Our future global climate

The natural greenhouse effect is the warming caused by naturally occurring greenhouse gases, such as water vapour and carbon dioxide, which absorb outgoing longwave radiation from the Earth's surface and re-radiate some of it back down. This keeps the planet about 33°C warmer than it would otherwise be, making it habitable, so the natural effect is essential for life.

The enhanced greenhouse effect is the additional warming caused by humans adding extra greenhouse gases to the atmosphere — mainly by burning fossil fuels, clearing forests and farming. The higher concentration traps even more outgoing heat, so the surface warms further.

The key difference is therefore that the natural effect is essential and keeps Earth habitable, while the enhanced effect is the human-caused extra warming that drives climate change — already about +1.1°C above the pre-industrial average.

Carbon dioxide (CO₂) is released mainly by the burning of fossil fuels such as coal, oil and gas for energy, industry and transport, and also by deforestation. It is the largest contributor by quantity and is very long-lived in the atmosphere.

Methane (CH₄) comes mainly from livestock such as cattle, from flooded rice paddy fields and from decomposing waste in landfill sites. It is far more potent per molecule than CO₂ but is shorter-lived.

Nitrous oxide (N₂O) is released mainly from the use of nitrogen fertilisers in agriculture, as well as from some industrial processes. It is a powerful and long-lived greenhouse gas.

These three gases differ in their sources, their potency and how long they remain in the atmosphere, which is why reaching net zero requires cutting emissions across energy, agriculture and waste rather than a single fix.

Net zero means that the amount of greenhouse gases added to the atmosphere is balanced by the amount removed from it, so the net addition is zero. Removal can occur through natural sinks such as forests and oceans, or through technology that captures and stores carbon. It does not mean zero emissions, but rather that any remaining emissions are offset by an equal amount of removal.

Reaching net zero is important because the climate only stops warming further once net emissions reach zero. Until that point, greenhouse gas concentrations in the atmosphere keep rising, so warming continues. Stabilising the climate therefore requires not just slowing emissions but balancing them with removals.

The timing also matters. Because carbon dioxide is very long-lived, the total amount of warming depends on the cumulative emissions before net zero is reached. Reaching net zero earlier — as in the 'very low' and 'low' projection categories — limits the total warming, keeping it closer to +1.4°C by 2100, whereas categories that never reach net zero this century lead to much greater warming, up to around +4.4°C.

Finally, even after net zero is achieved, some impacts such as sea-level rise continue for centuries because of the longevity of greenhouse gases, so reaching net zero limits but does not instantly reverse climate change.

Future climate change is non-linear because it does not follow a simple straight line; feedback loops and tipping points can speed it up or cause sudden shifts.

A positive feedback loop amplifies the original change. For example, in the ice–albedo feedback, warming melts bright, reflective ice and snow and exposes darker ocean or land. The darker surface has a lower albedo, so it absorbs more incoming sunlight instead of reflecting it, which causes more warming and melts still more ice. The loop reinforces itself, so warming accelerates rather than rising steadily. The permafrost–methane feedback works the same way: thawing permafrost releases methane and carbon dioxide, which cause more warming and more thawing.

A tipping point is a threshold beyond which part of the climate system shifts into a new state that is self-sustaining and often irreversible, even if the original forcing is removed. Examples include the collapse of an ice sheet, large-scale permafrost melt, or a shutdown of global ocean circulation. Once a tipping point is crossed, change can become rapid and locked in.

Together, positive feedbacks and tipping points mean the climate can respond abruptly and unpredictably rather than gradually, which is a major reason future projections are uncertain and why the 'very high' outcomes are especially dangerous.

A climate projection is a model-based estimate of how the global climate will change in the future under a given assumption about greenhouse gas emissions. Such projections, produced using General Circulation Models and summarised by the IPCC as five categories from 'very low' to 'very high', do contain real uncertainty. This essay weighs that uncertainty against the usefulness of the projections and argues that, while projections are genuinely uncertain, the uncertainty is understood, quantified and decision-relevant, so they remain useful and the statement is only partly correct.

There is a strong case that projections are uncertain. The largest source is emission, or scenario, uncertainty: future warming depends on how much greenhouse gas humans choose to emit, which cannot be known in advance. This is why projected warming by 2100 spans a wide range, from about +1.4°C in the 'very low' category to about +4.4°C in the 'very high' category. A second source is model uncertainty, including the exact climate sensitivity — how much the climate warms for a given rise in carbon dioxide — and how clouds, oceans and ice respond, which models represent imperfectly. A third and most serious source is the behaviour of feedback loops and tipping points, such as ice-sheet collapse or an ocean-circulation shutdown, which are non-linear and could push change beyond the central model ranges. Taken together, these mean no single confident figure for 2100 can be given.

However, it does not follow that projections are too uncertain to be useful. First, the uncertainty is quantified rather than unknown: projections are presented as ranges and as five clearly defined categories, with shaded bands showing the spread of results, so users know how confident to be. Second, the direction of change is unambiguous across all scenarios — every projection shows warming, rising sea levels and more frequent extremes — so the broad outcome is not in doubt, only its magnitude. Third, the models are validated against observed change and are regularly reviewed and updated by the IPCC in a transparent cycle, which improves confidence over time. Because of this, projections are already used effectively to set policy, such as net-zero targets and adaptation planning.

The decisive point is that the different kinds of uncertainty should be judged differently. Most of the 2100 spread is scenario uncertainty, which is not model failure but a reflection of human choice — and that makes the range informative, because it shows exactly how much difference our emissions decisions make. The narrower physical uncertainty within each scenario does not change the overall picture. And the uncertainty from tipping points is an argument for precaution and early action, not for ignoring the projections: if anything, it means the central estimates may understate the risk. So uncertainty, properly understood, strengthens rather than undermines the case for using projections.

In conclusion, I agree only to a limited extent that climate projections are too uncertain to be useful. They are genuinely uncertain, especially because future emissions and tipping points cannot be precisely predicted. But the uncertainty is bounded, quantified and dominated by a human choice we can influence, while the direction of change is clear and the models are validated and updated. Projections therefore remain highly useful: they reliably bound the likely range, identify emissions as the key lever, and justify acting early. Far from being too uncertain to use, they are among the most decision-relevant tools we have.

Future climate change, projected to raise global mean temperature by between about +1.4°C and +4.4°C above the pre-industrial average by 2100, will affect every environment, but not equally. This essay assesses whether the impacts will be greater for cold environments — high-latitude polar regions and high-altitude mountains — than for marine environments. It argues that cold environments will experience the most rapid and visible physical change, but that marine environments will suffer the most far-reaching human and global consequences, so 'greater' depends on whether magnitude of physical change or breadth of impact is being judged.

The case that cold environments are most affected is strong in terms of the speed and scale of physical change. High-latitude and high-altitude regions warm faster than the global average, a process known as polar amplification, so temperatures rise most here. The consequences are dramatic: snow and ice cover shrink, glaciers retreat and sea ice declines, transforming entire landscapes. Critically, thawing permafrost releases stored methane and carbon dioxide, which not only damages local ground stability and ecosystems but feeds back to amplify global warming. Because these changes are fast, large and partly self-reinforcing, the physical impact on cold environments is arguably the greatest of any environment.

However, the impacts on marine environments are wider-reaching and affect far more people, which is the stronger measure of 'impact'. Sea-level rise, driven by thermal expansion of warming water and the melting of land ice, causes coastal flooding that threatens densely populated deltas, low-lying small islands and major coastal cities — putting hundreds of millions of people and vast economic assets at risk. Warming sea-surface temperatures cause coral bleaching, which devastates the ecosystems that support fisheries and coastal protection, and they provide more energy for intense tropical cyclones. Changes to global ocean circulation, as meltwater freshens and warms the surface, could disrupt regional climates far beyond the oceans themselves. These impacts therefore extend globally and have profound human consequences.

The most accurate assessment recognises that the two environments are not separate but deeply interconnected, so ranking them in isolation is misleading. The melting of cold-environment land ice is itself a primary driver of marine sea-level rise; the freshening of the ocean from polar meltwater is what weakens global ocean circulation; and permafrost methane amplifies the warming that drives marine change. In this sense, the severe physical impacts in cold environments are the engine of the most damaging marine impacts. The interaction means that the impacts cascade from one environment into the other.

In conclusion, the impacts of future climate change are greatest in cold environments when measured by the speed and magnitude of physical change, because polar amplification, ice loss and permafrost melt are fast and self-reinforcing. But when measured by the breadth of consequences for people and the global system, marine impacts — especially sea-level rise and coastal flooding — are more serious. Since the two are linked, with cold-environment change driving marine change, the most defensible judgement is that impacts are not greater in one than the other in any simple sense; rather, the most severe outcomes arise from their interaction, with cold environments acting as the trigger and marine environments as the channel through which the largest human impacts are felt.