Sources and assessment
For each data source of the climate page this shows who collects it, what it provides, where on the page it is used, what it says about its own limits and under which terms it may be used.
Rules of this overview
- The statements come from the sources’ original pages (retrieved on 20 November 2026) or from the page’s own texts and the method register, not from memory.
- What could not be read is explicitly marked “not read” or “not retrievable”; there is then no summary of our own.
- The assessment reproduces what the source itself says about method and limits. It is not a rating of the source and not a statement about causes.
- For counter-positions and specialist articles see the section “Further sources and counter-positions” on the climate page.
IPCC, Sixth Assessment Report (AR6), Working Group I
not retrievable · To the source
- Who
- Intergovernmental Panel on Climate Change (IPCC).
- What it provides
- Report of Working Group I (2021) on the physics of the climate system, whose Summary for Policymakers (SPM) the page links in the “Positions” section.
- Used on this page
- Scenarios to 2100: SSP1-2.6, SSP2-4.5, SSP5-8.5 · Countries and scenarios: calculator for CO₂ savings, warming and costs
- Assessment: method and limits
- The page relies on the IPCC for the relation between emitted CO₂ and warming (0.45 °C per 1,000 bn t CO₂, likely range 0.27 to 0.63) and for the statement on human-caused warming 2010–2019 relative to 1850–1900 (in the “Positions” section).
- The scenario chart shows raw model means and is explicitly not the range assessed by the IPCC.
- Use and licence
- Not checked.
- Retrieval on 20 Nov 2026
- The original pages (ipcc.ch) could not be read when retrieved on 20 November 2026 (HTTP 403). There is therefore no summary of the report of our own here; the assessment follows only after checking the original.
CMIP6 (Coupled Model Intercomparison Project) via the Pangeo archive
partly read · To the source
- Who
- Climate research groups worldwide; the “Pangeo-CMIP6” archive is on Google Cloud.
- What it provides
- Monthly data from climate models for the past (historical) and for scenarios (ssp126, ssp245, ssp585).
- Used on this page
- Climate models against measurement: air temperature · Climate models against measurement: sea surface · Several runs per model · Time windows, internal variability and model difference · Quality check of the models · Europe, the Alps and Austria: models against measurements · Model map Alpine region: temperature trend, models against measurement · Scenarios to 2100: SSP1-2.6, SSP2-4.5, SSP5-8.5
- Assessment: method and limits
- The files keep the licence of their model: 20 models CC BY-SA 4.0, 3 models CC BY-NC-SA 4.0 (read from the datasets’ metadata).
- Models and runs are not independent, and nothing was selected by agreement with the measurement (no “hot” models filtered out).
- Use and licence
- Per model in the metadata (CC BY-SA 4.0 or CC BY-NC-SA 4.0); citation: Eyring et al. (2016), Geoscientific Model Development, doi:10.5194/gmd-9-1937-2016.
- Retrieval on 20 Nov 2026
- The licences were read from the datasets themselves. The project’s description page (wcrp-cmip.org) could not be read on 20 November 2026 (HTTP 403).
HadCRUT5 (Met Office Hadley Centre / Climatic Research Unit, UEA)
read at the original · To the source
- Who
- A joint product of the Met Office Hadley Centre and the Climatic Research Unit at the University of East Anglia.
- What it provides
- Global surface temperature anomalies on a 5° grid, monthly from January 1850, relative to 1961–1990; a blend of land air temperature (CRUTEM5) and sea surface temperature (HadSST4).
- Used on this page
- Climate models against measurement: air temperature · Several runs per model · Time windows, internal variability and model difference · Quality check of the models · Scenarios to 2100: SSP1-2.6, SSP2-4.5, SSP5-8.5
- Assessment: method and limits
- The uncertainty is available as an ensemble of 200 realisations; the page shows its 95 % band as a grey stripe (version 5.2.0.0, status of the source page: 18 September 2026).
- In the non-infilled dataset grid cells without observations stay empty; the infilled dataset is more globally complete and has additional uncertainty components.
- The measurement mixes air over land with water over the sea, while the models in the comparisons partly show pure air temperature; the page therefore also has a “blended” comparison.
- Use and licence
- Open Government Licence v3.0; citation: Morice et al. (2021), Journal of Geophysical Research: Atmospheres.
Berkeley Earth
read at the original · To the source
- Who
- Founded in 2010 by Richard and Elizabeth Muller; since 2013 an independent non-profit organisation (501(c)(3)), by its own account entirely supported by donations.
- What it provides
- Land temperatures (global from 1850, some land areas back to 1750) and a land-ocean series that combines the land data with a modified version of HadSST.
- Used on this page
- Quality check of the models · Map and globe: sea, land, El Niño, CO₂ in the air · Climate models against measurement: air temperature
- Assessment: method and limits
- Station records are checked for breakpoints against neighbouring stations and adjusted; the source data contain, by its own account, many quality-control and redundancy issues.
- The 1°×1° time series and grids have no longer been updated since the second quarter of 2025; the page’s land map therefore ends in 2024.
- Use and licence
- Creative Commons BY-NC 4.0 (non-commercial use only, attribution to Berkeley Earth requested).
NOAA ERSST (Extended Reconstructed Sea Surface Temperature)
read at the original · To the source
- Who
- NOAA National Centers for Environmental Information (NCEI).
- What it provides
- Global monthly sea surface temperature on a 2° grid from ship and buoy measurements (ICOADS 3.0) and Argo floats, reconstructed with Empirical Orthogonal Teleconnections; version 5 from January 1854.
- Used on this page
- Climate models against measurement: sea surface · Map and globe: sea, land, El Niño, CO₂ in the air · Quality check of the models
- Assessment: method and limits
- According to the source the data are more reliable after the 1940s; numerical uncertainty values are not given on the product page.
- The page uses version 5 (ERSSTv5). The NCEI product page now lists version 6 as the current one, version 5 under “previous versions”; whether the results differ with version 6 has not been checked.
- Use and licence
- Not stated on the product page.
NOAA Global Monitoring Laboratory: CO₂ at Mauna Loa
read at the original · To the source
- Who
- NOAA Global Monitoring Laboratory; the record began in March 1958 with C. David Keeling (Scripps Institution of Oceanography), NOAA has measured in parallel since May 1974.
- What it provides
- Monthly means of CO₂ at Mauna Loa Observatory (Hawaii, 3,400 m), according to the source the longest record of direct CO₂ measurements in the atmosphere.
- Used on this page
- CO₂ seasonal cycle, Keeling spiral and residence time · CO₂ in the air: amount, origin and fate
- Assessment: method and limits
- The source states an accuracy “generally better than 0.2 ppm” and points out that the Mauna Loa value need not equal the globally averaged surface CO₂; the series is therefore labelled as a station on the page.
- Measurements at Mauna Loa were suspended from 29 November 2022; December 2022 to July 2023 come from a site on Maunakea (about 21 miles north), after which measurements at Mauna Loa resumed. The latest year is preliminary.
- Since April 2019 a cavity ring-down analyser has been used, before that infrared absorption; monthly means come from hourly means of “background” hours.
- Use and licence
- Not stated on the pages read; Scripps keeps its own data at scrippsco2.ucsd.edu.
Global Carbon Budget (Global Carbon Project) and Our World in Data
partly read · To the source
- Who
- The Global Carbon Project created the Global Carbon Budget; the Global Carbon Budget Office at the University of Exeter (led by Pierre Friedlingstein) runs it. Our World in Data is produced by researchers at the University of Oxford with the non-profit Global Change Data Lab.
- What it provides
- An annual budget of global CO₂ emissions and sinks; Our World in Data provides the emissions per country from it.
- Used on this page
- CO₂ in the air: amount, origin and fate · Countries and scenarios: calculator for CO₂ savings, warming and costs
- Assessment: method and limits
- The page’s country data are territorial (production-based) emissions from fossil sources and industry: they allocate emissions by where they were released and do not adjust for traded goods; consumption-based data exist only from 1990 and only for some countries.
- Emissions from land-use change have, according to Our World in Data, significantly greater uncertainty than those from fossil fuels and are not included in the calculator; international aviation and shipping are not in the country totals.
- Use and licence
- Our World in Data: Creative Commons BY (CC BY 4.0), for third-party data the terms of the original authors apply. No licence was found for the Global Carbon Budget on the pages read.
- Retrieval on 20 Nov 2026
- The pages were read on 20 November 2026. The Our World in Data overview page was only about 60 % readable; the statements come from the readable part and the data-sources page.
NSIDC Sea Ice Index, version 4
read at the original · To the source
- Who
- National Snow and Ice Data Center (NSIDC), part of CIRES at the University of Colorado Boulder, with NOAA.
- What it provides
- Sea-ice extent and concentration for the Arctic and Antarctic from satellite microwave data (SMMR, SSM/I, SSMIS, AMSR2), 25 km grid, from late 1978; version 4 uses AMSR2 as input.
- Used on this page
- Sea ice and sea level
- Assessment: method and limits
- For trends the monthly products are, according to the source, better suited because errors in the daily values average out in the monthly mean.
- Due to funding limits the dataset was moved to a “Basic” service level (data integrity verified, basic user support); the anomalies refer to 1981–2010.
- The definition of extent (area with at least 15 % ice cover) and daily values “every second day until 1987” are given on the chart’s page; they were not read in the NSIDC user guide.
- Use and licence
- No licence named; use of the dataset requires citation (Fetterer et al., doi:10.7265/a98x-0f50).
University of Colorado Sea Level Research Group
read at the original · To the source
- Who
- Sea Level Research Group at the University of Colorado (CIRES), Boulder; radar altimeter data (RADS) from NOAA and TU Delft.
- What it provides
- Global mean sea level from satellite altimetry (TOPEX, Jason-1, -2, -3); version 2026_rel2, updated to 2026.5.
- Used on this page
- Sea ice and sea level
- Assessment: method and limits
- The correction for land uplift and subsidence since the last ice age (GIA) is applied: it raises the rate of rise by about 0.3 mm per year; the uncertainty of the rate is stated as ±0.4 mm per year, that of the GIA correction as at least 50 %.
- According to the processing page the satellite record begins in December 1992 with TOPEX; the page’s chart shows sea level since 1993.
- Use and licence
- Not stated on the pages read (only “© Regents of the University of Colorado”); citation: Nerem et al. (2018), PNAS.
E-OBS (Copernicus Climate Change Service / ECA&D)
partly read · To the source
- Who
- The access page is hosted by the Copernicus Climate Change Service; the station data come from ECA&D. Who exactly produces the dataset is not stated in the part of the page that was read.
- What it provides
- Daily gridded data (temperature, precipitation, pressure, humidity, wind, radiation) for Europe, 0.1° and 0.25°, from 1950; the page uses version 31.0e (daily mean temperature) and the matching elevation grid. According to the source page the current version is 34.0e to 30 June 2026.
- Used on this page
- Europe, the Alps and Austria: models against measurements · Alpine trend map and height profile (E-OBS) · Model map Alpine region: temperature trend, models against measurement
- Assessment: method and limits
- The dataset is built by conditional simulation (ensemble with 20 members); the best guess is the ensemble mean, the uncertainty is the difference between the 5th and 95th percentile (“90 % range”). The page uses the ensemble mean and does not include the uncertainty.
- The source points to low station density in some periods and areas; nothing about mountains is stated in the part read. The page itself therefore warns that E-OBS may be less certain in mountains.
- Use and licence
- Not stated in the part of the source page that was read (only about 70 % was readable); strict conditions, so the data package passes on only derived trends and no E-OBS series.
GeoSphere Austria (Data Hub)
read at the original · To the source
- Who
- GeoSphere Austria – Federal Institute for Geology, Geophysics, Climatology and Meteorology.
- What it provides
- Station data (point measurements) and spatial data; the page uses the monthly and daily values of the station Amstetten and temperatures from Sonnblick, Obergurgl and Rudolfshütte.
- Used on this page
- Europe, the Alps and Austria: models against measurements · Time travel: stations, thermometers and reconstructions · Glaciers: summer temperature and mass balance · Glacier map
- Assessment: method and limits
- The Data Hub gives no warranty for correctness, completeness and timeliness of the data; the terms of use are in the metadata of each dataset.
- A station is a point, not an area: it lies at a particular height and location; gaps (Amstetten 1989–1992) stay open and are not filled.
- Use and licence
- Attribution “Datenquelle: GeoSphere Austria – https://data.hub.geosphere.at”; the licence of the dataset used (station data v2, doi:10.60669/923n-p390) was read at the dataset as CC BY 4.0.
WGMS – World Glacier Monitoring Service (Fluctuations of Glaciers)
partly read · To the source
- Who
- World Glacier Monitoring Service, supported by ISC (WDS), IUGG (IACS), UNEP, UNESCO and WMO; the measurements come from the observers of the individual glaciers.
- What it provides
- Internationally collected, standardised data on glacier changes (length change and mass balance); the page uses the database in the version of 10 February 2026 (doi:10.5904/wgms-fog-2026-02-10).
- Used on this page
- Glacier map · Glaciers: length since 1850 · Glaciers: mass balance · Glaciers: summer temperature and mass balance
- Assessment: method and limits
- The page shows six selected Austrian glaciers, not all glaciers of the Alps; length reacts with a delay and depends on shape and debris cover, mass balance is the more direct climate signal.
- Use and licence
- Free use with correct citation (read at the dataset); the pages on data policy and guidelines were not read.
- Retrieval on 20 Nov 2026
- The product page “Fluctuations of Glaciers” was read (an older description of the printed volumes I–X); the statements on version and licence come from the dataset itself.
NASA GISS: research brief on water vapour and the greenhouse effect (Schmidt 2010)
read at the original · To the source
- Who
- NASA Goddard Institute for Space Studies; brief by Gavin Schmidt, October 2010 (according to the page no longer actively maintained).
- What it provides
- Assesses what share water vapour, clouds and CO₂ have in the absorption of the greenhouse effect and how water vapour acts as a feedback.
- Used on this page
- no chart of its own
- Assessment: method and limits
- According to the brief water vapour is responsible for about 50 % of the absorption, clouds for about 25 % and CO₂ for 20 %. These are shares of the absorption, not of the warming caused by one factor.
- Water vapour responds to changes in CO₂ as a feedback: in the GISS model a CO₂ doubling raises the absorption directly by 4 W/m², with water vapour and clouds to almost 20 W/m²; the capacity of the air for water vapour rises by about 7 % per degree.
- Use and licence
- Not stated on the page.
- Retrieval on 20 Nov 2026
- The brief is linked in the “Further sources and counter-positions” section.
Other sources named
These sources are named on the page; an assessment of our own is still to come (their statements on period and method are in the method register):
- Law Dome (MacFarling Meure et al. 2006, NOAA NCEI), Vostok ice core (Petit et al. 1999), Ljungqvist (2010)
- CenCO2PIP 2023, Westerhold et al. 2020, Scotese et al. 2021, Foster et al. 2017 (time travel over millions of years)
- Joos et al. 2013 (decay curve of CO₂)
- GPCP v2.3 and CMAP (precipitation products, NOAA PSL)
- IMBIE (Shepherd et al. 2018), Antarctic ice sheet
- KfW Research/Prognos, IHS/WIFO, European Commission SWD(2024) 63 (cost studies)
- Natural Earth, NASA GIBS “Blue Marble”, MapLibre GL, Wikimedia Commons (maps and images)
Status and upkeep
Status: 20 November 2026. The overview is maintained together with the method register; please report errors via the feedback form of the climate page.
