Environment: NASA GISTEMP and four major climate datasets disagree on whether 2025 ranked second or third hottest year on record

Findings

• Six of eight major global temperature datasets rank 2025 as the third-hottest year on record. NASA GISTEMP and DCENT rank it second (marginally edging 2023). The Delta Engine returned INDETERMINATE (deltaFinal=1.321 vs 0.05 threshold — the highest observed across all research runs to date on a factual ranking claim), because NASA's prominence score for '2025 was second' is 88 while Berkeley Earth's is 18: Berkeley Earth literally titled its 2025 annual report 'Third Warmest Year on Record.'
• The divergence is a legitimate methodological disagreement, not a measurement error. Different datasets apply different spatial coverage weights, sea-surface temperature records, and infilling approaches for data-sparse Arctic regions. NASA GISTEMP's weighted spatial average marginally places 2025 above 2023 by a difference within the margin of uncertainty; Berkeley Earth and Copernicus ERA5 place 2025 slightly below 2023 using different averaging. All datasets agree on the temperature magnitude: approximately 1.44°C above the 1850–1900 pre-industrial baseline (Copernicus: 1.47°C, WMO consolidated: 1.44°C ± 0.13°C).
• The 2nd vs 3rd distinction is not a minor rounding difference — it determines whether 2025 is framed as 'the second year in a row at or near record highs' (NASA framing) or 'a modest cooling from the 2024 record peak' (NOAA/Berkeley Earth framing). ESG reports, climate risk disclosures, and policy documents citing this claim without specifying the dataset are citing something that four of five major agencies disagree with if the cited ranking is 'second.'
• For sustainability analysts and climate risk professionals: the 2026 WMO State of the Global Climate report explicitly acknowledges the split — 'second warmest in GISTEMP and DCENT; third warmest in Berkeley Earth, CMST, ERA5, HadCRUT5, JRA-3Q, and NOAAGlobalTemp' — and uses the majority ranking. Any citation of '2025 was the second hottest year' in a regulatory filing that cites only a NASA press release without checking the broader dataset consensus is drawing from the outlier, not the consensus.