehicles faster than most projections anticipated, and the spread of net-zero commitments by major economies are all real developments that shift this psychology. Whether they shift it fast enough is the open question, and the answer depends less on physics than on politics.
The Aerosol Problem Is Also a Geoengineering Warning
The unintended experiment with shipping fuel regulations offers a preview of a larger debate. Solar radiation management, the deliberate injection of reflective particles into the stratosphere to cool the planet, has been proposed as a stopgap while emissions are reduced. But the shipping experience illustrates a fundamental vulnerability: if you start reflecting sunlight and then stop, the masked warming arrives all at once. The concept is sometimes called termination shock. With shipping aerosols, the effect was modest in global terms but measurable. With deliberate, large-scale geoengineering, the stakes would be vastly higher. Any interruption, whether from geopolitical conflict, funding collapse, or technical failure, would unleash decades of suppressed warming in a matter of years. The acknowledged primary risks of solar radiation management, including termination shock, geopolitical conflict over who controls the thermostat, and the moral hazard of reducing pressure to cut emissions, cannot be adequately assessed using smooth scenario models. They require thinking about shocks, disruptions, and the messy realities of international cooperation over timescales longer than any treaty has ever held.