50-Year U.S. Summer Temperature Trends: ALL 36 Climate Models Are Too Warm

From Dr. Roy Spencer’s Global Warming Blog

Roy W. Spencer, Ph. D.

I’ll get right to the results, which are pretty straightforward.

As seen in the accompanying plot, 50-year (1973-2022) summer (June/July/August) temperature trends for the contiguous 48 U.S. states from 36 CMIP-6 climate model experiments average nearly twice the warming rate as observed by the NOAA climate division dataset.

The 36 models are those catalogued at the KNMI Climate Explorer website, using Tas (surface air temperature), one member per model, for the ssp245 radiative forcing scenario. (The website says there are 40 models, but I found that four of the models have double entries). The surface temperature observations come from NOAA/NCEI.

The official NOAA observations produce a 50-year summer temperature trend of +0.26 C/decade for the U.S., while the model trends range from +0.28 to +0.71 C/decade.

As a check on the observations, I took the 18 UTC daily measurements from 497 ASOS and AWOS stations in the Global Hourly Integrated Surface Database (mostly independent from the official homogenized NOAA data) and computed similar trends for each station separately. I then took the median of all reported trends from within each of the 48 states, and did a 48-state area-weighted temperature trend from those 48 median values, after which I also got +0.26 C/decade. (Note that this could be an overestimate if increasing urban heat island effects have spuriously influenced trends over the last 50 years, and I have not made any adjustment for that).

The importance of this finding should be obvious: Given that U.S. energy policy depends upon the predictions from these models, their tendency to produce too much warming (and likely also warming-associated climate change) should be factored into energy policy planning. I doubt that it is, given the climate change exaggerations routinely promoted by environment groups, anti-oil advocates, the media, politicians, and most government agencies.



And here’s a musical coda from The Specials: Too Hot. — charles

4.8 18 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
Mike Lowe
October 20, 2022 10:26 pm

So instead of criticising the inaccuracy of these model results, I bet the “authorities” wish they would actually eventuate during the approaching winter. Unfortunately for them, their inaccuracy will become apparent!

Reply to  Mike Lowe
October 21, 2022 2:30 am

But the general populace won’t know about the innaccuracy because the MSM won’t report it, so the “authorities” will “get away with it” as usual.

Reply to  Mike Lowe
October 21, 2022 8:30 am

It’s never unfortunate for them because they’re never taken to account for their inaccuracies. So they just keep doing it again and again.

October 20, 2022 10:46 pm

While I never tire of articles about inaccurate climate models, this is a 40 year old story. These models, on average, have been over predicting global warming for many decades. In fact, that’s their purpose: To scare people. Accurate predictions are not a goal. We know that because the least inaccurate model, the Russian INM model. gets no attention. It should get 99% of the attention!

The fous on summer months is somewhat biased.
Greenhouse warming mainly affects TMINs in the colder months of the year.
So the models are likely to be less inaccurate for Winter months when there is more global warming. Not that the models are accurate for any months of the year.

I recommend this theme song for the models:

Ella Fitzgerald Too Darn Hot – YouTube

Reply to  Richard Greene
October 20, 2022 10:58 pm

“I recommend this theme song for the models:”

Or this:

Bob Bailey
Reply to  Richard Greene
October 22, 2022 7:12 am

Song was released in 1948. Ella was psychic! Her music is a regular feature on my Pandora.

October 20, 2022 10:47 pm

For clarity can you explain the difference between cmip and rcp. I often hear that extremists use rcp 8.5 to make their case and that rcp 8.5 is unrealistic. Is there a relationship between cmip and rcp?

Dave Fair
Reply to  Bob
October 21, 2022 12:11 am

CMIP (Climate Model Intercomparison Project) refers to an ongoing collaboration of climate modeling groups to contrast and compare the models, using some standard guidelines. It doesn’t really mean anything because the models vary widely in their climate metric outputs and computed ECSs (Equilibrium Climate Sensitivity of green house gasses). Despite decades and multi-billions of dollars going into the climate modeling community, they can’t even get the lack of a tropospheric hot spot right.

RCP refers to Representative Concentration Pathways, mental masturbation about how the various ways in which the Earth will develop economically and technologically. The numbers in each pathway’s designation is the assumed net forcing in W/m^2 of that pathway in 2100. The massive outlier is RCP8.5 which assumes most energy produced over time comes from coal. Suffice it to say that there is not enough coal in the ground to ever reach such levels.

The takeaway is that CliSciFi in its entirety is made up of venial liars. FJB, F UN and F Leftists.

Reply to  Dave Fair
October 21, 2022 10:34 am

Each of the RCP’s result in different CO2 levels during each year until 2100, and it is the CO2 levels that are supposed to be doing the forcing. When looking just at the atmospheric CO2 levels regardless of source, RCP 8.5 is on track and a legitimate way way to compare models to reality.

Reply to  Ted
October 22, 2022 5:46 am


h/t to stokes

Reply to  Yirgach
October 22, 2022 10:42 am

RCP 8.5 is the least wrong.

Last edited 3 months ago by MarkW
Mark BLR
Reply to  Ted
October 22, 2022 9:39 am

When looking just at the atmospheric CO2 levels regardless of source, RCP 8.5 is on track

No it isn’t.

Note that the actual, measured, CO2 levels fell below all of the CMIP6 (SSP) “projections” in 2018, two years before the “COVID dip” in emissions.

Mark BLR
Reply to  Mark BLR
October 22, 2022 9:46 am

Follow-up …

Several months ago I checked all twelve “Month X+1 of Year Y-1” to “Month X of Year Y” options for the Mauna Loa CO2 data to see which was the best match to the CMIP6 “Historical Data” numbers.

I ended up choosing “September to August” for the “MLO-prime” (thick black) line in my OP …

Reply to  Mark BLR
October 22, 2022 7:00 pm

The subject wasn’t the altered CMIP6 projections. The projection from the original RCP 8.5 was 415 ppm in 2020. Other RCP’s were at 409 to 412.


Moana Loa average was 414:


Mark BLR
Reply to  Ted
October 24, 2022 6:00 am

The projection from the original RCP 8.5

… is shown as the dashed red line in my OP.

I included the RCP lines to allow easy comparisons between CMIP5 (“Historical Data” to 2005), CMIP6 (“Historical Data” to 2014) and the actual empirical measurements (Mauna Loa, “tweaked” to calibrate them to the IPCC’s numbers) since those dates.

The projection from the original RCP 8.5 was 415 ppm in 2020. Other RCP’s were at 409 to 412.

Moana Loa average was 414

The numbers I downloaded from Malte Meinshausen’s “RCP Concentration Calculations and Data” webpage at the PIK (direct link) give 415.78 ppm in 2020 for RCP 8.5.

The “September to August averages” I used to “align” the Mauna Loa numbers to the (global ?) CMIP5 and CMIP6 “Historical Data” numbers allow me (/ us …) to calculate, and compare, all the way to “2022”.

RCP 8.5 : 421.86
RCP 6.0 : 413.22
RCP 4.5 : 415.64
RCP 2.6 : 416.52
MLO’ : 417.8

The current atmospheric CO2 concentration measurements are closer to the “original” RCP 2.6 and RCP 4.5 “projections” than they are to RCP 8.5.

Last edited 3 months ago by Mark BLR
Dave Fair
Reply to  Ted
October 22, 2022 1:50 pm

Do any of you people not know that RCP 8.5 has coal providing the bulk of the worlds electricity?

Dave Fair
Reply to  Ted
October 22, 2022 7:02 pm

Additionally, it is recognized that the top two emission scenarios are unlikely. Do a little research.

Reply to  Ted
October 23, 2022 11:42 am

Ted, it seems to me that RCP is a measure of how much CO2 will be added to the atmosphere under different government policies. It appears that 8.5 represents no government policy to lower CO2 emissions. The other lower numbers represent the success or not of the different government attempts to lower emissions. I guess my problem with this effort is labeling them with watts per meter squared. That is not what is being measured, the concentration of CO2 is what is being measured. The effect of the added CO2 is unknown. Labeling the different pathways with a number such as 8.5 or 6 or whatever would suggest we know the strength of CO2 as a greenhouse gas. I don’t think we do or if we do we know it can not raise average global temperatures on it’s own but causes positive feedback that raises temperatures. Negative feedback seems to be ignored or simply given no level of importance. Can RCP give us a measure of CO2 concentration? Yes, it can give us an idea of how much CO2 man is adding to the atmosphere but it says nothing about CO2 added to the atmosphere by nature. All pathways need to be labeled correctly in other words parts per million CO2 concentration. The watts per meter squared increase is dishonest and meaningless gibberish. in my view.

Reply to  Bob
October 21, 2022 1:42 am

My comic book analysis:
RCP — CO2 emissions growth rate predictions baloney
CMIP6 — climate computer game temperature predictions baloney
RCP and CMIP6 have no connection with reality
They have baloney in common.

Mark BLR
Reply to  Bob
October 21, 2022 4:05 am

Editorial note : Unfortunately “clarity” requires “length” … which reduces the overall “clarity”.

Apologies for the length of this post in advance, but hopefully it will make the differences between CMIP, RCP and SSP “clear” enough for you (and everyone else).

– – – – –

For clarity can you explain the difference between cmip and rcp.

They are related but independent concepts.

As “Dave Fair” noted, CMIP = Climate Model Intercomparison Project, of which there were 3 main rounds.

CMIP3 used the SRES (Special Report : Emission Scenarios) set of “scenarios” as inputs to the climate models, and reported the results in the TAR (2001) and AR4 (2007) IPCC reports.
NB : For AR4 the climate models had “improved / evolved”, but exactly the same inputs were used as for the TAR. The end results were very similar.

The most cited SRES scenario was “A1FI”, or the (worst case) “Fossil-fuel Intensive” one.

– – – – –

CMIP5 used the RCP (Representative Concentration Pathways) set of “pathways” as inputs, and reported the results in AR5 (2013).

The most cited RCP pathway was (and is) “RCP8.5”, which gave an increase in total GHG radiative forcing relative to “pre-industrial” levels of +8.5 W/m² in the year 2100.

Note that both SRES and RCP simply “dictated” GHG emission and atmospheric levels over time, with no consideration of how (un-)likely the “pathway” being described might be in the real world.

– – – – –

The CMIP6 round, however, used the SSP (Shared Socio-economic Pathways) set of “pathways” as inputs, and reported the results in AR6 (2021/2).

These attempted to include human behaviour while determining just how “likely” their various “storylines” actually were.

The AR6 (WG-I) report has a specific sub-section exploring “The likelihood of reference scenarios, scenario uncertainty and storylines” (, on pages 238 and 239 of the May 2022 “Final / Approved” version), which includes the following :

Among the five core scenarios used most in this report, SSP3-7.0 and SSP5-8.5 are explicit ‘no-climate-policy’ scenarios (Gidden et al., 2019; Cross-Chapter Box 1.4, Table 1), assuming a carbon price of zero. These future ‘baseline’ scenarios are hence counterfactuals that include less climate policies compared to ‘business-as-usual’ scenarios – given that ‘business-as-usual’ scenarios could be understood to imply a continuation of existing climate policies. Generally, future scenarios are meant to cover a broad range of plausible futures, due for example to unforeseen discontinuities in development pathways (Raskin and Swart, 2020), or to large uncertainties in underlying long-term projections of economic drivers (Christensen et al., 2018). However, the likelihood of high emission scenarios such as RCP8.5 or SSP5-8.5 is considered low in light of recent developments in the energy sector (Hausfather and Peters, 2020a, 2020b). Studies that consider possible future emission trends in the absence of additional climate policies, such as the recent IEA 2020 World Energy Outlook ‘stated policy’ scenario (International Energy Agency, 2020), project approximately constant fossil and industrial CO2 emissions out to 2070, approximately in line with the medium RCP4.5, RCP6.0 and SSP2-4.5 scenarios (Hausfather and Peters, 2020b) and the 2030 global emission levels that are pledged as part of the Nationally Determined Contributions (NDCs) under the Paris Agreement (Section 1.2.2; (Fawcett et al., 2015; Rogelj et al., 2016; UNFCCC, 2016; IPCC, 2018).

NB : the “new and improved” SSP3-7.0 and SSP5-8.5 pathways neatly bracket the “old” RCP8.5 pathway, so the IPCC is here effectively admitting that RCP8.5 should also be considered as a “counterfactual” emissions scenario.

– – – – –

I often hear that extremists use rcp 8.5 to make their case and that rcp 8.5 is unrealistic.

The new “worst case” pathways, one where no additional climate policies are implemented between now and 2070 ( / 2100), should be the “new” SSP2-4.5 and the “old” RCP4.5 ones.

The “official, IPCC approved” adjective is “counterfactual”, but however “unrealistic” RCP8.5 may be we are unlikely to see the “extremists” transition from hyperventilating about the “catastrophic” outputs of computer models that use RCP8.5 or SSP5-8.5 as inputs in our lifetimes.

To get an idea of just how bat-shit crazy these “counterfactual” (/ unrealistic) pathways are, look where they end up in the 23rd century, just by assuming that human beings would do absolutely nothing if CO2 levels were to “inexorably” rise into 4-digit territory.

Reply to  Mark BLR
October 24, 2022 11:16 am

Thank you.

October 20, 2022 10:57 pm

Warmists lie. Who’d a thunk it?

October 20, 2022 11:08 pm

“36 CMIP-6 climate model experiments”

No!. It’s one of the other.
Models are NOT experiments.

Calling them experiments is just more newspeak.
They are just “climate model runs”

Last edited 3 months ago by StuM
Michael 63
Reply to  StuM
October 20, 2022 11:52 pm

Model is more of a static thing – like a model airplane IMO.
Simulation is representation of “something” over time – used to be called testing. Again my understanding.
Simulation also implies “not-to-scale” and simplified 🙂 Very fitting for climate models, those aren’t to scale neither in maco nor micro. And they are simplified – eg phase change of water is very poorly represented if at all (“parameterized” – aka “fiddled for desired result”).

Reply to  StuM
October 21, 2022 1:44 am

They are computer games used for climate propaganda
Except for the Russian IMN, they are too inaccurate
to be called “models” of the climate on this planet.

Reply to  StuM
October 21, 2022 9:56 am

I would say they are experimenting with climate models and the experiments show that, except for one or two, they are not worth anything and should be trashed. That is a worthwhile conclusion from any experiment, except none of the organizations with the garbage results are willing to admit their model is garbage, and try building something that matches reality.

Dave Fair
October 20, 2022 11:53 pm

You do realize that NOAA, NASA, UK Met Office, Australian BOM, UN etc. ad infinitum are intimately familiar with this data, don’t you? All governments lie to you. In the West we used to accuse the Communist countries of lying to their populace. The commies couldn’t hold a candle to Western governments, NGOs, Meta (Facebook), Twitter, most all cable news and every major newspaper in blue U.S. cities.

This will not end well. Already there are rising groundswells of opposition. I’m on the civilian side that has the weapons. FJB and F Leftists.

Peta of Newark
Reply to  Dave Fair
October 21, 2022 12:42 am

Yes it will end well, when, as trends presently show, all chidren born in the western world will be autistic.

Autistic People Don’t Tell Lies
They simply don’t see the point of doing so

Another, some may say, odd thing about them is that they very rarely (i’ve yet to encounter one, drink alcohol or do any variation on Recreational Drugs.
That includes Trash TV and Main Stream Media

Expanding slightly on my current ‘bee in me bunnet’ and absolutely repeating myself:

re: That Government is now Too Big
(Drastic and pigs will be on transatlantic flight routes but will fix everything at a stroke)

From this moment forward, all folks who:

  • Is involved in politics in any way
  • Is political adviser or pundit
  • Is scientist, professor or teacher
  • Is police, doctor or senior nurse (ones who can prescribe medication)
  • Has any contact with folks under the age of 18
  • (I may have missed a few but, you get the gist?)

From this moment on they will:

  • Be certifiable T-Total as willing able to prove it at any time
  • Not be in any way obese or overweight (lets go for a BMI of less than 27.5)
  • Not watch any more than 60 minutes of TV per week

Because those are the liars.
In all those people, the words they utter are ALWAYS guided by an uncontrollable subconscious desire to get to their next plate of pasta (i.e. sugar) or glass of alcohol as soon as (in)humanly possible. This was Boris Johnson PM

And they routinely lie about that, not least because everyone around them also lies.
And once you start, the porkies, just like ‘government’, the porkies ‘creep’, ‘expand’ and grow in magnitude and extent

Certainly in the Western World, I’ve just ‘cancelled’ 80% of everybody over the age of 15, ain’t I just?

All the best ideas are really simple – then no-one can quibble.

Unless they’re liars
neat huh

Wei Zhang (Minnesota)
Reply to  Peta of Newark
October 21, 2022 8:34 am

Thanks. I understand everything now.

Reply to  Wei Zhang (Minnesota)
October 22, 2022 10:58 am

That makes one of us.

Reply to  Peta of Newark
December 23, 2022 1:23 am

Autistic people do lie sometimes. I knew someone once, autistic, low IQ, he could barely read, but he pulled a new girl every few days. Because he was very sweet, he bought gifts, he played into the fantasies of women who were fed up with unreliable boyfriends that someone simple couldn’t be deceitful. After a few days, he got bored, and started over again with a new woman.

Julian Flood
October 21, 2022 12:25 am

Dr Spencer,

Your analysis shows two things – the warming of the US is exaggerated, yes, but also that warming is happening. A particular interest of mine, having observed oil/surfactant smoothing of water bodies such as the Red Sea, North Sea, Mediterranean, Broad Lake in the grounds of UEA* is what this is doing to the surface temperatures of these. I make a Feynman guess that the lowered albedo and reduced. evaporation of oil/surfactant polluted water will cause warming.

There is a paper I believe – I’m away from my computer so I can’t reference it – – on the Eastern Mediterranean stratifying. The Sea of Marmora is warming at two to three times the rate of open ocean.

Could you please take a look?


Reply to  Julian Flood
October 21, 2022 12:55 am

“warming is happening”

Nooo, warming HAS happened, it requires a time machine to say that it “is” happening.

Tom Abbott
Reply to  climanrecon
October 21, 2022 5:12 am

The climate has cooled significantly over the last few years:

comment image

Robert Hanson
Reply to  Tom Abbott
October 21, 2022 3:14 pm

Yeah, but if you start at the bottom of the LIA, it’s been warming ever since.

AGW is Not Science
Reply to  Robert Hanson
October 21, 2022 6:42 pm

And if you start at the Holocene Climate OPTIMUM, aka the WARMEST AND THEREFORE BEST climate of the current epoch, the long term trend is down, down, down.

Ireneusz Palmowski
Reply to  Julian Flood
October 21, 2022 1:16 am

A 0.2 degree warming? This is very beneficial for vegetation.

Mark BLR
Reply to  Julian Flood
October 21, 2022 5:00 am

Your analysis shows two things – the warming of the US is exaggerated, yes, but also that warming is happening.

The ATL article includes a link to the NCEI website allowing downloading of all the CONUS data (from 1895).

It’s easy enough to plot all the “summer” (JJA) averages.

Why were the (absolute) “warm” temperatures in the 1930s of roughly the same amplitude as “now”, given that atmospheric CO2 levels back then were in the 305-310 range ?

Why did the 50-year (trailing) trends decrease from (1904-)1953 to 1979, even as CO2 levels were increasing from ~295 ppm (in the 190x decade) to ~335 ppm (in 1979) ?

AGW is Not Science
Reply to  Mark BLR
October 21, 2022 6:43 pm

Because atmospheric CO2 level IS NOT THE DRIVER of the Earth’s temperature.

Just like ALWAYS.

Ireneusz Palmowski
October 21, 2022 1:18 am

In the coming days, a wave of Arctic air will fall over the western US and the Rocky Mountains.
comment image

Carlo, Monte
Reply to  Ireneusz Palmowski
October 21, 2022 7:59 am

Hardly unusual for late October.

Ireneusz Palmowski
October 21, 2022 1:19 am

Does the ozone hole over Antarctica remain large? Yes, it remains.
comment image
comment image

Ireneusz Palmowski
October 21, 2022 1:20 am

Has solar activity declined again? Yes, it has decreased.
comment image
comment image

Ireneusz Palmowski
October 21, 2022 1:31 am

When will La Niña end? I think it’s just beginning.
comment image

October 21, 2022 1:47 am

36 models all giving results ABOVE the actual measured date is not only evidence of incompetence, it is evidence of intent to commit fraud. You can’t stuff up 36 times, without thinking perhaps the whole basis for our models is wrong.

AGW is Not Science
Reply to  Harves
October 21, 2022 6:45 pm

It is wrong – because it assumes adding CO2 causes warming.

Ireneusz Palmowski
October 21, 2022 1:51 am

If you look at sea surface temperatures, you might wonder if the warming trend is increasing.
comment image

October 21, 2022 1:52 am

The Climate Howlers focus on the global average temperature.
This article focuses on the US summer temperatures
That’s data mining and makes no sense.

If one wanted to show how inaccurate climate models are, it would have been better to data mine the USCRN average temperature since 2005, which only has a tiny uptrend.

If one wanted to be fair, the global average temperature should have been used, NOT the US Summer temperature.

Krishna Gans
Reply to  Richard Greene
October 21, 2022 6:01 am

For US policies US temperatures are relevant.

Jim Gorman
Reply to  Richard Greene
October 21, 2022 6:30 am

The real point is that if one large land area, that is well measured, is below the global average obtained from models, then where are temperatures that are well above the average in order for the average to be true? The modelers should be asked to reveal where these locations are! Hiding behind the excuse that they don’t calculate this is unresponsive and displays a lack of scientific rigor.

Dale S
Reply to  Richard Greene
October 21, 2022 6:57 am

This article specfically mentions the use of models in U.S. Energy Policy. It’s a rational view IMO that how the models have done in predicting CONUS temperature trends is more relevant to U.S. impacts than how the models have done in predicting GAT. There’s also the side bonus that CONUS temperature measurement has *much* better coverage than worldwide temperatures.

For CONUS, it is only summertime warming that could produce net negative impacts.

Of course, it’s also true that energy policy of the United States and other temperate first world nations haven’t prioritized local impacts. But politicians also show a frightening degree of either abject ignorance or wilful deceit when discussing global impacts, so it hardly matters. You’d never guess from the media that the *actual* impact of even the unrealistic RCP 8.5 is projected to be that the much richer world of the future will be somewhat less rich. The president of the USA says with a straight face that climate change is an “existential threat”, and yet those focused on “climate misinformation” totally ignore such dangerous ignorance and concentrate on those claiming that warming is nonexistent, when it’s really just undetectable without instruments.

Michael Jankowski
Reply to  Richard Greene
October 21, 2022 3:46 pm

A few points.
(1) the climate howlers often focus on regional or even local temperatures. There are countless articles about “X is warming twice as fast as the rest of the globe.” The latest big local climate howls is the cancellation of crab seasons in Alaska.
(2) showing large scale model failure is an important part of showing that the global models are wrong. If a model gets regional, continental, or large national (e.g., US) areas wrong but sums up to being close to “right” on a global scale, that just shows the “right” answer was arrived via intermediary garbage.

Ireneusz Palmowski
October 21, 2022 2:37 am

How can Venus heat up from the surface if only 2.5% of the available solar radiation reaches it?
“First global map of Venus nightside surface temperature using Akatsuki infrared measurements reveals hot Venus surface with an average surface temperature of about 698 K. Surface temperatures do not show any significant variation with changing latitudes because only a small amount (~2.5%) of solar energy reaches the surface. Surface temperatures are relatively colder at higher altitude regions as compared to lower altitude regions.”
“Venus surface temperatures indicate a hot (~698 K) with little to almost no interaction with solar energy. Assuming a solar constant of 2600 Wm−2, and 2.5% absorption by the surface the dayside temperature would be higher by about 1–2 K than nightside temperature.”
“The global average nightside surface temperature of Venus is about 698 K with a spatial variation of over 230 K. Due to very thick and reflective atmosphere Venus surface absorbs about 2.5% of incident solar flux as compared to about 50% on Earth.”

Ireneusz Palmowski
Reply to  Ireneusz Palmowski
October 21, 2022 2:56 am

On Venus it is apparent that the denser the atmosphere due to its chemical composition and gravity, the more it absorbs the solar energy reaching the planet.

Don Vickers
Reply to  Ireneusz Palmowski
October 21, 2022 5:00 am

The atmospheric pressure ?

Krishna Gans
Reply to  Don Vickers
October 21, 2022 6:35 am

As the pressure is constant, that doesn’t change the temperature.

Ireneusz Palmowski
Reply to  Don Vickers
October 21, 2022 8:22 am

Gravity is the main player, pressure is secondary. The denser the atmosphere, the more particles collide.

October 21, 2022 4:17 am

Given that climate is a perfect example of the Rumsfeld doctrine – known knowns, known unknowns etc – and that everybody, especially the plasticine modellers, know that they are no better than the blind men and the elephant, why would their models be correct?

It’s, er, pretty much impossible.

Imagine a car, it has many components, systems, mechanisms – even door handles.

But if you only had one or two bits of it how could you really model it, let alone model how it works?

The honest answer is you can’t.

Reply to  strativarius
October 21, 2022 1:58 pm

But some of our favourite computer gamers – CliSci modellers, to their families – think they have half a dozen components, including the cigar lighter and, maybe a floor mat or two.

And we are to dismantle our civilisation on their vapourings…… ??

Auto, coloured unimpressed.

John Shotsky
October 21, 2022 4:59 am

The one thing these models have in common is that they all use CO2 to predict climate. CO2 has nothing whatever to do with climate. Ice cores show that CO2 follows temperature changes, so CO2 is a red herring. It is a RESULT, not a CAUSE of climate change. The earth is 70+% covered with water, and when water temperature rises, as it does after an ice age, it releases CO2. It is a slow process. We are only 170 years away from the end of the little ice age.
In spite of decades of evidence that the models are wrong, they (modelers) steadfastly believe that CO2 is the climate control knob and build it into their models. All they have done is prove that the models are wrong.
At this point, we probably have enough climate history to run those models with CO2 backed out and see what happens. Even with all the fudged temperature readings, I’d suggest that the models will all come closer to reality. None of them would DARE to try that, as it would make them all look silly.

Rod Evans
October 21, 2022 5:10 am

Hey, come on guys. Who needs actual observations when you have got modelled temperatures to rely on….

James F. Evans
October 21, 2022 5:49 am

This is a theme I’ve been hammering on.

“Almost all the models have been wrong.” (All of them in this example.)

This is a response that can be made in conversation with an AGW proponent.

(And their eyes won’t glaze over. It’s short, to the point, and easy to understand.)

October 21, 2022 6:06 am

So, there were no dancing angels? tsk, tsk

Jim Gorman
October 21, 2022 6:36 am

Here is a study about Chinese temperatures being warmer than actually measured.


Sweet Old Bob
October 21, 2022 6:57 am

So , what are summer HIGH temperatures doing ?
Last I saw , they were decreasing .
For last 80 years ? (east of Rockies )

Wei Zhang (Minnesota)
Reply to  Sweet Old Bob
October 21, 2022 8:30 am

Yes. Almost all the Eastern USA has had fewer days >95, >100, >105. The West slightly more but overall USA fewer.

James Allen
October 21, 2022 7:21 am

Off topic a bit, but kudos to whomever is selecting the headline artwork today. The “painterly” illustrations look great.

Reply to  James Allen
October 21, 2022 10:15 am

thank you.

but not one person commented on my musical selection 🙁

Last edited 3 months ago by Charles Rotter
Geoff Sherrington
Reply to  Charles Rotter
October 22, 2022 4:23 am

For my taste, not much good new music was composed after 1900.
Oh for more of the mastery of Beethoven.
Geoff S

Wei Zhang (Minnesota)
October 21, 2022 8:20 am

Important to note that these is not a FORECAST verification. Most of the time period (1973-2022) is a HINDCAST. So, the models are already wrong on the backtest without including UHI impact. When did the actual forecast start? 2006?

October 21, 2022 10:08 am

Because of the many well documented problems with UHI and micro-site contamination, the so called observed temperature ends up being higher than the actual temperature.

Johne Morton
October 21, 2022 11:02 am

Not only that, but how much warming is spurious already (UHI, site changes etc)?

Dave Fair
Reply to  Johne Morton
October 22, 2022 2:27 pm

Anthony Watt’s Surface Station Project (latest 2022) will give that answer for CONUS.

October 21, 2022 11:34 am

Aren’t there more than 36 models in CMIP6? What about the full model population?

October 21, 2022 12:57 pm

@moderator Rotter
How long dioes it take to generate those AI graphics? Some are very related.
Understand the need for infringement problems. Useful tech.

October 21, 2022 1:15 pm

Joe Biden’s climate emergency proclamation is two years old. According to NOAA it is 0.052 deg C warmer on average now than when he took office. I’m pretty sure his 3 trillion climate emergency bill will be spent judiciously to prevent any more of this disastrous heating.

October 21, 2022 2:16 pm

The biggest problem with computer models is getting them to matchup to reality.

michael hart
October 21, 2022 2:36 pm

The modellers done too much, much too young.

Gary Kerkin
October 21, 2022 3:29 pm

And the elephant in the room is? … 0.1º change per decade is NOT alarming..

Reply to  Gary Kerkin
October 21, 2022 4:46 pm

For the US, it’s around 0.26C over 50 years. Closer to 0.05C per decade.

Reply to  MarkW
October 21, 2022 6:07 pm

It’s 0.26°C / decade. About 1.3°C over 50 years.

Geoff Sherrington
Reply to  Bellman
October 22, 2022 3:37 am

Before you quote a change of anything over 50 years, you need to understand the measurement uncertainty 50 years ago, like 1970. In their 2020 State of the Climate report, BOM claim
Australia’s climate has warmed on average by 1.44 ± 0.24 °C since national records began in 1910″
Apart from national T records actually going back to 1856, the BOM appear to think that uncertainty has been constant and symmetrical since 1970 or so.
One small example disproves this. Metrication was on 1 Sept 1972. BOM estimate about 0.1 to 0.15C of change from metrication, when rounding from Fahrenheit to Celcius. They do not include this error in their figures because they claim inability to distinguish it from the Great Pacific Climate Shift of the mid-1970s. So the error remains. It is not symmetrical, it is only towards seeming warmer, so their +/- expression is wrong. It is also (maybe) 0.15 of the 0.24 total uncertainty, a guesstimate by the BOM of uncertainty that is quite implausible, a product of choice of certain statistics that seem inappropriate as well as weightings for land ar4eas in claculating national figures.
When you account for all of the irregularities, not just metrication, you get uncertainty much larger than 0.24. (All numbers here are 2 sigma).
A group of us who have worked on these numbers from the late 1980s have considerable familiarity with the uncertainty. There is no consensus number, but many of us seem not far away from drawing a line through 1.44 +/- 0.24 C and a better estimation of numbers giving 0.6 +0.7/-0.4C for Australian land Taverage change since 1910. (It is less if we start in 1890). There are, as yet, unexplained large changes since year 2000, as colleague Chris Gillham explains in link 3 below.
Geoff S
Uncertainty Of Measurement of Routine Temperatures–Part Three – Watts Up With That?

Reply to  Geoff Sherrington
October 22, 2022 2:40 pm

I’ve asked what Dr Spencer considers the uncertainty to be of this 50 year trend. For some reason nobody seems to object to quoting the trend with no confidence interval as long as the claim is that there has been less warming than the models predict.

But when I correct a trivial factual error , that the warming quoted is 0.26 per decade, and not over 50 years, suddenly it’s back to measurement uncertainties.

By my calculations, and nothing to do with measurement uncertainties, the 95% confidence interval for the 50 year trend is around ±0.09°C / decade.

Dave Fair
Reply to  Geoff Sherrington
October 22, 2022 6:48 pm

Your reference to Part Three of your uncertainty work in WUWT is really good. It is a prima facie case of BOM malfeasance. People that haven’t ought to review it, Geoff.

Reply to  Bellman
October 22, 2022 11:04 am

Not according to the data.

Reply to  MarkW
October 22, 2022 2:31 pm

What data would that be?

I’m talking about the data that is used by Dr Roy Spencer in this very article. It’s clearly stated in the graph at the top that the warming rate is 0.26°C / decade, and not 0.26°C / half century. You can check the source yourself.

If you think this is wrong, take it up with Dr Spencer, not me.

Jim Gorman
Reply to  Bellman
October 22, 2022 5:38 pm

From Dr. Spencer’s blog.

Reply to  Jim Gorman
October 22, 2022 6:41 pm

The official NOAA observations produce a 50-year summer temperature trend of +0.26 C/decade for the U.S., while the model trends range from +0.28 to +0.71 C/decade.

From Dr. Spencer’s blog.

Mike Maguire
October 21, 2022 8:29 pm

Almost half of the all time hottest temperatures in the 50 US states were set in the 1930’s.


Screenshot 2022-10-21 at 22-24-45 June 2021 historic heat W US - MarketForum.png
Mike Maguire
Reply to  Mike Maguire
October 22, 2022 7:24 am



“The IPCC Fifth Assessment Report predicted 0.5080.102 Wm−2RF resulting from this CO2 increase, 42% more forcing than actually observed. The lack of quantitative long-term global OLR studies may be permitting inaccu-racies to persist in general circulation model forecasts of the effects of rising CO2 or other greenhouse gasses.”

Geoff Sherrington
October 21, 2022 11:49 pm

Here is a related data presentation about Australia.
Our CSIRO government research agency has 2 models in this CMIP, named ACCESS CM2 and ACCESS ESM1.5. On Roy’s bar chart comparison their anomaly temperatures for the US are 0.52 and 0.57, so they come in amongst the highest projections of the 36.
CSIRO produced a CMIP6 comparison shown in this figure showing CSIRO’s Australian data:
I have added some of theie text to the figure caption, plus the link to the larger CSIRO description.
I have also added as big green asterisks the UAH satellite temperature anomalies for the years 1980, 1990, 2000, 2010, 2015 and 2020. Tes, I know these are Version 6 with a different baseline to the animalies, but look at the trend!
CSIRO has simulated Australia cooling far faster than UAH.
Geoff S

Geoff Sherrington
Reply to  Geoff Sherrington
October 22, 2022 2:51 am

Exceeded my best typo count.
“anomalies”, “Yes”, “their”, “warming” not cooling.
No excuses. Apologies.
But look at the implications. Geoff.

Dave Fair
Reply to  Geoff Sherrington
October 22, 2022 6:55 pm

The UN IPCC CliSciFi propaganda mill uses various graphing tricks so as to hide the fact that their models do not replicate the late 19th Century and early 20th Century cooling nor the 1910 to 1945 global warming.

Last edited 3 months ago by Charlie Skeptic
%d bloggers like this:
Verified by MonsterInsights