Key points:
- Under the Paris Agreement, countries committed to limiting global temperature increase to 1.5°C or ‘well below’ 2°C above pre-industrial levels. However, even in an optimistic emissions scenario, the chance of limiting warming to 1.5°C by the end of the century is now virtually zero.
- Our journey to achieving Paris Agreement goals will likely entail ‘temperature overshoot’, where warming exceeds a specified level (typically 1.5°C to 2°C) for 10 years or more before returning to that level in the future.
- Ninety percent of the mitigation pathways in the IPCC’s Global Warming of 1.5°C report that limit warming to 1.5°C by 2100 include temperature overshoot.
- Pathways that limit warming to 1.5°C with limited or no overshoot require deep, rapid, and sustained reductions in emissions, as well as some carbon dioxide removal to compensate for sectors that cannot be decarbonised.
- Even temporary overshoot increases the risk of Earth systems reaching ‘climate tipping points’, catalysing large and often irreversible changes. For example, deforestation in the Amazon could trigger a deforestation-drought feedback loop, rapidly transforming the region into a savanna.
- Every increment of avoided warming can make a difference. Keeping overshoot as short and at as low a temperature as possible is critical to avoid sudden, cascading and irreversible climate impacts.
- Without rapid emissions cuts now, we will need to rely more heavily on large-scale carbon dioxide removal to bring temperatures back down after overshoot, which comes with risks and uncertainties.
- Other proposed large-scale climate interventions, known as geoengineering, fail to address the root causes of climate change and introduce additional severe risks.
- Strategically targeting ‘positive socio-economic tipping points’ – which are conditions that accelerate the deployment of technologies or practices to reach net zero – could help drive decarbonisation and allow us to meet the 1.5°C target.
Surpassing the Paris Agreement limits
Last year, the Earth’s average temperature reached around 1.5°C warmer than pre-industrial times. The average warming for the decade between 2015 and 2024 was 1.24°C above pre-industrial times. The heating is predominantly a result of human activities1Human-cased warming is estimated to have caused 1.36°C of warming in 2024, relative to 1850–1900., primarily the burning of fossil fuels. This warming has had widespread adverse impacts on people and nature, which will worsen as the climate continues to warm.
The Paris Agreement was adopted at COP21 in 2015 by most countries worldwide – countries that are collectively responsible for 98% of human-caused emissions – with the primary goal of avoiding the most devastating impacts of climate change. The agreement aimed to limit “the increase in the global average temperature to well below 2°C above pre-industrial levels” and to pursue efforts “to limit the temperature increase to 1.5°C above pre-industrial levels” by the end of the century. The US, currently responsible for 13% of global emissions, announced its withdrawal from the Agreement in January 2025.
It is now clear that achieving the Paris Agreement goals will mean stricter measures to reduce emissions. Net carbon dioxide emissions will need to fall by 48% by 2030 to keep warming to 1.5°C, according to the latest IPCC assessment, published in 2022. More recent estimates suggest emissions would need to fall >around 50% over the same timeline. The 2024 Emissions Gap Report finds that, even under an optimistic scenario where all NDCs and net zero pledges are met, the chance of limiting warming to 1.5°C by the end of the century is “virtually zero”. The 2023 Emissions Gap Report gave a 14% chance, as, at the time, there were more opportunities to scale down emissions.
Countries had agreed to submit more ambitious national climate plans by 2025 as part of the Paris Agreement. However, as of October 2025, countries have only pledged to reduce their emissions by 1.6 billion tonnes more than in their previous NDCs, leaving a gap of 26.5 billion tonnes to stay within the 1.5°C limit.
The latest stocktake of the global carbon budget – the amount of carbon that can be emitted before we reach 1.5°C of warming – estimates that we will burn through the remaining carbon budget in about six years. Other estimates suggest that we may have only three years left to reduce emissions sufficiently to limit warming to 1.5°C, and that for a 50% chance of keeping warming to 1.5°C, greenhouse gas emissions would have needed to peak before 2025. Yet, emissions continue to rise and national commitments to reducing emissions remain insufficient.
Every fraction of a degree of warming increases the odds of additional, and often extreme, impacts. As greenhouse gases build up in the atmosphere, heatwaves are becoming hotter and more frequent and rainfall is becoming more intense and variable. Consequently, droughts and floods are worsening globally, triggering crop failure, infrastructure damage and humanitarian crises.
As emissions continue to rise, it is increasingly likely that our journey to limiting warming will entail a period of ‘temperature overshoot’. Unprecedented efforts are needed, soon, to limit the amount of overshoot and the impacts of temperature rise.
What is temperature overshoot?
Temperature overshoot is the term used by the Intergovernmental Panel on Climate Change (IPCC) to describe scenarios, or ‘pathways’, in which the Earth exceeds a specified global warming level, typically between 1.5 and 2°C, before returning to that level at some point in the future.
The magnitude (how much the specified level is exceeded) and the duration (how long it is exceeded for) differ across pathways. In most overshoot pathways, the duration of the overshoot is at least one decade and can extend to several decades, while the magnitude reaches up to 0.5°C. However, in ‘high overshoot’ pathways where the temperature overshoots by as much as 0.5°C, it is unlikely that warming could be returned to ‘well below 2°C’ by 2100. If we follow these high-overshoot pathways, we will not meet the goals of the Paris Agreement.
Most IPCC scenarios foresee some degree of temperature overshoot
Pathways with temperature overshoot are not exceptional. In the IPCC’s Global Warming of 1.5°C report, published in October 2018, 90% of the mitigation pathways that limit warming to 1.5°C by 2100 include a period of overshoot.
The pathways that limit warming to 1.5°C with limited or no overshoot require deep, rapid and immediate reductions in emissions, with carbon dioxide removal (CDR) only used to compensate for historical emissions and in sectors where no mitigation measures are available.‘2Limited-overshoot’ pathways overshoot 1.5°C by no more than 0.1°C. In comparison, pathways that delay emissions cuts and overshoot the temperature targets rely more heavily on large-scale CDR to bring global temperatures back down by the end of the century.
Explaining 1.5°C scenarios with and without overshoot
Figure 1 illustrates three hypothetical overshoot scenarios for pathways that aim to achieve the 1.5 °C target. In the upper panel, the grey line represents a scenario with no temperature overshoot. The blue line represents a scenario with a low magnitude and short duration of overshoot. The red line represents a scenario with a high magnitude of temperature overshoot that persists for a long time. In the blue and red scenarios, temperatures return to 1.5°C by the end of the century
The lower panel shows emissions trends in each of the pathways. The scenario represented by the grey line reaches net-zero emissions, at which point temperatures stabilise at 1.5°C, by rapidly reducing emissions and without the need for negative emissions solutions. In contrast, the scenarios represented by the blue and red lines require negative emissions solutions, such as large-scale CDR, to reach net-zero emissions. In the red scenario of high and long-lasting overshoot, negative emissions solutions will need to be deployed far beyond the end of the century.
The greater and longer the overshoot, the more negative emissions solutions will be needed in order to stabilise temperatures at 1.5°C.
Figure 1: Illustration of 1.5°C scenarios with and without overshoot
Source: Reversing climate overshoot, Nature Geoscience 16, 467 (2023).
Increasing overshoot means increasing climate impacts
Every increase in the magnitude and duration of overshoot increases the severity, frequency and duration of climate impacts, such as heatwaves, droughts and floods. In one analysis, the frequency of agricultural drought increases from 24% in a 1.5°C scenario with little to no overshoot to 31% in an overshoot scenario that reaches just below 2°C by the end of the century. In the same scenarios, the frequency of major heatwaves increases from 29% to 44%. Brazil, North Africa, and southern Africa are projected to be the most severely impacted by heatwaves exacerbated by temperature overshoot.
Sticking to the 1.5°C target also brings economic benefits – climate impacts from temperature overshoot will lead to higher mitigation costs and economic losses later in the century.3While higher initial investments are needed to keep temperatures down, this is outweighed by the economic benefits later in the century.
For nature, exceeding a temperature threshold for even a short amount of time may push species beyond their tolerance limits, causing extinction, migration and knock-on effects for entire ecosystems. As different parts of the world warm and cool at different rates, species worldwide will be unevenly exposed to dangerous conditions. Temperature overshoot will be particularly critical for species already living close to their thermal limits, such as those in the tropics.
The IPCC estimates that the area of global land at risk of changing from one ecosystem type to another – such as a forest changing to a grassland – is 50% lower at 1.5°C of warming compared to 2°C. Some of the world’s most important ecosystems, such as the Amazon Basin, the Pantanal and the Coral Triangle, could be irreversibly transformed with just a few years of overshoot.
In the ocean, overshoot is projected to decrease ecosystem habitability for centuries to come. If we do not curtail emissions now, marine ecosystems in the Indo-Pacific, Caribbean and West Africa could experience sudden collapse as early as the 2030s, with knock-on effects for the people who rely on these ecosystems for food and income from tourism.
Every increment of warming counts
2025 has been hot. It is on track to be the second or third warmest year on record. It is also likely to be the second year since pre-industrial times where the average global temperature exceeded 1.5°C. This occurred for the first time in 2024, which was recorded as the warmest year on record since pre-industrial times and broke multiple heat records on land and in the sea. Although a breach of the Paris Agreement target would require average annual temperatures to be above 1.5°C for at least 20 consecutive years, the fact that we are increasingly seeing individual years surpass this threshold implies that we are getting closer.
A ‘safe’ limit to global warming does not exist, and we are already seeing devastating impacts at current warming levels. IPCC scientists urge that 1.5°C shouldn’t be viewed as a guardrail. The target of 1.5°C was chosen as a threshold beyond which the impacts of warming become increasingly intolerable to humans and nature. This assessment was made based on criteria such as food security, extreme weather events, health, biodiversity loss, water supply and economic growth.
But, what is considered an acceptable level of damage from global warming is highly subjective. At just 0.56°C of warming, 506 people died from climate change-induced heat stress in Paris during the summer of 2003. At 0.95°C of warming, wildfires in Australia in 2019-2020 led to the death or displacement of three billion wild animals and caused AUD 4-5 billion in losses to the Australian food system.
While we have not permanently surpassed the 1.5°C target, we are already experiencing significant impacts of global warming. In 2024, climate change was responsible for an additional 41 days of dangerous heat, and extreme weather events worsened by climate change resulted in the death of thousands of people and billions in damages. These represent only a tiny fraction of all impacts resulting from human-caused warming.
We should do everything in our power to keep warming down as much as possible. As emphasised by IPCC scientist Prof. Mark Howden in 2018: “Every half a degree matters. Every year matters. Every choice matters.”
What are climate tipping points?
Overshooting temperature targets, even temporarily, risks the creation of positive feedback loops – self-perpetuating cycles that speed up warming. These positive feedback loops can trigger climate ‘tipping points’, after which an Earth system transforms into an alternative stable state, completely different to its original state. A single tipping point can have cascading and catastrophic effects at regional and global scales.
For example, the Amazon is drying and burning due to human-caused warming, which reduces tree cover. Reduced tree cover, exacerbated by deforestation, causes further reductions in rainfall due to decreased evapotranspiration, whereby moisture is evaporated into the atmosphere from trees. Over time, this deforestation-drought feedback loop could pass a tipping point, transforming the region from a rainforest into a savanna. This would switch it from one of the most important global carbon sinks to a carbon source and could trigger potentially catastrophic changes to global rainfall patterns, impacting agriculture.
The transformation of the Amazon is a fast-onset tipping point, in that when the tipping point is transgressed, the change in the climate system would occur rapidly – in a matter of decades. For these tipping points, overshoot could cause sudden and irreversible changes to the system. Scientists are uncertain when the Amazon’s tipping point might be crossed. Some estimates suggest this could happen at 40% forest loss, and around 26% of the Amazon was already deforested or degraded as of 2022. Some scientists believe that the first warning signals of this shift are already here.
Other tipping points are slow-onset tipping points, whereby the change to an Earth system occurs over a much longer timescale, like over many centuries. For these tipping points, briefly overshooting temperature targets might not cause immediate and irreversible changes to the climate system, as long as the overshoot is not longer than the time needed for the system to recover. One example is the melting of the Greenland ice sheet. While we may be close to this tipping point, it is estimated that ice sheet loss could be mitigated as long as temperatures are brought back down to 1.5°C or lower relatively quickly once the tipping point is reached.Even temporary overshoot increases the risks of surpassing critical tipping points by 72% compared to scenarios with no overshoot. Keeping the magnitude and duration of overshoot as low as possible is critical and rapid decarbonisation is key: While delaying emissions reductions to beyond 2030 could still allow us to meet 1.5 °C by the end of the century, this would result in higher temperature overshoot over many decades, with the potential for adverse consequences.
Boundaries for keeping the Earth habitable
Humans have altered the Earth, such as by depleting the ozone layer, reducing biodiversity, changing land cover and warming the atmosphere. Scientists have tried to estimate how far these processes could continue to be altered until a global tipping point is reached, causing the Earth to transform irreversibly into a state that could endanger humanity and render the planet uninhabitable.
Planetary system boundaries
The concept of planetary system boundaries involves tracking changes to nine crucial processes, identified by scientists as responsible for keeping the Earth habitable (Figure 2). These processes are ranked on a scale from ‘safe operating’, meaning the process happens in a way that is safe for humanity, to ‘high-risk’, i.e. the process poses a high risk to humanity.Of the nine identified planetary boundaries, the Earth is now outside the safe zone for seven: Climate change, biosphere integrity (the quality of living organisms and ecosystems, impacted by, for example, decreasing species diversity), land-use change, biogeochemical flows (for example of nitrogen and phosphorous – aggravated by agribusiness and industry), novel entities (the release of novel chemicals such as plastics), freshwater change and ocean acidification. The safe space for ocean acidification was added as breached in September 2025, reflecting worsening trends.
Figure 2: Planetary system boundaries
Source: Seven of nine planetary boundaries now breached – ocean acidification joins the danger zone, Potsdam Institute for Climate Impact Research (2025)
There is great scientific uncertainty regarding how much longer we can continue to push these boundaries before the total collapse of the Earth’s system happens. However, if we rapidly decarbonise, we could reduce this risk and stabilise the Earth within a safe operating space.
Earth system boundaries
Scientists have also proposed a set of ‘safe and just’ Earth system boundaries that quantify the safety of humans and the stability of the planet (Figure 3). Safe boundaries are those “where biophysical stability of the Earth system is maintained and enhanced over time, thereby safeguarding its functions and ability to support humans and all other living organisms”. Just system boundaries minimise the exposure of countries, communities and people to significant harm, including “loss of lives, livelihoods or incomes; displacement; loss of food, water or nutritional security; and chronic disease, injury or malnutrition”.
This framework differs from previous frameworks in that the impacts on people are measured in comparable units to impacts on the planet. While other frameworks only assess how human activities have impacted Earth systems, using comparable units allows for a better understanding of the harm that changes to Earth system boundaries will do to humans. The framework focuses on all species, not just humans, attempting to “define the environmental conditions needed not only for the planet to remain stable, but to enable societies, economies and ecosystems across the globe to thrive”. The framework also incorporates information on climate, the biosphere, and other Earth system tipping points into the Earth system boundaries
Figure 3: Safe and just Earth system boundaries
Source: Safe and Just Earth System Boundaries published in Nature, Global Commons Alliance (2023).
According to this framework, the ‘just’ Earth system boundary for climate is 1°C, while the ‘safe’ boundary is 1.5 °C. The Earth has already heated by more than 1.2 °C, meaning current global warming, while still in the ‘safe’ zone, is unjust (Figure 3), emphasising the need for urgent action. As the framework incorporates interspecies justice, intergenerational justice and intragenerational justice (including race, class and gender), it can be used to inform sustainability targets and practices.
However, these frameworks have been criticised, with some suggesting that they are too simplistic or that they do not distinguish between thresholds, that can be breached, and hard limits, that cannot be breached. They may also shift political focus to the wrong areas or dampen political action. Others warn against allowing one group of scientists to define what constitutes a safe set of boundaries for everyone on the planet, which could be viewed as divisive.
Relying on carbon removal is risky – emissions cuts must come first
To bring average global temperatures back down to between 1.5 and 2°C, overshoot scenarios rely to different extents on CDR. This includes nature-based solutions, such as afforestation (planting forests) and bioenergy with carbon capture and storage (BECCS). Other solutions include direct air capture (DAC) and storage, where CO2 is directly captured from the air and then stored for the long term.
However, there are major risks and uncertainties with these approaches. With nature-based CDR and BECCS, there is a risk that ecosystems could be replaced with bioenergy crops or plantations, endangering wild plants and animals. Food crops could be supplanted, threatening food security. DAC technology is still in its infancy. These technologies are also not proven to be effective at the scale needed.
These uncertainties and risks emphasise that climate action needs to be boosted in the near-term to reduce our reliance on these approaches for bringing temperatures back down later in the century.
Geoengineering cannot replace rapid and deep emissions cuts
Rapid and deep emissions cuts provide the best chances of limiting temperature overshoot and avoiding tipping points. At the same time, approaches that remove carbon from the atmosphere can help to offset emissions that cannot be reduced and mitigate overshoot.
Technologies that aim to intervene with Earth systems at a large scale to counteract the effects of a warming climate, referred to as geoengineering, have also been increasingly proposed in recent years.4CDR is also defined as a geoengineering approach by the IPCC. A prominent solution is solar radiation management (SRM), which aims to reduce the amount of sunlight absorbed by the Earth by reflecting it back into space or preventing it from reaching the Earth’s surface altogether. This would result in reduced warming, but would have no impact on greenhouse gases that are being released into the atmosphere or that are already there, meaning it does not tackle the underlying causes of climate change and instead introduces new and additional risks.
Research remains in early and theoretical stages, with many unknowns and potential unintended consequences. Studies suggest that SRM would have significant impacts on water cycles, could modify monsoon systems, and could lead to drought in some tropical regions. SRM could also impact renewable energy generation, which is crucial for mitigating emissions. Additionally, as its deployment would impact the climate of the entire planet, SRM comes with substantial geopolitical risks.
SRM approaches would have to be implemented continuously until emissions are reduced to safe levels to prevent warming. If they are stopped before this happens, this would result in very rapid warming, and “severely stress ecosystem and human adaptation”, according to the IPCC. The IPCC is very clear that, even in best-case scenarios, SRM is a supplement to rapid and deep emissions cuts.
Positive tipping points can help us reach net zero
In contrast to climate tipping points, positive tipping points – otherwise known as positive socio-economic tipping points – occur when a set of conditions is reached that can accelerate the deployment of technologies or practices to achieve net zero. For instance, a new technology begins to outcompete an old technology. Sales of the new technology facilitate further development, which in turn reduces its costs, allowing the new technology to become widespread and replace the old. An example of this is the rapid development of renewable energy, where over just 10 years, solar and wind technologies have become the cheapest source of power in many parts of the world.
Targeted interventions in socioeconomic, technological and political systems can be used to advance climate change mitigation, and strategic investment can help bring down the costs of technologies that facilitate decarbonisation. For example, oil and gas companies have been accused of having overly optimistic projections of long-term oil prices, resulting in an inflated picture of future economic performance. Updating the accounting standards or disclosure guidelines for these companies could cause prices to decrease, thereby curbing investment and catalysing investment into renewables.
The focus on these positive tipping points should not distract us from the need to rapidly decrease emissions. Rather, these positive tipping points should be viewed in the context of driving strategic interventions to encourage decarbonisation.
This briefing was originally published in December 2023. This updated version was published in October 2025.
