To no-one’s surprise 2024 was the warmest year on record – and by quite a clear margin.
Another year, another data point. Unlike the previous year, 2024 was anticipated to be a record breaker even before it began (I predicted a record – despite the huge anomaly in 2023 – with a 55% probability). It did fall at the higher end of the prediction, so maybe we are seeing the same issues that we saw in 2023 continue into another year. We shall see!
Anyway, we will be updating the climate graphs and model-observation graphs over the next week or so, but in the meantime, a couple of things are worth exploring this year are the estimates of the anomaly from the pre-industrial (nominally 1850-1900), the very large difference between the surface records and the MSU-based records.
1.5ºC or not 1.5ºC
The people have spoken, and they have collectively agreed that ‘pre-industrial’ can be thought of as the average of 1850 to 1900. There were other candidates – but the influence of IPCC AR6 is too strong to fight against. So, while I’ve been holding on to ‘late 19th Century’ (in practice 1880-1899) as a baseline, I have bowed to the inevitable and started producing anomalies with respect to the earlier baseline. But that raises a problem – how do you produce an anomaly with respect to a baseline that isn’t in your data set? This is not just a problem for GISTEMP, but it’s also an issue for the reanalyses like ERA5 and JMA-3Q.
There are a number of options. The one we chose is to make an alignment in the 19th Century over the common period in the records (1880-1899) and then use the average of three datasets that go back further to make a small adjustment. The difference between 1880-1899 in HadCRUT5/NOAAv6/Berkeley Earth and 1850-1900 is small (about 0.038ºC currently with the 1880-1899 period being slightly cooler). So for GISTEMP, you need only add 0.19 to the standard baseline anomalies (which are based on 1951-1980). If we included DCENT in this alignment, it would be a 0.022 offset, a net change of 0.20.
This method does however depend on the period over which you choose to align the records. The more recent a period, the closer the long term difference will align with the average of the three longer records and the less independent any estimate will be. The reanalyses which only go back to 1940 and 1948 respectively have fewer options. ERA5 (Copernicus) chooses to align the 1991-2020 period to the average of the three data sets. That would give an offset of 0.889ºC at present. However, this estimate was first calculated for the 2021 AR6 report, at which point the value was 0.88ºC (with a calculated uncertainty of ±0.12ºC). Copernicus has chosen to stick with this same number subsequently (even though the datasets it was based on have updated slightly since then). The semi-official ‘WMO’ assessment (which John Kennedy puts together), does something else again. They add a 0.69ºC (±0.12ºC) offset to the average of the records (GISTEMP, NOAAv6, HadCRUT5, ERA5, and JMA-3Q) baselined to 1981-2010. That too is a function of the calculation made in 2021, and if you did it again now you’d get 0.697ºC. These are minor differences, but they do underline that the pre-industrial values are more uncertain, and as datasets evolve (with more digitization of old data, updated corrections for non-climate inhomogeneitites, these estimates will evolve too.
What’s the best way to visualize all this? The uncertainty is all in the earliest period, but if you use a pre-industrial baseline that would move all the variance to the modern period. So, the alternative is to align the records in the modern period and then apply a average offset for the baseline change for all of them.
The value for 2024 in this graph is then 1.54±0.07ºC. However, this is sensitive to the inclusion of the DCENT record and to any further updates in the earlier parts of the records. There isn’t a perfect answer here, but comments on what to do (for different purposes) are welcome below.
Ummm… MSU
With much of the focus on the longer-term records, it seems to have flown under the radar a little how oddly the MSU/AMSU records have been behaving over the last year or so. As with the surface records, the satellite products (UAH, RSS, NOAA STAR) all have 2024 and 2023 as the warmest and second warmest years, but unlike the surface records, 2023 was not such a outlier (~0.06ºC above 2016), while 2024 was huge (with records broken by ~0.32ºC). Additionally, the time over which the peak temperatures have lasted (17 months or so) is much shorter than the peaks around 2016 or 1998 (7 months). I don’t have much insight into why this is happening, but it might hold some clues about the drivers of the recent anomalies.
Thanks for this excellent article.
Gavin says: ” I don’t have much insight into why this is happening, but it might hold some clues about the drivers of the recent anomalies. “End of the year, but before we start with the summaries of 2024, what does 2025 look like? These predictions are based on the long term trend plus an anomaly based on predicted ENSO for the DJ period.” https://bsky.app/profile/climateofgavin.bsky.social/post/3lemrq6jgcs2j
Please note Gavin’s graph does not even show Temps breaking the +1.5C anomaly line in 2025. Will it cool down? We will have to wait another year to find out, yet again. The same as every year.
And earlier has reported: “Gavin Schmidt @climateofgavin.bsky.social 21d
“Actual long term climate trends are being driven by a set of complex human and natural drivers that have no obligation to follow your simplistic formulae.”
https://bsky.app/profile/climateofgavin.bsky.social/post/3ldtuvcql3k2s
Excellent.
_________________________________________________________________
Summary of Climate Crisis Insights
1. Accelerating Global Temperature Increases:
Global temperatures in 2025 are starting off alarmingly high, with ocean surface temperatures already +0.25°C warmer than in 2023.
Ocean Heat Content (OHC) surged in 2024, adding approximately 15 Zettajoules (ZJ) of heat to the oceans. This equates to an energy release of 3.4 Hiroshima bombs per square mile of open water—a staggering amount.
2. Decreasing Global Albedo and Earth’s Energy Imbalance (EEI):
Earth’s albedo (reflectivity) has significantly diminished, amplifying energy absorption. Dr. James Hansen’s analysis equates this effect to a rise of over 100 ppm in CO₂ concentration.
The decreased albedo, along with greenhouse gas (GHG) emissions, has accelerated EEI, intensifying warming trends beyond what climate models predicted.
3. Oceanic and Ice Dynamics:
Despite a strong El Niño in 2024, which typically releases heat from the Pacific, the oceans have continued to absorb and retain heat.
The cooling effect of melting ice, which absorbs significant heat as latent energy, is being overwhelmed by other warming forces. Greenland alone has lost 6,000 billion tons of ice from 2002 to 2024.
The oceans are projected to warm further in 2025, indicating a persistent imbalance.
4. Forest Carbon Sink Failures:
Terrestrial carbon sinks, including forests, are failing. In 2023, forests absorbed far less CO₂, with some regions (e.g., Finland) becoming net carbon emitters. This reversal has negated decades of emissions reduction efforts globally.
5. Climate Models and Uncaptured Variables:
Observed warming (+1.6°C over baseline by 2024) has outpaced projections by mainstream climate models, exposing significant gaps:
Models fail to fully capture albedo variations and the impacts of ice-ocean interactions.
Cooling effects from La Niña cycles have not materialized as predicted.
Hansen’s 2023 projections anticipated these rapid temperature anomalies, contrasting with the underestimated risks in IPCC models.
6. The Illusion of “Net Zero”:
The failure of natural carbon sinks and increasing emissions from energy use (e.g., burning coal and oil) render the “net zero” target increasingly unattainable.
This underscores the inadequacy of current mitigation strategies in addressing the systemic and accelerating nature of the climate crisis.
7. Implications for the Future:
With the oceans heating relentlessly, a “monster” El Niño may emerge, exacerbating global temperatures further.
Hansen warns of a decadal acceleration in warming driven by aerosol climate forcing and unresolved ocean overturning dynamics.
The scientific community is grappling with models that are unable to account for sudden shifts, forcing a reassessment of predictions and strategies.
8. Call to Action:
Evidence suggests we are entering a catastrophic phase of climate change, driven by compounding feedback loops and systemic failures. Immediate and unprecedented action is required to address Earth’s energy imbalance and halt further degradation of natural climate buffers. And yet nothing is being said, nothing is happening. There is no global leadership for action of any kind.
AI ChatGTP maybe weak at calculating modelled climate and enso data but it is excellent at up to date summaries of what is. If you use it right.
One small observation:
> Cooling effects from La Niña cycles have not materialized as predicted.
>
AIUI, the cooling effect hasn’t materialised because La Niña itself didn’t materialise in 2024. It’s here now, but it’s a weak one.
Mammon 11 Jan “Thanks for this excellent article”
Specifically, which part was “ excellent“?
– the 4/5 of the article about the reference preindustrial temp. that you ignored entirely?
– the remaining 1/5, the introduction, in which Gavin points were … opposite to what you argue in your “response”?
Gavin: “ 2024 was anticipated to be a record breaker even before it began (I predicted a record – despite the huge anomaly in 2023 – with a 55% probability)
You – compliment Gavin and then … proceed to eviscerate the credibility of Gavin and mainstream science for the inability to model and therefore failure to anticipate what
…. Gavin and other HAVE anticipated.
Therefore, it seems as if it …. does not matter what Gavin actually wrote – as if you have just looked for a pretext to plant your prepared in advance list of your talking points,
rushed it in to be the first post in the discussion (so anybody reading Gavin’s article will see yours), with the opening “ excellent article” compliment to soften any resistance and to suggest to the reader that Gavin has just validated your criticism of … him and the mainstream science.
The impression strengthened by your following with:
Mammon: “Please note Gavin’s graph does not even show Temps breaking the +1.5C anomaly line in 2025
The graph you speak of ends close to + 1.6C anomaly, so … what are you talking about?
Your analysis points to only one conclusion. We have passed threshholds for synergistic positive feedbacks and only rapid cessation of fossil fuel burning can prevent a global transition. See Trajectories of the Earth System in the Anthropocene Steffen et al.. 2018 http://www.pnas.org/cgi/doi/10.1073/pnas.1810141115
Since we are definitely NOT reducing burning of fossil fuels, but in fact fossil fuel burning is still increasing, we are on track to transition rapidly to an ice free world. Absent a silver bullet solution happening very soon, such as a sufficiently better battery that rapidly makes fossil fuels obsolete, because they cost more than solar and wind with battery backup, there is a vanishingly tiny chance of avoiding a rapid transition to a world that is ten degrees C warmer than pre industrial and that transition will be rapid. This is the result, absent rapid reduction in fossil fuel production and use, that is predicted by Hansen based on multiple papers in the past decade.
A battery that does not require scarce elements like nickel and cobalt or lithium (not scarce but economically efficient sources are limited), would at a stroke do what natural gas has done to the use of coal. Render coal plants and natural gas power plants too costly to compete, so US corporations decide to close them and build new. China is another matter now with many coal plants existing and slated to be completed, and to operate another 50 years. Because their demand for power is so great and their domestic coal is cheap, now all the solar and wind and nuclear they can build simply supplements fossil fuel plants. . National security considerations are another decision driver in China. However existing coal fired plants have closed in the US since building and operating a new natural gas plant is more economical than operating an existing plant in most locations. If China’s huge productive capacity for solar were combined with better battery systems even China would rapidly cease burning coal.
There is one strategy that would offer hope, namely a very large set of prizes as well as subsidies for better batteries, both stationary for power plants and mobile for transport. Prizes plus subsidies have been effective in spurring innovation. There is no prospect of this soon given Trump’s election.
Political considerations, NOT scientific data, drive the actions of the autocracies that produce most of the fossil fuels worldwide. Expecting Putin’s Russia or the middle eastern kingdoms and dictatorships to cut production is naive. The leaders depend on that income to hold power.
“Immediate and unprecedented action is required to address Earth’s energy imbalance and halt further degradation of natural climate buffers. And yet nothing is being said, nothing is happening.”
The reason is very clear. Transportation fuels are required to take any action. Eight billion stakeholders need food and all of the materials required to complete the transition phase-in of renewables and Evs. That means more oil will be needed and not less during the lengthy phase-out.
Either way, it seems we are at the dreaded 1.5C ‘guardrail’, and poised to crash through it–if indeed we haven’t already done so–no? I know that some would say, probably correctly from a technical point of view, that you can’t really make such a call based on a couple of years. But honestly, how much of a decline are we likely to see in 2025, and how long is it likely to be before we once again exceed 1.5C? At the very least, as, IIRC, Ralph Keeling wrote about the 400 ppm milestone, we would seem to be experiencing the first ‘flickers’ of a post-1.5C reality.
Gavin says: – but the influence of IPCC AR6 is too strong to fight against. So, while I’ve been holding on to ‘late 19th Century’ (in practice 1880-1899) as a baseline, I have bowed to the inevitable and started producing anomalies with respect to the earlier baseline. But that raises a problem –
What a terrible conundrum.
I must agree though, the the influence of IPCC AR6 is too strong to fight against in every respect.
Can someone provide link to an updated temperature time series plot at what Wikipedia reports to be worlds longest running meteorological station with mention of no urban heat island effect? Plot at Wikipedia looks to end before 2010:
https://en.m.wikipedia.org/wiki/Hohenpei%C3%9Fenberg_Meteorological_Observatory
This record begins in 1781 and provides a visual of temperatures (aka kinetic energy of molecules) for decades before the 1850-1900 pre-industrial baseline.
“Ce qu’il faut faire”, c’est avant tout que la vérité sur le climat soit dite, que TOUS les avis scientifiques soient pris en compte, et pas uniquement les avis alarmistes-carbocentristes, des milliers de scientifiques ne sont pas d’accord avec les thèses du GIEC et les avis comme celui de Monsieur Gavin.
Il faut pouvoir expliquer clairement, dans le détail, comment et par quels procédés une trace anthropique de CO2, (principalement évoquée), peut contribuer à l’augmentation de la température moyenne globale et être le “bouton de commande du climat mondial”. Il faut aussi, entre autres, montrer une formule qui permettrait de donner la température à partir de la concentration en CO2, et arrêter de dire que “l’effet de serre est bien connu, sur des principes scientifiques bien établis”, en prenant comme référence des chercheurs du XIXe siècle, comme Fourier, Tyndall ou Arrhenius, qui eux, n’ont JAMAIS évoqué le terme “effet de serre” ! Et en parlant de ce dernier, il serait plus correct de dire “effet thermique atmosphérique”, car même s’il peut y avoir quelques similitudes, l’atmosphère n’est pas une serre en verre ou en plastique, et ce qui s’y passe est bien plus complexe que ce qui se passe dans une serre.
I completely agree with you that we are facing a climate crisis and as the Los Angeles catastrophe is showing we as a society are not remotely prepared. I have recorded the local weather in my part of Canberra Australia for 33 years. The occurrence of above average temperatures has increased markedly in the last decade. We have less frost days than 30 years ago which is affecting our fruit trees which need frost days to produce a good crop. Last year a very late severe frost wiped out our peaches, apples and apricot crops and impacted citrus and figs too. The weather is far more changeable these days with extreme rain events quickly followed by snap droughts. We have seen flooding rains in Queensland while the rest of the country is experiencing heatwave and drought!
When are policy makers going to wake up and act or will we see repeats of Barcelona floods or Los Angeles bush fires the like of which has never been recommended before?
Gavin O’Brien FRMetS
> the time over which the peak temperatures have lasted (17 months or so) is much
> shorter than the peaks around 2016 or 1998 (7 months)
s/shorter/longer/ ?
“The people have spoken, and they have collectively agreed that ‘pre-industrial’ can be thought of as the average of 1850 to 1900.”
What was that average value…and what should be added to that number to obtain the current value? NOAA says that the 20th century average is 13.9°C. 57°F.
About the different behavior of MSU vs. Surface: MSU data usually show a much stronger deviation during El Niño events (especially for strong ones) than surface temperature. A possible reason might be that the heat over the tropical Pacific is spread over a larger area in higher layers (e.g. by the zonal circulation over the tropics and the Hadley cell) than at the surface (just guessing …).
And a second observation concerning El Niño that struck me was the unusual difference between the Nino3.4 index and the MEI index during the recent Niño event: MEI index was only about half of Nino3.4 and started later, while during earlier strong Niños the two indices were very similar. Since the MEI index involves a much larger region and other variables than Nino3.4 this might point out that the larger-scale atmosphere (and ocean) over the Pacific was in a rather unusual state during 2023/2024.
However, this doesn’t give a clearer picture of “what has been (or still) is unusual and why” either … maybe just another piece of the puzzle.
Dear Dr. Schmidt,
As a layman, I had to Google what does the acronym MSU / AMSU actually mean.
Perhaps you could, for the sake of clarity, any time you use such terms that may be unfamiliar for a broader public, add (analogously as did ChatGPT in the comment of 11 Jan 2025 at 7:47 PM above) also a “full text” reading of the acronym.
Or, alternatively, an explanatory link, e.g.
https://www.remss.com/missions/amsu/
Thank you in advance and best regards
Tomáš
In 1875 an Icelandic volcano named Askja had a wet eruption. What followed was a large temperature spike which, so far, is almost identical to the one following the HT volcano. The similarities cannot be ignored. I can’t find any evidence that this connection has ever been discussed in this forum.
https://localartist.org/media/HTvAkjsa.png
I plan on monitoring any discussion on this, but don’t plan on participating further on this topic. Water vapor injected at low, or high latitudes into the stratosphere is not really an area I’m qualified to discuss. However, I would not be surprised if we return to 2016 temperatures by early 2026 and stay there, or even cool slightly for the next decade.
Er, except it clearly does, if you are talking about the second one shown–the one with the 1981-2010 offset. And if the helpful dotted 1.5C line isn’t enough, then you have the accompanying text which says, “The value for 2024 in this graph is then 1.54±0.07ºC.”
In hindsight, it appears that Edgar saved countless lives:
https://localnewspasadena.com/2025/how-two-words-from-a-24-year-old-pasadena-climate-specialist-saved-hundreds-of-lives/
Thanks E McG