GCRI · Climate and Hazards · Foundations
Climate baselines: the foundations of measuring a changing climate
Climate baselines are the reference periods against which every anomaly, record and extreme is judged, and the choice of baseline can move a number by almost a degree. This foundational guide explains the three main reference periods, how normals and global datasets are built, and why baselines must change as the climate does.
- Aggiornato al
- 28 October 2026
- Lettura
- 15 minutes
- Evidenza
- 16 sources, cited in order
- Metodo
- A synthesis of WMO regulations and guidance on normals, IPCC conventions and the 2025 releases of the main global temperature datasets, comparing how each reference period is defined, calculated and used.
- Dati
- Normals are published by WMO with NOAA's National Centers for Environmental Information and by national services; global temperature values come from the WMO State of the Global Climate 2025, the Copernicus Global Climate Highlights 2025, NOAA and Berkeley Earth, each from its custodian.
Metodi ed evidenze
- Domanda
- What are climate baselines, how are they defined and calculated, and how should they be used as the climate changes?
- Approccio
- A synthesis of WMO regulations and guidance on normals, IPCC conventions and the 2025 releases of the main global temperature datasets, comparing how each reference period is defined, calculated and used.
- Evidenza
- 16 sources, cited in order and listed under References
- Data availability
- Normals are published by WMO with NOAA's National Centers for Environmental Information and by national services; global temperature values come from the WMO State of the Global Climate 2025, the Copernicus Global Climate Highlights 2025, NOAA and Berkeley Earth, each from its custodian.
- Limiti
- Datasets differ by a few tenths of a degree, observations are sparse in some regions, recent reanalysis months are preliminary and dataset versions change past values. The article compares published values and does not recompute them.
01Definizione
What climate baselines are
Reference periods and their statistics: the yardstick for every anomaly and every extreme.
Climate baselines are the reference periods, and the averages and distributions computed over them, against which present weather and climate are compared. A baseline turns a raw measurement into a statement that matters: a month is warm or dry only relative to what is usual for that place and season, and a rainfall total is extreme only relative to the distribution it is drawn from.
The World Meteorological Organization (WMO) sets the vocabulary. Its technical regulations define period averages as averages over at least ten years, normals as period averages over a uniform period of at least three consecutive ten-year periods, and climatological standard normals as averages over a fixed sequence of 30-year periods 2. Since the Seventeenth World Meteorological Congress in 2015, the standard normal has been the most recent 30-year period finishing in a year ending in 0, while 1961–1990 is kept as the reference for assessing long-term change 2.
Global warming is measured against a different baseline. The Intergovernmental Panel on Climate Change (IPCC), in its Sixth Assessment Report on the physical science, uses 1850–1900 as an approximation for pre-industrial conditions because it is the earliest period with sufficiently complete global observations to estimate global surface temperature 6. A baseline is therefore always a choice made for a purpose: tracking human-caused warming, describing the current climate of a place, or setting a design value for infrastructure.
This article sets out those choices, how normals are calculated, how global datasets build their records and why baselines must change as the climate changes. It belongs to the Climate and Hazards platform's climate baselines 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 baselines matter now
One year, several numbers: why the baseline decides what a record means, and why design values are at stake.
Records and thresholds are only as meaningful as the baseline behind them. In 2025 the global average surface air temperature was 14.97 degrees Celsius in the ERA5 reanalysis reported in the Copernicus Global Climate Highlights 2025, 0.59 degrees above the 1991–2020 average and 1.47 degrees above 1850–1900 1. The Copernicus Climate Change Service puts the difference between the two periods at 0.88 degrees 1. A reader who misses the baseline can misread the state of the climate by almost a degree.
Policy goals are stated against the pre-industrial baseline. The WMO State of the Global Climate 2025 places 2025 at about 1.43 degrees Celsius above 1850–1900, with an uncertainty of plus or minus 0.13 degrees, and finds 2015 to 2025 the 11 warmest years on record 7. ERA5 shows 2023 to 2025 as the first three-year period to average more than 1.5 degrees above pre-industrial 1. Whether that level has been crossed in the sense of the Paris Agreement depends on the averaging period, which the IPCC sets at 20 years 8.
Engineering and services also rest on baselines. Drainage, dams, building codes and insurance tables often use statistics drawn from past records on the assumption that the climate they describe is stable. The U.S. National Oceanic and Atmospheric Administration (NOAA) is replacing that assumption in its national precipitation frequency standard: NOAA Atlas 15 is described as a shift from a stationary to a non-stationary assumption, in which extreme precipitation changes over time 5.
Warming continues while baselines stand still. WMO's latest decadal update predicts annual global temperatures of 1.3 to 1.9 degrees above 1850–1900 for each year from 2026 to 2030 9, so a 1991–2020 normal already describes a cooler climate than the one people now live in.
03Reference periods
Three reference periods and what each is for
Pre-industrial, fixed monitoring and current normal: three baselines with three different jobs.
1850–1900, the pre-industrial approximation. The IPCC states global warming relative to 1850–1900 and found global surface temperature 1.09 degrees Celsius higher in 2011–2020 than in that period, with a very likely range of 0.95 to 1.20 degrees 6. The IPCC uses it as an approximation of pre-industrial conditions: it is the earliest period with enough global observations, and keeping it fixed allows assessments to be compared over time 6.
1961–1990, the fixed monitoring reference. WMO keeps 1961–1990 as a standard reference period for long-term climate change assessment 2 and still recommends it for historical comparison and for tracking global anomalies against a fixed, common baseline 3. Many long national series and older publications use it, so it remains the bridge between past and present reporting.
1991–2020, the current standard normal. WMO recommends that 30-year standard periods be updated every decade and that 1991–2020 be adopted globally, so that sectors such as water, energy and agriculture plan with information that reflects the recent climate 3. The next standard normal, 2001–2030, can only be computed after 2030 2.
Other baselines exist for specific purposes. NOAA publishes 15-year normals for 2006–2020 alongside its 30-year normals, using the same methods, for uses that need more recent climate information 10. Projections often use 1995–2014 as their recent baseline; the IPCC's sea level and precipitation projections are stated relative to that period 6. A careful reader always checks which period a number refers to before comparing it with another, and converts between periods only with a stated offset from the same dataset, such as the 0.88 degrees Celsius that ERA5 uses between 1850–1900 and 1991–2020 1.
| Base di riferimento | Main purpose | Set or used by |
|---|---|---|
| 1850–1900 | Approximation for pre-industrial conditions; global warming and policy goals | IPCC 6 |
| 1961–1990 | Fixed reference for long-term change and historical comparison | WMO 2 |
| 1991–2020 | Current climatological standard normal | WMO and national services 3 |
| 2006–2020 | 15-year normals for uses needing recent information | NOAA 10 |
| 1995–2014 | Recent baseline for projections | IPCC 6 |
04Normals
How a normal is calculated
The WMO rules for turning daily observations into a 30-year normal, and the global collection that results.
A normal is built from the bottom up. Daily observations become monthly values, monthly values become 30-year averages, and each step has completeness rules. The WMO guidelines on the calculation of climate normals recommend not calculating a monthly mean from daily values if observations are missing for 11 or more days in the month, or for 5 or more consecutive days 2.
The 30-year step has its own test. A monthly normal needs data for at least 80 per cent of the years in the period, which for a standard normal means 24 or more of the 30 years, and should not be calculated if values are missing in 3 or more consecutive years 2. The guidelines also explain the origin of the 30-year convention: it was set mainly because only 30 years of data were available when the recommendation was first made 2.
Homogeneity matters as much as completeness. A station that moves, changes instruments or finds a town growing around it can show a step or trend that has nothing to do with the climate, and the WMO guidelines devote a section to homogeneity and to composite stations 2. Methods for detecting and correcting these breaks are the subject of the platform's methods and evidence article on climate baselines, Measuring climate baselines: evidence and its limits.
The result is a global collection. For 1991–2020, WMO collected normals from 141 countries and territories across its six regions, and NOAA's National Centers for Environmental Information, which hosts the World Data Center for Meteorology, publishes more than 6,500 station files 11. National collections go further: the United States 1991–2020 normals draw on more than 15,000 stations reporting precipitation and more than 7,300 reporting temperature 10.
05Dataset
From station to globe: datasets and reanalysis
How global records are built, why they differ by tenths of a degree, and what that spread means.
Global temperature records combine land stations and sea surface measurements into a gridded field and then an average. The HadCRUT5 dataset of the Met Office Hadley Centre and the Climatic Research Unit runs monthly from January 1850 on a 5-degree grid and is published as an ensemble of 200 realisations that sample its uncertainty 12. NOAA maintains its own surface record from 1850 4, and the independent Berkeley Earth project publishes another 13. Each record makes its own choices about which stations to include, how to adjust for changes at those stations and how to estimate temperature between them, so independent records are a test of one another as well as a source of numbers.
Reanalysis takes a different route. ERA5, produced for the Copernicus Climate Change Service, combines model data with observations from across the world into an hourly, 0.25-degree record from 1940 onwards, updated daily with a delay of about five days 14. It is one of six international datasets used to track global temperature 1.
The datasets agree on the trend and differ in the detail. For 2025, the estimates above 1850–1900 are 1.34 degrees Celsius from NOAA 4, 1.39 degrees from HadCRUT5 and 1.47 degrees from ERA5 1, and 1.44 degrees, plus or minus 0.09, from Berkeley Earth 13. WMO consolidates them to 1.43 degrees with an uncertainty of plus or minus 0.13 7. The differences come mainly from how each treats regions with few observations, such as the Arctic, and how it merges land and ocean data.
Methods also move the past. The IPCC notes that methodological advances and new datasets, including a more complete representation of the Arctic, raised its estimate of global warming by about 0.1 degrees Celsius between assessments without representing any additional physical warming 6. A baseline value is therefore a property of a dataset version as well as of a period, and serious comparisons cite both.
06Non-stationarity
Baselines in a changing climate
Why a fixed baseline drifts away from the present, and how standards are responding.
A normal assumes, for practical purposes, that the 30 years it covers share one climate. In a warming world that assumption weakens. The IPCC expresses changes in extremes relative to 1850–1900: an extreme heat event that occurred once in 50 years in that climate now occurs about 4.8 times as often, and would occur about 8.6 times as often at 1.5 degrees of warming and 13.9 times at 2 degrees 6. A return period drawn from older records will understate today's odds, and more so as warming continues.
Updating the normal every decade narrows the gap but does not close it, because a 30-year average still centres about 15 years in the past. This is why NOAA offers 15-year normals alongside its 30-year set for uses that need more recent information 10, and why some services describe the current climate with trend-adjusted statistics rather than a plain average.
Design standards are where this becomes concrete. NOAA Atlas 15 will provide two volumes: one that accounts for temporal changes in historical observations, and one that projects estimates into the future using downscaled climate model information 5. Preliminary estimates for the contiguous United States are scheduled no earlier than September 2026 and published estimates no earlier than September 2027, after which Atlas 15 will replace Atlas 14 as the national standard 5.
Baselines also anchor attribution and projection. Attribution studies compare today's odds of an event with those in a cooler counterfactual climate 15, and projections report change against a recent baseline. The platform's climate projections topic explains how those future changes are produced, and the heat and drought topic shows how shifting baselines change what counts as a heatwave.
07Evidenza
How baseline evidence is measured
The observations, collections and datasets behind climate baselines, and their limits.
Baselines rest on sustained observations. The Global Climate Observing System lists 55 Essential Climate Variables across atmosphere, land and ocean, from surface air temperature and precipitation to sea level and soil moisture, the variables that together characterise the climate 16. A baseline can only be as good as the record of the variable it describes.
Coverage is uneven. The normals collection depends on what each country submits 11. Reanalysis fills gaps with a physical model, which gives a complete field but carries the model's own biases, and its most recent data are a preliminary release (ERA5T), which may differ from the final release two to three months later if serious flaws are found 14.
Uncertainty is part of the evidence. HadCRUT5 publishes its estimate as 200 realisations 12, Berkeley Earth gives an uncertainty for each year 13, and WMO reports its consolidated value with a range 7. A careful use of a baseline quotes the dataset, the version, the period and the uncertainty together. The platform's methods and uncertainty and source and data register pages record how each dataset is used.
| Fonte | Che cosa fornisce | Custodian |
|---|---|---|
| Global 1991–2020 normals | Station normals from 141 countries and territories | WMO with NOAA NCEI 11 |
| National normals | Station and gridded normals, including 15-year normals | NOAA NCEI for the United States 10 |
| Surface temperature datasets | Global gridded temperature from 1850 with uncertainty | Met Office and CRU 12; NOAA 4; Berkeley Earth 13 |
| Rianalisi | Hourly, gridded reconstruction from 1940 | Copernicus Climate Change Service 14 |
| Consolidated indicators | Multi-dataset global values with uncertainty | WMO 7 |
08Strumenti
The rules and standards behind baselines
WMO regulations and guidance, IPCC conventions and national design standards, as of 28 October 2026.
The governing rules sit with WMO. Its technical regulations define period averages, normals and climatological standard normals, and the 2017 guidelines on calculating normals set the completeness and homogeneity practice that countries followed for 1991–2020 2,11. The decadal update cycle and the retention of 1961–1990 for long-term monitoring were endorsed in 2015 2.
The IPCC sets the conventions for global assessment: warming relative to 1850–1900, projections relative to 1995–2014, and the 20-year averaging used to judge whether a warming level has been reached 6,8. These conventions make one assessment comparable with the next.
National standards translate baselines into design values. In the United States, NOAA Atlas 14 remains the national precipitation frequency standard until Atlas 15 is published and supersedes it 5. The governance of how such standards are adopted, and by whom, is covered on the Global Risks Forum's climate risk governance topic.
09Frontiers
What is changing in climate baselines
Four areas where practice is moving.
Non-stationary design values. National standards are starting to build trends and projections into return levels rather than assuming a stable climate 5.
Shorter and trend-adjusted normals. Fifteen-year normals 10 and trend-aware statistics are being used where a 30-year average lags too far behind.
Reanalysis as a baseline. Reanalysis such as ERA5 can supply a baseline where station records are thin 14, which raises questions about how model bias enters the reference.
Warming-level baselines. Assessments increasingly describe change at global warming levels rather than in calendar periods 6, which ties baselines to the physics of warming rather than to dates.
10Consorzio
Three faces of climate baselines across the Consortium
Science, governance and finance read the same baseline for different decisions.
A baseline is a scientific choice with policy and financial consequences. Within the Consortium, The Global Centre for Risk and Innovation (GCRI) leads on science, technology and evidence: how baselines are built, tested and documented, supported by the platform's climate science for hazards and extreme value statistics 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 the institutions that adopt and revise standards, and its Diplomacy platform covers the international cooperation, such as shared observing systems, on which global baselines depend.
The Global Risks Alliance (GRA) translates baselines into the questions finance asks, such as how a shift in a return level changes physical climate risk in a portfolio. 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 baselines 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. Values quoted here belong to named datasets and versions and will change as those datasets are updated. 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.
Domande
Domande e risposte
What is a climate normal?
A climate normal is the average of a climate variable over a uniform 30-year period. WMO's current climatological standard normal is 1991–2020, the most recent 30-year period ending in a year ending in 0, and it is updated every decade 2,3.
Why do some reports use 1850–1900 and others 1991–2020?
They answer different questions. 1850–1900 approximates pre-industrial conditions and is used to measure human-caused warming 6. 1991–2020 describes the current climate of a place. In ERA5, 2025 was 0.59 degrees Celsius above 1991–2020 but 1.47 degrees above 1850–1900 1.
Why do global temperature datasets give different values for the same year?
Each centre uses its own observations and methods to fill gaps, especially in the Arctic and other sparsely observed regions. For 2025 the estimates above 1850–1900 range from 1.34 to 1.47 degrees Celsius 4,1, and WMO consolidates them to 1.43 degrees with an uncertainty of plus or minus 0.13 7.
How much data is needed to calculate a normal?
WMO guidance requires data for at least 80 per cent of the years, 24 of 30 for a standard normal, and no gap of 3 or more consecutive years. A monthly value from daily data is not calculated if 11 or more days, or 5 or more consecutive days, are missing 2.
Termini
Glossario
- Climatological standard normal
- The average of a climate variable over the most recent 30-year period ending in a year ending in 0, currently 1991–2020.
- Anomaly
- The difference between an observed value and the baseline average for the same place and time of year.
- Homogeneity
- The property of a climate record whose variations come only from the climate, not from station moves, instrument changes or surroundings.
- Rianalisi
- A reconstruction of past weather that combines observations with a weather model to produce a complete and consistent record.
- Non-stationarity
- The condition in which the statistics of a climate variable, such as the frequency of extremes, change over time.
Reuse and cite
Text CC BY 4.0. Cite the page with its date; data tables are linked under each figure.
Fonti
Riferimenti
16 sources, numbered in the order they are first cited
- Copernicus Climate Change Service (2026-01-14). Global Climate Highlights 2025. climate.copernicus.eu. Accessed 2026-10-28.
- World Meteorological Organization (2017). WMO Guidelines on the Calculation of Climate Normals (WMO-No. 1203). www.ncei.noaa.gov. Accessed 2026-10-28.
- World Meteorological Organization (2021-05-05). Updated 30-year reference period reflects changing climate. wmo.int. Accessed 2026-10-28.
- NOAA National Centers for Environmental Information (2026-01-14). Assessing the Global Temperature and Precipitation Analysis in 2025. www.ncei.noaa.gov. Accessed 2026-10-28.
- NOAA Office of Water Prediction (Informational page, accessed 2026-10-28). NOAA Atlas 15 informational page. water.noaa.gov. Accessed 2026-10-28.
- 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.
- World Meteorological Organization (2026-03-23). State of the Global Climate 2025. wmo.int. Accessed 2026-10-28.
- 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.
- World Meteorological Organization (2026-05-28). Global Annual to Decadal Climate Update 2026–2035. wmo.int. Accessed 2026-10-28.
- NOAA National Centers for Environmental Information (1991–2020 normals, v1.0.1). U.S. Climate Normals. www.ncei.noaa.gov. Accessed 2026-10-28.
- NOAA National Centers for Environmental Information (1991–2020 collection, product page accessed 2026-10-28). WMO Climate Normals (1991–2020 global collection). ncei.noaa.gov. Accessed 2026-10-28.
- Met Office Hadley Centre and Climatic Research Unit, University of East Anglia (HadCRUT.5.2.0.0). HadCRUT5 global temperature dataset. www.metoffice.gov.uk. Accessed 2026-10-28.
- Berkeley Earth (2026-01-14). Global Temperature Report for 2025. berkeleyearth.org. Accessed 2026-10-28.
- Copernicus Climate Change Service (Climate Data Store dataset, updated daily). ERA5 hourly data on single levels from 1940 to present. cds.climate.copernicus.eu. Accessed 2026-10-28.
- Philip et al., Advances in Statistical Climatology, Meteorology and Oceanography (2020-11-10). A protocol for probabilistic extreme event attribution analyses. Advances in Statistical Climatology, Meteorology and Oceanography 6, 177–203. doi.org. Accessed 2026-10-28.
- Global Climate Observing System (Current ECV list, accessed 2026-10-28). Essential Climate Variables. gcos.wmo.int. Accessed 2026-10-28.
Registrazione
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 the WMO State of the Global Climate 2026, new releases of the cited temperature datasets and the publication of NOAA Atlas 15.
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 baselines: reference periods, normals and data. Climate and Hazards. https://therisk.global/climate-hazards/climate-baselines/
@online{gcri2026climatebaselines,
author = {{GCRI editorial desk}},
title = {{Climate baselines: reference periods, normals and data}},
publisher = {Climate and Hazards},
year = {2026},
month = {10},
url = {https://therisk.global/climate-hazards/climate-baselines/},
language = {en}
} TY - ELEC AU - GCRI editorial desk TI - Climate baselines: reference periods, normals and data PB - Climate and Hazards PY - 2026 DA - 2026/10/28 UR - https://therisk.global/climate-hazards/climate-baselines/ LA - en ER -
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 baselines: the foundations of measuring a changing climate — GCRI editorial desk, Climate and Hazards. First published 28 October 2026 at <a href="https://therisk.global/climate-hazards/climate-baselines/">https://therisk.global/climate-hazards/climate-baselines/</a>. Republished under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).</p>