⬆ 5th or 6th highest surface temps
⬆ Warmest summer on land
⬆ Warmest year for 25 countries + 1.8 billion people
⬆ Record ocean heat
⬆ Record high GHGs
⬆ Record high sea levels
⬇ Record low glacier mass
1/18
2021 was a bit cooler than the last few years due to a moderate La Nina event. La Nina tends to result in cooler temps globally, though the global response tends to lag 3-4 months after peak conditions. Here is what global temps look like since 1970 with and without ENSO removed:
The years since 2015 – 2021 included – are quite a bit warmer than any years that came before. Barring a Pinatubo-sized eruption in the next few years, its exceedingly unlikely we will ever see a year as cool as 2014 again:
Land temperatures – where we all live – are warming 40% faster than the world as a whole (which is mostly oceans). The world's land has warmed by around 1.8C already since preindustrial times:
Areas home to 1.8 billion people saw their warmest year on record during 2021, with 25 countries – including China, South Korea, Bangladesh and Nigeria – setting all-time annual temperature records. No parts of the world set cold records for the year.
2021 saw the warmest summer (N Hem summer – June, July, August) on record for the world's land areas:
These high temperatures – and the long-term warming trend – contributed to a number of extreme events both in the summer and across the year: carbonbrief.org/guest-post-rev…
Checking in on climate model projections (from the CMIP5 models that provide future projections after 2005), temperatures are pretty well-in-line with what models think they should be:
Note that I'm not featuring a comparison with CMIP6 models; they are less well suited to a multimodel mean approach given a subset of high-sensitivity outliers. The new assessed warming ranges (which downweight too-warm models) in the AR6 only start in 2015 making comparison hard
In the lower troposhere we saw 2021 as the 6th warmest (RSS) or 8th warmest (UAH) year on record. Note that the troposhere tends to see a larger influence of La Nina and El Nino events than the surface.
The stratosphere continues to see cooling temperatures. This is a clear fingerprint of climate change from greenhouse gases, which warm the lower part of the atmosphere by trapping heat while cooling the upper atmosphere as less heat escapes.
We saw record high sea levels in 2021. Global sea levels have risen by around 0.2 metres (200mm) since 1900, and there is evidence of accelerating sea level rise over the post-1993 period when high-quality satellite altimetry data is available.
The figure below shows the change in global average glacier mass from 1950 through to the end of 2020 (2021 values are not yet available). We see consistent loss of ice mass associated with warming temperatures:
Greenhouse gas concentrations reached a new high in 2021, driven by human emissions from fossil fuels, land use and agriculture. Methane concentrations in particular have seen a sharp rise over the past decade after a plateau in the 2000s.
Arctic sea ice was at the low end of the historical (1979-2010) range for most of 2021, but saw few new all-time daily low records set outside of brief periods in February and July. The summer minimum extent was the 12th lowest since records began in the late 1970s.
Finally, we can use current conditions (and El Nino/La Nina forecasts) to estimate where temperatures will end up in 2022. Four different groups (including a new @CarbonBrief estimate) have projections for 2022, and the differ a fair bit!
Our projection and that of the @metoffice has 2022 looking pretty similar to 2021, driven down a bit by the current "double dip" La Nina event. @BerkeleyEarth has it in the middle, while @ClimateOfGavin has 2022 threatening to top 2016 and 2020 as warmest year on record. 18/18
• • •
Missing some Tweet in this thread? You can try to
force a refresh
El Niño is coming, and it is shaping up to be a big one.
Over at The Climate Brink I've put together a compilation of the latest forecasts by different modeling groups. They suggest that we might see an event comparable in strength to what we saw in 2016.
This is based on a collection of 11 different models (and 455 individual ensemble members) all updated since the start of March. I've put an interactive version of the data up on the Climate Dashboard here: dashboard.theclimatebrink.com/#enso
While there remains a big spread in models (and some models only run through August), more than half the runs show a strong (>1.5C Nino3.4) event developing by August and a very strong event (>2C) by the end of the year.
As a rare climate scientist working in Silicon Valley, I've been drinking from the AI firehose a lot more than my peers. I thought it would be helpful to lay out my experiences of both the promise and pitfalls of using AI to accelerate scientific research.
As a bit of background, I've been working with these tools since late 2022, and seen firsthand how they have dramatically improved over time. I’ve also worked with frontier AI labs to evaluate how well LLMs answer climate questions, and to help enable AI tools to support scientific collaboration.
So what do AI tools do well for scientific work? In short, coding.
Scientists are generally not software engineers. Much of their coding is self-taught, and many struggle with writing code quickly, producing well-documented reproducible code, and fixing errors.
My new State of the Climate report over at @CarbonBrief finds that 2025 had the:
⬆️ Warmest ocean heat content
⬆️ Tied as second warmest surface temps
⬆️ Second warmest troposphere
⬆️ Record high sea level and GHGs
⬇️ Record low winter Arctic ice
Read the article here:
Ocean heat content increased by 23 billion trillion joules, which was around 39 times greater than global primary energy use this year. This is the largest rise in OHC since 2017; overall OHC has increased by over 500 zettajoules since the 1940s.carbonbrief.org/state-of-the-c…
@CarbonBrief 2025 tied with 2023 as the second warmest surface temperatures. It was nominally the second warmest in NASA and DCENT datasets, and third warmest in NOAA, Hadley, Berkeley, Copernicus, JRA-3Q, and China-MST. In all cases uncertainties overlap with 2023.
After a modest decline over the first half of the year (and after record 2024 warmth), global temperatures are ticking back up. The past two days have been the warmest on record for this time of year in ERA5 and the highest temperature anomalies since January.
With 26 days of October now reporting in ERA5, October 2025 will be the third warmest on record after 2023 and 2024.
Weather models expect global temperatures to remain relatively flat over the coming week as extreme Northern Hemisphere warmth persists, and anomalies (departures from normal) will be at or above the levels the highest levels any we've seen earlier in the year
Last week the German Meteorological Society warned that "the 3-degree limit could be exceeded as early as 2050".
While not possible to fully rule out, the assessed warming scenarios we published in the IPCC AR6 report find this to be extremely unlikely.
If we look a the full ensemble of CMIP6 models we see a small number (3 of 37 models) reaching 3C by 2050. However, these three have both too much historical warming (~2.2C in 2024) and what an unrealistically high climate sensitivity (>5C per doubling CO2) as we noted here: nature.com/articles/d4158…
However, if we constrain CMIP6 to match recent observed global temperatures, we see no models reaching 3C until at least 2060: carbonbrief.org/analysis-what-…
The EPA cited my paper in their argument against the endangerment finding today. However, their point is completely backwards: my paper actually supports the EPA's 2009 range of 1.8C to 4C warming by 2100. nature.com/articles/d4158…
Specifically, in our paper we argue that RCP4.5 or RCP6.0 are more realistic representations of 2100 warming under current policy than the increasingly implausible RCP8.5 scenario. But the lower of those two – RCP4.5 – gives a 2100 warming range of 1.8C to 4C!
Its only the high end warming outcomes of >4C that have become increasingly unlikely as the world has moved toward lower emissions scenarios. The wide range of climate sensitivity and carbon cycle feedbacks still makes it impossible to rule out up to 4C: journals.sagepub.com/doi/10.1177/29…