Skip to content

GCRI · Climate and Hazards · Foundations

Climate projections: the foundations of exploring future climate

Climate projections show how the climate would respond to different paths of emissions and land use, and they now inform plans and design standards that will last for decades. This foundational guide explains scenarios, model ensembles, downscaling and how to read projection ranges, with the latest IPCC assessment and the new CMIP7 scenario set.

По состоянию на
28 October 2026
Чтение
15 minutes
Доказательная база
15 sources, cited in order
Метод
A synthesis of the IPCC Sixth Assessment Report, the published designs of the CMIP6 and CMIP7 scenario experiments, WMO decadal predictions and public downscaled projection products, following a projection from scenario to local information.
Данные
Assessed warming ranges come from the IPCC Working Group I Summary for Policymakers; model output is archived through the Coupled Model Intercomparison Project; downscaled products are published by NASA, the Met Office and the regional downscaling programme of the World Climate Research Programme; each is available from its custodian.
NEXUS TWIN · WF-1 · SURFACESEED 885c7f57G-P02.CLIMATE-PROJECTIONS.FOUNDATIONSxyzSCHEMATIC · SURFACEST-01 · 0.665 0.368
WF-1 · surface Schematic wireframe twin (surface, catastrophe model). An illustration, not data.

Методы и доказательная база

Вопрос
How are climate projections produced from scenarios, models and downscaling, and how should their ranges be read and used?
Подход
A synthesis of the IPCC Sixth Assessment Report, the published designs of the CMIP6 and CMIP7 scenario experiments, WMO decadal predictions and public downscaled projection products, following a projection from scenario to local information.
Доказательная база
15 sources, cited in order and listed under References
Data availability
Assessed warming ranges come from the IPCC Working Group I Summary for Policymakers; model output is archived through the Coupled Model Intercomparison Project; downscaled products are published by NASA, the Met Office and the regional downscaling programme of the World Climate Research Programme; each is available from its custodian.
Ограничения
Projections are conditional on scenarios that do not cover every possible future, models differ in sensitivity and regional patterns, downscaling inherits the errors of its driving model, and the next IPCC assessment and the CMIP7 simulations are not yet available.

01Определение

What climate projections are

Conditional simulations of the future climate, and how they differ from forecasts and predictions.

Climate projections are simulations of the response of the climate system to scenarios of future emissions, concentrations and land use, produced with climate models. They answer a conditional question: if emissions follow this path, how would temperature, rainfall, sea level and extremes change? The Intergovernmental Panel on Climate Change (IPCC) assessed projections for five illustrative scenarios that start in 2015 and span very low to very high greenhouse gas emissions 1.

Projections are not forecasts. A weather forecast predicts a specific state days ahead from today's observations. A decadal prediction, such as the World Meteorological Organization (WMO) annual to decadal update, gives probabilities for the next five years 6. A projection over decades depends mainly on the scenario, so no single year in it should be read as a prediction.

Every projection carries a reference period. The IPCC states global warming relative to 1850–1900 and notes that changes relative to the recent 1995–2014 period can be approximated by subtracting 0.85 degrees Celsius, the best estimate of warming between the two 1. The platform's climate baselines topic explains those reference periods.

This article sets out how scenarios, models and downscaling combine into projections, how to read their ranges and what is changing in the next generation. It belongs to the Climate and Hazards platform's climate projections topic, and its hub is the platform's foundational guide, Climate science and natural hazards: the foundations of a discipline, reachable from the Climate and Hazards platform.

02Stakes

Why climate projections matter now

The near term is already near 1.5 degrees, and long-lived decisions need numbers for the decades ahead.

The near-term future is largely set. In the IPCC Working Group I assessment, the best estimate of warming for 2021–2040 is 1.5 degrees Celsius under four of the five scenarios and 1.6 degrees under the highest, with very likely ranges starting at 1.2 degrees, or 1.3 under the highest 1. The Synthesis Report finds 1.5 degrees more likely than not to be reached in the near term even under very low emissions 7.

Observations are tracking that path. WMO places 2025 at about 1.43 degrees Celsius above 1850–1900 8, and the WMO annual to decadal update gives a 91 per cent chance that at least one year between 2026 and 2030 temporarily exceeds 1.5 degrees, and an 86 per cent chance that one of those years beats 2024 as the warmest on record 6. In the ERA5 reanalysis reported in the Copernicus Global Climate Highlights 2025, 2023 to 2025 averaged 1.52 degrees above pre-industrial, the first three-year average above 1.5 degrees in the instrumental period 9. These are observations and short-range predictions, not the 20-year averages that define a warming level, but they show the projected near term arriving on schedule.

The long term depends on choices. By 2081–2100 the very likely range is 1.0 to 1.8 degrees under the very low scenario and 3.3 to 5.7 degrees under the very high one 1. Bridges, drainage, power networks and land-use plans built now will still be in service then, so planners need to know what each path implies for the hazards they design against.

Standards are starting to use projections directly. The next United States national precipitation frequency standard, NOAA Atlas 15, will include a volume that projects estimates into the future using downscaled climate model information 10. How projections enter such standards is a scientific question about method and a governance question about who decides.

03Сценарии

Scenarios: the questions projections ask

How emission scenarios are named and built, and the new scenario set for the next model generation.

A scenario is a coherent description of how emissions, land use and air pollution might evolve. The IPCC labels its scenarios SSPx-y, where SSPx is the Shared Socio-economic Pathway describing the underlying social and economic trends and y is the approximate radiative forcing in watts per square metre reached in 2100 1. The five illustrative scenarios run from SSP1-1.9 and SSP1-2.6, with carbon dioxide emissions falling to net zero around or after 2050, through SSP2-4.5, with emissions near current levels until mid-century, to SSP3-7.0 and SSP5-8.5, with emissions roughly doubling by 2100 and 2050 1.

The IPCC is neutral about the assumptions behind the scenarios and notes that they do not cover all possible futures 1. They were designed for the sixth phase of the Coupled Model Intercomparison Project, whose scenario activity, ScenarioMIP, set out eight 21st-century scenarios and a large initial-condition ensemble, divided into two tiers by priority 2.

The next generation reframes the set. The ScenarioMIP design for CMIP7, published in April 2026, proposes seven marker scenarios named by their emission trend: High, High-to-Low, Medium, Medium-to-Low, Low, Very Low and Low-to-Negative 3. Most are intended to run in emission-driven mode for carbon dioxide, so that the carbon cycle responds inside the model, and the set is expected to span roughly 1.5 to 3.5 degrees Celsius of warming in 2100 3.

Choosing a scenario is choosing a question. A stress test of critical infrastructure may use a high scenario to explore high-end risk, while a check of whether a policy target is consistent with physics uses a low one. The CMIP7 design states this purpose openly, with high scenarios included to explore high-end climate risks 3.

04Модели

Models and ensembles

The model generations behind projections, how they are tested and why their spread is constrained.

Climate projections come from Earth system models that represent the atmosphere, ocean, land, ice and, increasingly, the carbon cycle. Modelling centres coordinate their experiments through the Coupled Model Intercomparison Project: its sixth phase made a common set of baseline and historical simulations the entry requirement for participating models and organised 21 endorsed intercomparison projects around them 11.

Models are tested against the past. The IPCC reports that the CMIP6 historical simulations have an ensemble-mean global surface temperature change within 0.2 degrees Celsius of observations over most of the historical period, and that the models improved the simulation of the recent mean state of most large-scale indicators, while differences remain, for example in regional precipitation patterns 1.

Not every model is equally plausible. CMIP6 models show a wider range of climate sensitivity than earlier models and a higher average, traced to an amplifying cloud feedback about 20 per cent larger 1. The IPCC's own assessment, from multiple lines of evidence, puts equilibrium climate sensitivity at a best estimate of 3 degrees Celsius with a likely range of 2.5 to 4 degrees 1. For the first time, the assessed projections of global temperature, ocean warming and sea level combined model results with observational constraints rather than taking the raw model average 1.

The practical lesson is to use ensembles with care. An unweighted average of all available models can be pulled by a few high-sensitivity members; assessed ranges, or results organised by global warming level, are more robust for most uses.

05Неопределённость

Reading the ranges: where uncertainty comes from

Scenario, model and natural variability, and the rules for reading a projected range.

Three sources of uncertainty shape every projection. Scenario uncertainty is about which emissions path the world follows and dominates late in the century. Model uncertainty is about how different models respond to the same forcing. Internal variability is the climate's own year-to-year and decade-to-decade fluctuation: the IPCC finds that decadal variability has both enhanced and masked human-caused change in the historical record and will continue to do so 1.

The balance shifts with scale. Over a small region and a few decades, internal variability can amplify or dampen the human-caused signal, and near-term cooling at a particular place would still be consistent with global warming 1. For precipitation, near-term regional changes are also uncertain because of model differences and uncertainty in aerosol forcing 1. Modellers separate these sources by design. The design of ScenarioMIP, the scenario activity of CMIP6, consists of eight alternative 21st century scenarios plus one large initial condition ensemble, and it addresses the relative contributions to uncertainty from scenarios, climate models and internal variability 2.

External shocks add to this. The IPCC finds it likely that at least one large explosive volcanic eruption will occur during the 21st century, which would cool the surface for one to three years and temporarily mask part of the human-caused warming 1. Projections do not include such eruptions on a schedule, because their timing cannot be predicted.

Two reading rules follow. First, warming levels are judged on 20-year averages, so a single year above 1.5 degrees does not mean the 1.5-degree level has been reached 1. Second, the likelihood language matters: in the IPCC's calibrated terms, 'very likely' means a 90 to 100 per cent probability 1.

06Downscaling

From global to local: downscaling and warming levels

How coarse global projections become local information, and the trade-offs of each method.

Global models run on grids too coarse for many decisions, so their output is downscaled. The World Climate Research Programme's Coordinated Regional Climate Downscaling Experiment advances and coordinates regional climate downscaling, lists 14 regions, from Region 1 (South America) to Region 14 (South-East Asia), and aims to evaluate and improve regional climate downscaling models and techniques 12.

Statistical downscaling and bias correction relate model output to local observations. NASA's NEX-GDDP-CMIP6 dataset provides global, bias-corrected daily projections derived from CMIP6 models at 0.25 degrees, about 25 km, for the four first-tier scenarios 4. National services combine both approaches: the Met Office's UK Climate Projections run at 60 km globally and 12 km over the United Kingdom, and are further downscaled to 2.2 km to simulate local high-impact events such as summer heavy rainfall 5.

Finer grids are not automatically more certain. Downscaling adds local detail, but it inherits the large-scale errors of the driving model, and bias correction assumes that the relation between model and observations stays fixed as the climate changes. Users should treat a local projection as one member of a family that includes the global ensemble behind it.

Global warming levels offer a complementary frame. The IPCC finds that robust geographical patterns of many variables can be identified at a given level of warming, common to all scenarios and independent of when that level is reached 1. Every additional increment of warming brings larger changes in extremes 1. The numbers are large. An extreme heat event that occurred once in 50 years in the 1850–1900 climate is projected to occur about 8.6 times as often at 1.5 degrees of warming, 13.9 times at 2 degrees and 39.2 times at 4 degrees, while a once-in-10-years heavy rainfall day becomes about 1.5, 1.7 and 2.7 times as frequent at the same levels 1. This is why the platform's heat and drought, flood and storm and compound hazards topics report hazard change by warming level where possible.

07Доказательная база

How projection evidence is produced and checked

Model archives, assessed ranges and downscaled datasets, with their custodians and limits.

Projection evidence is layered. Model output from coordinated experiments is archived and shared under common data standards, which the CMIP6 design made a condition of participation 11. Assessments then combine the models with observations and other lines of evidence into ranges with calibrated likelihoods 1. Downscaled and bias-corrected products turn those into regional and local fields 4,12.

Checking a projection means checking each layer. Is the scenario appropriate to the question? Does the model ensemble reproduce the relevant past behaviour, such as the regional rainfall pattern? Is the downscaling method suited to the variable, for example extremes rather than averages? The platform's methods and uncertainty page and the reviewed models and scenarios list record how the platform answers these questions for the products it uses.

Projections cover more than temperature, and each variable has its own baseline. The IPCC projects global mean sea level rise by 2100, relative to 1995–2014, of 0.28 to 0.55 metres under the very low scenario and 0.63 to 1.01 metres under the very high scenario (likely ranges), and notes that sea level will keep rising for centuries 1. The table below gives the assessed global warming ranges for the five scenarios, the reference values against which downscaled and sectoral projections can be checked.

Сценарий2021–20402041–20602081–2100
SSP1-1.9 (very low)1.5 (1.2 to 1.7)1.6 (1.2 to 2.0)1.4 (1.0 to 1.8)
SSP1-2.6 (low)1.5 (1.2 to 1.8)1.7 (1.3 to 2.2)1.8 (1.3 to 2.4)
SSP2-4.5 (intermediate)1.5 (1.2 to 1.8)2.0 (1.6 to 2.5)2.7 (2.1 to 3.5)
SSP3-7.0 (high)1.5 (1.2 to 1.8)2.1 (1.7 to 2.6)3.6 (2.8 to 4.6)
SSP5-8.5 (very high)1.6 (1.3 to 1.9)2.4 (1.9 to 3.0)4.4 (3.3 to 5.7)
Table 1. Assessed change in global surface temperature relative to 1850–1900, 20-year averages, best estimate and very likely range in degrees Celsius, IPCC Sixth Assessment Report 1.

08Инструменты

Conventions and instruments for using projections

The assessment conventions, the assessment cycle and the standards that apply projections, as of 28 October 2026.

The IPCC sets the reference conventions: warming relative to 1850–1900, 20-year averages to judge warming levels, the SSPx-y naming of scenarios and a calibrated likelihood language 1. Its Sixth Assessment Report remains the latest assessment. The seventh cycle will produce three working group reports and a Synthesis Report by late 2029, but at its March 2026 session the Panel made no formal decision on the working group timeline 13.

Coordinated experiments are the second instrument. CMIP6 and its scenario activity defined what models ran for the last assessment 11,2, and the CMIP7 scenario design now sets the experiments for the next 3.

Standards translate projections into practice. ISO 14091:2021 gives guidelines for assessing the risks related to the potential impacts of climate change, and describes how to understand vulnerability; it can be used for assessing both present and future climate change risks 14, and national design standards such as NOAA Atlas 15 build projected change into engineering values 10. How governments decide which scenarios to plan for is covered by the Global Risks Forum's adaptation institutions topic.

09Передний край

What is changing in climate projections

Four areas where projection science is moving.

Emission-driven projections. Running most CMIP7 scenarios with emissions rather than prescribed concentrations 3 brings carbon-cycle uncertainty into the projection itself.

Digital twins of the climate. The European Destination Earth initiative has the European Centre for Medium-Range Weather Forecasts (ECMWF) implementing its first two digital twins, one of them a Climate Change Adaptation twin that provides multidecadal simulations 15.

Constrained and warming-level projections. In the IPCC's sixth assessment, for the first time, assessed future changes in global surface temperature, ocean warming and sea level were constructed by combining multi-model projections with observational constraints, and robust projected geographical patterns of many variables can be identified at a given level of global warming 1.

Overshoot pathways. Scenarios that exceed 1.5 degrees and then return, such as Low-to-Negative in the CMIP7 set 3, raise new questions about which changes reverse and which do not.

10Консорциум

Three faces of climate projections across the Consortium

Science, governance and finance use the same projections for different decisions.

A projection is used by a scientist to test understanding, by a government to set a planning standard and by a financial institution to test a portfolio. Within the Consortium, The Global Centre for Risk and Innovation (GCRI) leads on science, technology and evidence: how projections are produced, evaluated and documented, with support from the platform's climate science for hazards and computing for climate and hazard science pages. Nexus Foundry is operated at national and regional level by Nexus Consortiums through their host and lead organisations, within a multilateral programme led by these three pillar not-for-profit institutions on a public-private-planet partnership.

The Global Risks Forum (GRF) leads on governance, leadership and cooperation. Its Climate and Nature Governance platform covers how institutions adopt projections in plans and standards, and its Diplomacy platform covers the international cooperation, including the assessment process, that projections feed.

The Global Risks Alliance (GRA) translates projections into the questions finance asks, such as how scenario analysis uses climate pathways to test exposures. Work moves between the three in a cycle: a question is stated, evidence assembled, analysis built and tested, findings published in Reports, competent authorities decide, and lessons reset the next cycle.

What this article does not claim: it explains climate projections and the public evidence on it. It is not a forecast, a warning or advice for any decision; national meteorological, hydrological and geological services are the authorities for warnings. Projections are scenario results, not predictions for a place or year, and the ranges quoted are those assessed by their issuers. GCRI does not regulate or certify, GRF does not regulate or finance, and GRA does not lend, underwrite or rate. Naming an organisation, framework or dataset is not a claim of partnership, endorsement or deployment.

Вопросы

Вопросы и ответы

How are climate projections made?

Modelling centres run Earth system models under common scenarios of emissions and land use through the Coupled Model Intercomparison Project 11. Assessments such as the IPCC's combine those runs with observations into ranges 1, and downscaling turns global output into regional and local detail 12.

What is the difference between a climate projection and a forecast?

A forecast predicts a specific future state from today's observations, such as tomorrow's weather or the next five years' temperature probabilities 6. A projection describes how the climate would respond to a scenario of emissions over decades, so it is conditional on that scenario and does not predict a particular year.

What do SSP scenario names such as SSP2-4.5 mean?

The first part names the Shared Socio-economic Pathway that describes the social and economic trends behind the scenario. The number after the dash is the approximate radiative forcing, in watts per square metre, reached in 2100 1. SSP2-4.5 is the intermediate scenario, with best-estimate warming of 2.7 degrees Celsius by 2081–2100 1.

Why do climate models disagree?

Models represent processes such as clouds differently, so they respond differently to the same forcing. CMIP6 models show a wider and higher range of climate sensitivity than the IPCC's assessed very likely range of 2 to 5 degrees Celsius, and a higher average than its best estimate of 3 degrees, which is why the IPCC combined models with observational constraints 1.

Has the world already passed 1.5 degrees of warming?

Not in the sense the IPCC uses, which is a 20-year average. Single years can exceed 1.5 degrees without the level being reached 1; WMO gives a 91 per cent chance that at least one year from 2026 to 2030 will do so temporarily 6.

Условия

Глоссарий

Climate projection
A simulated response of the climate system to a scenario of future emissions, concentrations or land use.
Shared Socio-economic Pathway
A storyline of future social and economic development used to build emission and land-use scenarios.
Equilibrium climate sensitivity
The long-term global warming that follows a doubling of atmospheric carbon dioxide once the climate has adjusted.
Downscaling
Methods that derive finer-scale climate information from coarse global model output, using regional models or statistical relations.
Global warming level
A level of global surface temperature increase above 1850–1900, such as 1.5 or 2 degrees Celsius, used to organise projected changes.

Reuse and cite

Text CC BY 4.0. Cite the page with its date; data tables are linked under each figure.

Источники

Источники

15 sources, numbered in the order they are first cited
  1. Intergovernmental Panel on Climate Change (2021-08-09). Climate Change 2021: The Physical Science Basis. Summary for Policymakers (Working Group I contribution to the Sixth Assessment Report). www.ipcc.ch. Accessed 2026-10-28.
  2. O'Neill et al., Geoscientific Model Development (European Geosciences Union) (2016-09-28). The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6. Geoscientific Model Development 9, 3461–3482. doi.org. Accessed 2026-10-28.
  3. Van Vuuren et al., Geoscientific Model Development (European Geosciences Union) (2026-04-07). The Scenario Model Intercomparison Project for CMIP7 (ScenarioMIP-CMIP7). Geoscientific Model Development 19, 2627–2656. doi.org. Accessed 2026-10-28.
  4. NASA Center for Climate Simulation (Versions 1.0 to 1.2). NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP-CMIP6). www.nccs.nasa.gov. Accessed 2026-10-28.
  5. Met Office (UKCP18, recommended national dataset as of 2026-10-28). UK Climate Projections (UKCP). www.metoffice.gov.uk. Accessed 2026-10-28.
  6. World Meteorological Organization (2026-05-28). Global Annual to Decadal Climate Update 2026–2035. wmo.int. Accessed 2026-10-28.
  7. Intergovernmental Panel on Climate Change (2023-03-20). Climate Change 2023: Synthesis Report. Summary for Policymakers (Sixth Assessment Report). www.ipcc.ch. Accessed 2026-10-28.
  8. World Meteorological Organization (2026-03-23). State of the Global Climate 2025. wmo.int. Accessed 2026-10-28.
  9. Copernicus Climate Change Service (2026-01-14). Global Climate Highlights 2025. climate.copernicus.eu. Accessed 2026-10-28.
  10. NOAA Office of Water Prediction (Informational page, accessed 2026-10-28). NOAA Atlas 15 informational page. water.noaa.gov. Accessed 2026-10-28.
  11. Eyring et al., Geoscientific Model Development (European Geosciences Union) (2016-05-26). Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geoscientific Model Development 9, 1937–1958. doi.org. Accessed 2026-10-28.
  12. World Climate Research Programme, CORDEX (Programme page, accessed 2026-10-28). About CORDEX: the Coordinated Regional Climate Downscaling Experiment. cordex.org. Accessed 2026-10-28.
  13. Intergovernmental Panel on Climate Change (2026-03-27). IPCC wraps up 64th plenary session in Bangkok (press release). www.ipcc.ch. Accessed 2026-10-28.
  14. International Organization for Standardization (2021-02). ISO 14091:2021 Adaptation to climate change: Guidelines on vulnerability, impacts and risk assessment. committee.iso.org. Accessed 2026-11-17.
  15. European Commission, Destination Earth initiative (Initiative site, accessed 2026-10-28). Destination Earth. destination-earth.eu. Accessed 2026-10-28.

Запись

Editorial record

Facts and policies are stated as of 28 October 2026. This article is reviewed at least every 180 days and whenever a cited assessment, dataset or instrument changes, including any agreed timeline for the IPCC Seventh Assessment Report, the first CMIP7 scenario simulations and the next WMO annual to decadal update.

Published by the GCRI editorial desk for the Climate and Hazards platform. To report an error or request a correction, use the platform's corrections and concerns route; corrections are dated and listed with the article.

Text is published under the Creative Commons Attribution 4.0 licence (CC BY 4.0): reuse it with credit to GCRI and a link to this page. Third-party data keep their own terms.

Cite this article

GCRI editorial desk (2026, October 28). Climate projections explained: scenarios, models and ranges. Climate and Hazards. https://therisk.global/climate-hazards/climate-projections/

Republish

Republish or translate this article free of charge, in print or online, under CC BY 4.0: keep the byline and Climate and Hazards, link to the original and note any change. Figures, maps and twins made by the institutions carry the same licence; third-party data keeps its own credit line.

Attribution to paste
<p>Climate projections: the foundations of exploring future climate — GCRI editorial desk, Climate and Hazards. First published 28 October 2026 at <a href="https://therisk.global/climate-hazards/climate-projections/">https://therisk.global/climate-hazards/climate-projections/</a>. Republished under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).</p>

Related across the Consortium

Доля

EmailLinkedInBlueskyMastodon

The Global Centre for Risk and Innovation (GCRI Canada) is a Canadian not-for-profit organisation, not a registered charity, with its registered office in Ontario.