Resolution algorithm:
At the end of January 2026, consider the most recent estimates of annual global temperatures (including 2025 data) that have been published online by 1) NASA, 2) NOAA, and 3) Berkeley Earth.
For each of these sources, check whether the estimate for 2025 is strictly higher than the estimates for all other past years (since 1900, say). (The most precise published version of the data counts. If a source publishes one dataset where it's 1.15 versus 1.15, and another where it's 1.152 versus 1.151, that counts as strictly higher. If it publishes only the former dataset, it does not count as strictly higher, even if the latter is true "under the hood".)
The market resolves YES if:
Three sources provided estimates, and at least two showed 2025 as the warmest.
Two sources provided estimates, and both showed 2025 as the warmest.
One source provided estimates, and it showed 2025 as the warmest.
The market resolves NO if:
Three sources provided estimates, and at most one showed 2025 as the warmest.
Two sources provided estimates, and at most one showed 2025 as the warmest.
One source provided estimates, and it did not show 2025 as the warmest.
The market resolves N/A if none of the sources provided estimates.
The resolution will ignore everything that isn't a list of numbers. For example, if NASA writes in an article that 2025 was the warmest year on record, but its most recently published list of annual numbers has 2025 as lower or equal compared to some other year, that counts as NASA not showing 2025 as the warmest year.
Since I've done my best to make the resolution criteria well-defined and based on public data, I'll probably bet in this market.
Update 2026-01-14 (PST) (AI summary of creator comment): The creator has indicated they would be fine with resolving early (before the end of January 2026 as originally specified) or waiting until the end of January as per the original criteria. All three sources have published data showing 2025 was less hot than 2024 by a fair margin.
1,000