SCIENCE · INVESTIGATE AND EXPLAIN

Compare climate indicators with their metadata

Preserve temperature and CO₂ units, geographic scopes, coverage and source uncertainty.

Learning outcome: Build a paired evidence ledger without replacing missing observations or treating concentration as emissions.

Starting recap

Temperature is a global anomaly; this CO₂ record is an observing-site dry-air mole fraction, not a global emissions rate.

All lessons are open directly. Your saved experiments remain when you leave and return.

Try

The full selection has 145 temperature values but only 66 CO₂ annual values from 1959–2024.

Open the experiment

Predict

Does matching the requested analysis window guarantee identical observed coverage for both series?

Your prediction and reason

Experiment

Compare full-record and shared-coverage windows. Inspect missing early CO₂, annual source uncertainty and the 2022–2023 observing-site note.

Investigate recorded climate series

Recorded climate estimates · frozen historical series · Frozen climate-records-v1, selected through 2024. NASA GISTEMP v4 global land–ocean annual anomalies and NOAA/Scripps Mauna Loa annual dry-air CO₂ mole fractions are different observational products. Rebasing subtracts a constant. Descriptive ordinary least-squares trends are not causal attribution, forecasts or uncertainty intervals. No live data are requested.

Analysis interval: 1959–2024 Temperature reference period: Source baseline: 1951–1980

Reset keeps your prediction, captures and practice history.

Snapshot climate-records-v1 retains annual temperature estimates for 1880–2024 and Mauna Loa CO₂ for 1959–2024. The complete bundled timeline remains visible; gold shading marks the selected interval. These are processed observational products, not direct readings from one global thermometer.

Descriptive statistics for available annual values in the selected interval
Quantity	Temperature anomaly (°C)	CO₂ mole fraction (ppm)
Actual coverage	1959–2024	1959–2024
Available / unavailable years	66 / 0	66 / 0
Arithmetic mean of available annual values	0.3868	360.2495
Least-squares slope per decade	0.1768	16.5669
First-to-last slope per decade	0.1923	16.7123
Subtracted temperature baseline offset	0.000000 °C	Not applied to CO₂

The fit uses every available annual value in the interval with equal weight: slope per decade = 10 × Σ[(year − mean year)(value − mean value)] / Σ[(year − mean year)²]. The endpoint slope uses only the first and last available values. Neither describes every year's change. No confidence interval or significance test is calculated; annual climate values are not assumed to be independent random samples.

Inspect all selected annual source values

Frozen annual estimates, with source and rebased anomalies kept separate
Year	Source anomaly °C (1951–1980)	Selected anomaly °C	CO₂ ppm	Reported CO₂ uncertainty ppm
1959	0.03	0.0300	315.98	0.12
1960	-0.02	-0.0200	316.91	0.12
1961	0.06	0.0600	317.64	0.12
1962	0.03	0.0300	318.45	0.12
1963	0.05	0.0500	318.99	0.12
1964	-0.20	-0.2000	319.62	0.12
1965	-0.11	-0.1100	320.04	0.12
1966	-0.06	-0.0600	321.37	0.12
1967	-0.02	-0.0200	322.18	0.12
1968	-0.08	-0.0800	323.05	0.12
1969	0.05	0.0500	324.62	0.12
1970	0.03	0.0300	325.68	0.12
1971	-0.08	-0.0800	326.32	0.12
1972	0.01	0.0100	327.46	0.12
1973	0.16	0.1600	329.68	0.12
1974	-0.07	-0.0700	330.19	0.12
1975	-0.01	-0.0100	331.13	0.12
1976	-0.10	-0.1000	332.03	0.12
1977	0.18	0.1800	333.84	0.12
1978	0.07	0.0700	335.41	0.12
1979	0.16	0.1600	336.84	0.12
1980	0.26	0.2600	338.76	0.12
1981	0.32	0.3200	340.12	0.12
1982	0.14	0.1400	341.48	0.12
1983	0.31	0.3100	343.15	0.12
1984	0.16	0.1600	344.87	0.12
1985	0.12	0.1200	346.35	0.12
1986	0.18	0.1800	347.61	0.12
1987	0.32	0.3200	349.31	0.12
1988	0.39	0.3900	351.69	0.12
1989	0.27	0.2700	353.20	0.12
1990	0.45	0.4500	354.45	0.12
1991	0.41	0.4100	355.70	0.12
1992	0.22	0.2200	356.54	0.12
1993	0.23	0.2300	357.21	0.12
1994	0.31	0.3100	358.96	0.12
1995	0.45	0.4500	360.97	0.12
1996	0.33	0.3300	362.74	0.12
1997	0.47	0.4700	363.88	0.12
1998	0.61	0.6100	366.84	0.12
1999	0.38	0.3800	368.54	0.12
2000	0.39	0.3900	369.71	0.12
2001	0.53	0.5300	371.32	0.12
2002	0.63	0.6300	373.45	0.12
2003	0.62	0.6200	375.98	0.12
2004	0.53	0.5300	377.70	0.12
2005	0.68	0.6800	379.98	0.12
2006	0.64	0.6400	382.09	0.12
2007	0.66	0.6600	384.02	0.12
2008	0.54	0.5400	385.83	0.12
2009	0.66	0.6600	387.64	0.12
2010	0.72	0.7200	390.10	0.12
2011	0.61	0.6100	391.85	0.12
2012	0.65	0.6500	394.06	0.12
2013	0.67	0.6700	396.74	0.12
2014	0.75	0.7500	398.81	0.12
2015	0.89	0.8900	401.01	0.12
2016	1.01	1.0100	404.41	0.12
2017	0.91	0.9100	406.76	0.12
2018	0.85	0.8500	408.72	0.12
2019	0.98	0.9800	411.65	0.12
2020	1.01	1.0100	414.21	0.12
2021	0.85	0.8500	416.41	0.12
2022	0.89	0.8900	418.53	0.12
2023	1.17	1.1700	421.08	0.12
2024	1.28	1.2800	424.61	0.12

NOAA defines the annual uncertainty column from the standard deviation of differences between independently determined NOAA and Scripps annual means; it is not a regression confidence interval. The source temperature table supplies no annual uncertainty column here. Decimal precision does not imply zero uncertainty. Temperature rebasing uses the mean of 30 rounded annual anomalies; the native 1951–1980 annual mean can differ slightly from zero because of rounding.

Mauna Loa is a background-air observing location, not a global mean or emissions inventory. Its record includes Maunakea observations during the 2022–2023 eruption interruption. The bundled analysis stops at 2024 to use a fixed historical window; it is not a claim about the latest year. Upstream revisions do not silently replace this snapshot.

Sources: NASA GISTEMP v4, Lenssen and colleagues (2024), observational uncertainty, and NOAA / Scripps CO₂ data. Retrieved 25 October 2026. Two rising series alone do not establish causal attribution; physical mechanisms, alternative drivers and independent evidence are also needed. There is no CO₂-to-emissivity calibration or forecast in this investigation.

Temperature trend 0.1768 °C per decade across 66 annual estimates; CO₂ trend 16.5669 ppm per decade across 66 available annual means.

Selected interval 1959–2024. Temperature reference shift 0.000000 °C. CO₂ coverage 1959–2024; 0 years unavailable in this selection. Missing values are not zero. Co-trends alone do not identify a cause.

Observation to attach to a capture

Captures preserve interval and reference-period selections in climate-records-v1 and replay descriptive calculations from frozen NASA and NOAA/Scripps annual products. They are not new measurements, live updates or causal experiments.

Understand

Identify what each series measures

The two panels show different quantities with different observational coverage. Temperature is NASA’s global annual land–ocean anomaly in degrees Celsius. CO₂ is an annual dry-air mole fraction from the Mauna Loa record, expressed in parts per million. Separate vertical axes preserve the units without suggesting that the heights of two curves are directly comparable.

NOAA defines the CO₂ quantity as the fraction of dry-air molecules that are carbon dioxide. A value of 400 ppm corresponds to a fraction of 0.000400. It is not 400 tonnes of emissions per year. NOAA also distinguishes this observing record from its global surface-average CO₂ product. NOAA CO₂ record description.

At 2024, this snapshot gives +1.28°C on the native temperature reference and 424.61 ppm CO₂. These describe one year using two indicators; they are not interchangeable measurements of the same variable. Write a paper row containing value, unit, geographic scope and source for each before comparing their trends.

Preserve the difference in coverage

The temperature series covers 1880–2024 with 145 annual values. The frozen CO₂ series covers 1959–2024 with 66 annual values. Earlier CO₂ entries are missing, represented as null. They are not zeros, interpolated measurements or proof that atmospheric carbon dioxide was absent.

Select 1880–2024 and inspect the actual coverage reported for each series. A requested window can contain fewer usable observations than its full number of years. The model computes CO₂ summaries over available records, so a temperature summary for 1880–2024 and a CO₂ summary for 1959–2024 do not describe identical intervals.

Then select 1959–2024. Both now span the same calendar years, although their geographic definitions remain different. Calendar alignment improves comparability but does not turn Mauna Loa into a global measurement or make the observations independent controlled treatments.

Read provenance alongside the numbers

The NOAA record includes observations from Maunakea during the interruption following the Mauna Loa eruption: December 2022 through early July 2023. The annual values for those years therefore require that provenance note. A familiar record name should not hide a documented change in observing location.

The downloaded annual file supplies CO₂ uncertainty values, which are preserved. Treat them as source-provided uncertainty, with the source’s definition, rather than inferring a new quality grade. They are not confidence intervals for the fitted trend, and they do not describe uncertainty in the temperature series. The NOAA download page also notes that quality-control revisions can change published values.

Make an evidence ledger with separate columns for observation, derived calculation and metadata. Annual source values belong in the first column; means and slopes computed by this lab belong in the second; coverage, reference periods and site notes belong in the third. All three are needed to interpret the result.

Distinguish concentration change from emissions

A CO₂ slope in ppm/decade describes change in atmospheric concentration at the record’s observational scope. It does not directly measure the emission rate from human activities, volcanoes or ecosystems. Atmospheric amounts reflect both additions and removals, and this investigation does not solve a carbon budget.

Similarly, a rising temperature anomaly is not itself a radiation measurement. The earlier energy and greenhouse models offer mechanisms to investigate, but the record viewer contains no calibration converting ppm to atmospheric grey emissivity. Do not join the two labs by inventing that missing relationship.

Changing the temperature reference shifts only the temperature panel and corresponding means. It does not change CO₂ values, available years or either series’ slope over unchanged observations. Predict this invariance and verify it before drawing a combined conclusion.

Transfer to a missing-data report

A report fills all CO₂ years before 1959 with zero to “match” the temperature timeline, then calculates a full-period mean. Explain the error and propose a defensible comparison.

Check the climate-indicators transfer

Missing records do not mean zero concentration. Filling them with zeros invents observations and distorts summaries. Preserve the missing interval and either report each series' actual coverage separately or compare their shared 1959–2024 interval, retaining the different geographic scopes.

Apply

Public practice checks a specific claim. It does not establish independent mastery.

Distinguish values, derived summaries and metadata, and explain why ppm/decade is not an emissions rate.

Your explanation and limitations

Continue

Completion requires a saved prediction, at least one capture and a reflection. Presence checks do not grade your reasoning.

Skipping retains experiments and awards no mastery.

Previous: Compare climate windows and trend summaries Next: Separate climate observations from attribution claims

Save a science backup or import one

Science work is stored locally in this browser, separately from mathematics and AI. It does not sync automatically. Use this science-only backup; this checkout's older all-subject export does not include science.

A legacy file is available only when the complete snapshot fits the 36 MB limit. For larger snapshots, prepare a complete file set below; never discard notebooks to make a smaller file.

Choose a science backup (maximum 36 MB)

Matching imported notebook drafts replace their current drafts. Omitted lessons and known first attempts remain. Export your current work first if you want to retain both drafts.

Complete file sets and course backups

A file set can hold more work than one science file. Keep its manifest and every part together. Each file is limited to 36 MB. This format needs this version of LearnAI or a compatible newer version; older versions can use only the legacy single-file backup above.

Export scope

The set includes every saved science course and global continuation at this moment. Later edits need a new export. A conflicting tab exports its own visible work, which may differ from saved work in another tab.

Choose the manifest and all parts together

0 file(s) selected. File order and filenames do not determine identity. Missing, changed or conflicting parts block the entire restore; no good part is applied by itself.

Science storage: loading. No account or remote computation is required.