By Kenneth Richard on 7. August 2025
The urban heat island effect adds far more non-climatic heat to temperature station records than can be reliably controlled for.
An analysis of 10 cities from across the globe (Kara and Yavuz, 2025) reveals airports and industry centers are, on average, 2.5°C to 2.8°C warmer than neighboring green spaces.
“Airports exhibited a mean daytime land surface temperature (LST) that was 2.5 °C higher than surrounding areas, while industrial zones demonstrated an even greater temperature disparity, with an average increase of 2.81 °C.”
The urban heat island effect can leave airports and industrial centers as much as 12°C warmer than nearby vegetated, forested areas.
“Mexico City’s green spaces are up to 12.13 °C cooler than its urban core.”
Warming trends in recent decades are generally confined to urban areas, whereas non-urban areas have been cooling. For example, from 2001-2021 urban areas warmed +0.04°C/yr, but vegetated, bareland, and water body areas cooled -0.07°C, -0.03°C, -0.04°C/yr, respectively, in the city of Chongqing.
“In contrast, cold spots characterized by dense vegetation showed a notable cooling effect, with LST differences reaching −3.7 °C. Similarly, proximity to water bodies contributed to temperature mitigation, as areas near significant water sources recorded lower daytime LST differences, averaging −4.09 °C.”
Image Source: Kara and Yavuz, 2025
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Fits my experience from driving around.
The variability between rural and built up areas has always been way larger than I have ever seen to be recognised.
8 degrees in Calgary, Canada. The city has a heat map that shows it (I don’t have access to it at the moment).
So pretty much global warming is an artifact of where the temperatures have historically been measured…in urban zones at night during winter. But not so much in Forests or Grassy Plains or in water zones where global cooling is the order of the day.
Outside of North America and parts of Europe the only long term consistent measurements were in cities so most of the “world” records are hopelessly contaminated.
Even in N. America and Europe, most of the long-term records were taken areas that are now urbanized.
Good thing CC Zealots want to drive humanity out of rural areas and into 15 minute city zones making urbanized areas even more populated and thereby increasing the UHI effect
This is a paper in a MDPI pay to play journal, so be wary.
It isn’t telling us much. It has no data on measured air temperature. It only tells of radiation from surfaces. No-one will be surprised to learn that these are higher in cities.
Except all those who point to about a few hundredths of a K ‘record’ temperature warning being evidence of Catastrophic Anthropogenic Global Warming. At a airport.
But nobody here would be as silly as that, would they?
Be wary of the truth , Nick !!
Nick has no trouble with partial truths. It’s whole truths that he avoids.
The papers you really have to be wary of are ones PAID to play…
… as found in such low end climate psychotic rags like “Nature-Climate” etc !
Wait a minute. Isn’t it radiation from the surface that warms CO2 (the “air”) that, in turn, causes back radiation that causes the surface to get even warmer causing more radiation causing higher CO2 temperatures that causes more back radiation that causes even higher surface temps that causes more radiation causing higher CO2 temps that …….
There is an abundance of temperature data that confirms this if you just look.
Here is one graph for a single month in Topeka, Ks where I live. Same for all months.
Here is another from a world away from me.
There are tons of others. How does the saying go? There are none so blind as those who refuse to see!
looks like the only data indicating an increase of temperatures is the “Adjusted” data
Re your Hachijojima graph……
https://briangunterblog.wordpress.com/2021/02/03/japan-temperatures/
“DATA USED
Monthly records of mean maximum and minimum temperatures are available for many Japanese stations on the website of the Royal Netherlands Meteorological Institute (KNMI), and these were used in this analysis. Raw data (ie unadjusted, as recorded) were used.”
“Hachijojima has a warm, humid, subtropical climate with little temperature variation throughout the year. The island is influenced by the warm Kuroshio Current and volcanic activity. It has an average annual temperature of 18°C (64°F) and experiences high annual precipitation of around 3,202mm. ”
Is this supposed to be a *typical* global weather station trend?
I ask (rhetorically), as it is very far from that, as it’s small Island in a large ocean well south of mainland Japan ….
…. and as such is entirely subject to the surrounding SST for it’s climate.
(Another fave temp trend from a small island in a v large ocean being Macquarie Isl – 930 mls south of Tasmania)
Oh, and Hachijojima weather is also taken at an airport.
So the above study finds a trend of +0.3C for maxes and +0.5C for Mins (corresponding the global measured greater warming trend for mins).
So was that down to increasing traffic at the airport (sarc).
Or the slight increase in the surrounding ocean SST in those 103 years? + a small influence in GHE warming ( mins).
Essentially Hachijojima (like any small island ) has a perpetual sea-breeze, and the temperature of it (on the beach of a mainland) is the same temp as the SST immediately upwind.
This is misleading as the islands are dominated by an Ocean climate.
And oceans are warmed BY THE SUN, not human CO2.
In fact, NOTHING is warmed by human CO2, anywhere.
There is certainly no evidence of human CO2 warming in the UAH data.
Al load of garbage. You have obviously never been to Waimea Bay
The point is that IT IS a land temperature. You can explain why all you want, but it is included in the GLOBAL mean temperature. The temperature here has little change over a century and should have some (however small) effect on the changing of the global mean.
If you want to make an impression, examine a number of random rural stations across the globe. Funny how so many have no CAGW trend.
Daily minimum and maximum aren’t useless for determining temperature averages. However, they are close enough to useless that the difference isn’t worth talking about.
That some climate “scientists” admit that Urban Heat Island (UHI) warming exists is true. However, most claim that it is only a degree or two and that it has been controlled for.
Once again Nick gives a partial truth, in order to hide the whole truth.
I don’t think he’s hiding it, just ignoring it.
This paper currently appears to be open and available for free PDF download at https://www.mdpi.com/2413-8851/9/4/115 .
And how many MANN Articles are in “Pay to Play Journals”?
From the article:
“While urbanization contributes minimally to overall global temperature trends, it significantly intensifies land surface warming at local and regional scales, ”
so the paper itself says that the urban heat island is not significant on global scales. As you would expect from the satellite measurements 70% of which cover the oceans which have no urban heat island effects.
And overall global temperature trends are insignificant. 😉
So population and buildings increase temperatures. Who knew.
‘As you would expect from the satellite measurements 70% of which cover the oceans which have no urban heat island effects.’
So we should not worry too much about climate reports about the temperature of the land, as it is only the oceans which are measured accurately?
Over the past 50 years, the temperature above the Pacific Ocean—measured as the sea surface temperature (SST)—has increased significantly due to global warming. Recent research estimates that the ocean has warmed about 0.11°C per decade over the last 50 years, resulting in an approximate total increase of 0.55°C from 1975 to 2025.
That will be an increase of 1.1 degrees per century.
Exactly.
And possibly as much as half of that might be caused by and additional 0.75% of greenhouse gases. Nobody can say for sure, however. It could mostly or all be natural variation, as past warm periods were.
In any case, that isn’t cause to panic, obviously.
No, oceans warmed due to reduced cloud cover, no CO2 involved.
No signal factor is the control knob.
Yes, reduced/increased cloud cover has an effect.
Yes. minor trace CO2 increases has a trivial effect.
1975-2025
Yes, and from 1940-1970 global cooling is what was observed. There’s plenty of reason to believe that the 1930s were warmer than it is at present.
And one should consider ALL ocean temperatures if one wants to make a global estimation, just like with the poles. When one warms the other might cool. The warministas only look at hotspots, even the ones that do not exist but should, according to their hypothesis.
No. It’s not. Nor is human habitation, at global scales, despite what the Malthusians would have you believe.
But at weather stations, where temperatures supposedly providing ‘evidence’ of Catastrophic Anthropogenic Global Warming are recorded? Oh, yes.
The point is not that UHI increases temperatures on a global scale, it’s that it increases measured temperatures, which are then used to suggest that global temperatures are increasing as well.
I’d be really interested to see what the data shows if we only looked at Class 1 sites, has anyone done this?
Especially when those urban temperatures are extrapolated or interpolated to non-urban turf.
The points are (a) we do not know what the average global temperature is, (b) the way it is calculated is bogus, (c) we do not know what the optimum is (assuming it has any meaning whatsoever).
Just adding to the valid points you make.
Which does not stop the Met Office using these airport temps as all time record temp headlines.
Satellite measurements in the last 45 years show no warming except at El Nino events.
Surface measurements show continual contamination by urban warming and “adjustments”
In an effort to explain a nagging difference between the surface temperature record and the satellite temperature record, the keepers of the satellite record determined, using changes in population density and the raw surface temperature data, that UHI accounts for 65% of temperature increase in the surface record. I believe this was for US stations only.
Your ability to read and comprehend is far below grade level. The problem is that published global temperatures USE these not fit for purpose to generate averages.
Your statement is just another way of saying only urban areas containing measuring devices have warmed while the rest of the globe has not.
Satellite measurements yes. However, most cliministas prefer to use the ground based measurements that are hopelessly corrupted.
Beyond that, the almost constant claims of high temperature records proving that the world is going to burn up due to CO2 is also disproved by these facts.
Even satellite measurements have problems, inaccuracies, and errors.
They are, however, the best we have today.
Not going to do the deep dive. Not now.
“contributes minimally”
What does that mean?
No, I don’t think the paper says that at all. Here is the very first sentence under the cited paper’s 5. Conclusions section:
“In conclusion, this study observed an overall increase in land surface temperature
(LST) both during the day and at night across major cities on all continents.”
And here’s some simple math for you:
If 70% of a given data set used for temperature trending has “no change”, but the remaining 30% has a net change of, say, +3 deg-C over a specified time interval arising from area- & magnitude-averaged UHI, then the full data set will show an average increase of +1 deg-C over that same time interval.
I do believe that the IPCC and other AGW/CAGW alarmists have indeed been “wringing their hands” over just a 1.5 deg-C increase in average global LAT since about 1760 (ref. Paris Agreement). HAH!
It’s not significant on solar, planetary, galactic or universal scales either. So what’s your point? The only place it’s significant is where it counts; where people live at where they are directly affected. Pacific Islanders don’t give a rat’s @ss about the UHI in, say, Atlanta, but the people in Atlanta sure as heck do.
28 million square miles converted from natural to hard surface over 100 years. But no affect on measured global temperatures? I call bullshit on that one.
Even in smaller towns like Scunthorpe the UHI effect has a very noticeable effect on minimum temperatures. Listed below are the max and min temps of central Scunthorpe compared to the local rural weather stations at Hatfield and Thorne Moors for the first 7 days of August.
1st max min
S 19.2C 15.1C
H 19.3C 13.3C
T 20.3C 13.2C
2nd
S 22.1C 13.5C
H 22.3C 11.6C
T 21.9C 10.1C
3rd
S 22.9C 12.3C
H 24.2C 10.3C
T 24.4C 9.7C
4th
S 20.5C 14.0C
H 22.2C 14.6C
T 22.8C 14.1C
5th
S 19.2C 12.7C
H 20.3C 10.6C
T 19.9C 11.5C
6th
S 21.0C 11.9C
H 21.9C 9.4C
T 21.6C 8.8C
7th
S 22.8C 14.5C
H 23.4C 12.4C
T 24.0C 12.3C
It’s during clearer night’s where the UHI effect has the greatest impact.
But notice how often central Scunthorpe record’s lower maximum temperatures then the local rural areas. Now that is down to the fact that l record temperatures with a LiG thermometer in open shade rather then with a electronic thermometer in a Stevenson screen.
And just how much of my tax money was pissed away to point out the obvious?
No One denies the uhi effect. but unless the city is growing by considerable amounts the effect should be constant with time – has this been checked where is the data. E.g. how much has the core of Mexico city increased over time compared to the surroundings. The plots provided show very little increase in day temperature from 2001 to 2021. And night temps seem to decrease!
To remove the uhi content from temperatures is simple – just use ocean heat content/sea surface temperatures and you will see a rising plots. Strange huh? Perhaps the world is warming?
Of course not. A global loss of 44 TW is cooling, not warming.
As Fourier pointed out, the Earth loses all the heat it receives from the Sun, plus a little internal heat (being hotter than the surrounding environment at about 4 K.
Hence, after four and a half billion years, the Earth has liquid water, and a congealed solidish crust of some depth.
You are free to believe in a GHE, as many ignorant and gullible people do. No consistent and unambiguous description exists of the fabulous GHE, of course. Its believers can only imply its existence by asking silly questions like “Perhaps the world is warming?”.
Religion, not science.
Of course it’s warming, we are coming out of an Ice Age and a Little Ice Age, thank goodness we are.
I don’t fancy mile high Ice Glaciers.
We were coming out of the Ice Age, however the Holocene Optimum ended around 5000 years ago, the world has been on an overall cooling trend since then.
This cooling has been interrupted by 5 warm periods, about 1000 years apart and each one cooler than the previous.
Minoan, Egyptian, Roman, Medieval and modern.
The problem is that most ground based sensors are located in areas that were once rural and are now suburban or even urban.
That is where most of the temperature stations have been located.
There are very few cities that aren’t growing.
The cores of these cities are irrelevant, since there are no climate stations there.
Most of the sensors are at the edge of cities, and these areas are warming rapidly.
If the claim you want to make is that urbanization accounts for the bulk of the observed land surface warming trend, you have to demonstrate that the adjustments applied explicitly to account for urbanization bias do not work. None of you has yet to come close to doing this.
Another Appeal to Anonymous Authority. If you are going to reference experts, you need to show the references you are appealing to. We have no way to know what changes you are referring to, nor what studies contain the information you are claiming to know.
See:
https://crudata.uea.ac.uk/cru/data/temperature/#sciref
https://www.giss.nasa.gov/pubs/
https://berkeleyearth.org/papers-climate-science/
https://www.ncei.noaa.gov/products/land-based-station/global-historical-climatology-network-monthly
https://crudata.uea.ac.uk/cru/data/temperature/#sciref
Funny how this document has NO MENTION of UHI (or urban).
Not going to read anymore as you’ve just demonstrated your reading ability.
I linked you to the references section of the HadCRUT website, you need to access and read the cited publications. If you aren’t capable of that, then I do think it’s best if you sit this one out.
Yet you could have posted the relevant section instead you blow fog and evasions.
The problem is not mine, it is you who cannot provide sufficient evidence to support your assertions.
I could copy and paste the whole page you referenced but I assure you my software could not find UHI or urban.
That is your fault.
My apologies, I take full responsibility for unfairly assuming basic literacy on the part of my opponent. Now that you know you need to study the research papers referenced in the link provided, perhaps you’ll be able to take it from there? Let me know if you need more hand-holding.
Studying links in a link is not my job, IT IS YOURS. If a link on a link has pertinent information, then IT IS UP TO YOU to post the link and you should give a quote showing how it supports your assertion. Anything less is just an Appeal to Authority.
Teaching you basic information is certainly not my job. I do not care if you wish to remain uneducated. I am satisfied that you are unable to meaningfully contest any of the methodology employed in any of the published research. Your inability or unwillingness to engage is a tacit concession that I am happy to accept.
Is Ad Hominem all you got now? Why am I not surprised?
No quotes, how lazy of you.
Mr. tommy: As you know, Mr. J can quote or link to a biased source that tells him what he wants to hear. You’re correct to note that, here, he doesn’t start out with a cite or quote, unusual for him. He’s a one-trick-pony act, linking to some preferred source, then telling you that you got no links. Here, upon request, he posts four links to articles, if you read them you lose IQ points. I just like to mock him, and thank this site for the opportunity.
“you have to demonstrate that the adjustments applied explicitly to account for urbanization bias do not work.”
why would you make adjustments in the first place?
joe x
Primarily because the goal of a surface temperature analysis is to isolate the signal arising from changes in the surface temperature, and not changes in the observing network or non-climatic changes arising from regional effects like urbanization.
rule #1…if you need to adjust raw data, than you need to re-design your experiment.
That isn’t a rule of anything, and these temperature indexes are analyses of historical datasets – there is no possible way to “redesign the experiment” save inventing time travel.
“That isn’t a rule of anything,”
i should be if you truly want accurate analysis.
“there is no possible way to “redesign the experiment” save inventing time travel.”
sure there is, just think out side the box. i don’t know what the re-designed experiment should looklike but i have some ideas. here is a question..why are we measuring air temperature at all?
regarding inventing time travel, see manns hockey stick.
What are your ideas?
Because knowing what the weather outside is doing is important for a lot of human enterprise. Are you asking why we estimate long-term global/hemispheric temperature change? That is because we want to monitor large scale climate change.
More horse crap propaganda.
Tell us how the temperatures from various standard climates areas are weighted to account for temperature variance. How do you weight the various heat environments in Miami, Phoenix, and Bangor, Maine?
Knowing what the temperature outside is, is one thing, being able to use those same readings to calculate a global average is something else all together.
It’s really amazing how ignorant your average climinista is. Especially AJ.
In other words, it doesn’t matter how invalid the data is, if it’s all you’ve got, just use it and pretend there are no problems.
so we take a temp reading in the middle of a thermal mass environment (steel and concrete city) adjust that temp data to what we think or feel it would be as if the thermal mass environment never existed? that is called putting your finger on the scale in my book.
There is no “we think or feel” involved. UHI is handled with transparent, testable methods like comparing urban stations to nearby rural ones, using satellite data to flag urban sites, or statistically modeling the urbanization effect. The goal is to separate true climate change from local development effects, just like calibrating a thermometer, and the raw data and methods are public so anyone can check the math.
By invariably COOLING the past !!
Oops.. The “adjustments for UHI should be cooling the present. !
More hokum from a troll. “Isolate THE SIGNAL” my foot. You are trying to high flying language to insure folks that you know what signal processing is and how it works.
The first thing you need to do is to tell everyone exactly what is interfering with the temperature data such that extracting a signal is necessary.
Then you need to describe how subtracting does anything more than simple scaling of values. It certainly isn’t a filtering process to remove an interfering signal.
The trend for a station record is the sum of the regional climate, local variance, and non-climate bias. The goal is to identify and remove the third term.
Ahhh, the old perceived bias in measurements.
You do realize that bias in measurements means the calibration for the device is incorrect?
What do the JCGM docs have to say about bias.
JCGM 100:2008
JCGM 104:2009
JCGM 200:2012
The primary issue is that bias in a measurement cannot be determined by comparing a measurement to another field device, it can only be assessed thru calibration. Other field devices have their own “biases”, which makes them ineligible as a reference.
You cannot assume measurement bias exists when comparing different device types. Each device may be in calibration, yet their implementation of a measurement model may give varying readings.
Yawn.
Instrument calibration is not the sole potential source of systematic bias in station records. You know this, let’s not rehash your boring old tripe for the umpteenth time.
We will rehash because you have no clue about making measurements and evaluating them.
Instrument calibration is the source used to determine what correction factors should be applied to correct measurement bias. Remember, the entire system is what must be calibrated, not just a single component. That is why you can’t determine system uncertainty by comparison to another device.
Your definition of bias is different readings between two different measuring devices. That is not measurement bias and “corrections” to make them agree to a preconceived value is not correcting bias.
At a theoretical level, that is true. However it is completely impossible to tease out how much of a measured increase comes from each of those components. Not to mention the contribution of components not yet thought of.
Unless you properly control the area around the sensors, then you have no idea what the readings you get mean.
Your belief that you can allow sensors to be contaminated by local influences and still use their outputs, is as unscientific as most of your bizarre claims.
And you have to demonstrate that the adjustments applied are correct.
This is not a call for you to cut and paste someone else’s work.
But you will. You cannot stop yourself.
There has yet to be a competent challenge from the contrarians that might cast any doubt on the viability of the adjustments. There is nothing for me to prove. The published literature speaks for itself, I suggest you become familiar with it.
WE don’t have to demonstrate squat. The onus is NOT on us to prove a negative. You got the null hypothesis backwards again.
The onus is on you to substantiate any argument you make. Scientists have extensively defended their methodology in the peer reviewed literature. You are saying they are wrong. Now you get to put up or shut up.
NO, it isn’t. Scientists may have “extensively defended their methodology in the peer reviewed literature” but that doesn’t mean they are right, or that peer review gives them some sort of Holy Imprimatur that must be believed unquestioningly like religious dogma. They have been wrong on every prediction they have made over the last 30-plus years and I don’t have to put up, and I don’t have to shut up.
It doesn’t make them right, it just means they have put up a compelling scientific argument, backed by the preponderance of evidence, and you have failed to meaningfully challenge any of it.
Of course there is the ever-present implicit third option: babble like an ignorant fool. But I assumed you wouldn’t want to go that route. Apologies if I was mistaken.
Kinda like what you are doing with this article.
As Phil R points out, past predictions can’t just be forgotten. Past performance does influence one’s trust in future predictions.
If temperatures at one’s location has little change over decades of life, climate change becomes very low on one’s priorities.
I am saying nothing about scientists being wrong. I am saying you need to provide quotes from their work to support YOUR assertions. You need to be able to study their work until you can quote by heart with understanding.
Anything less means you are simply using Appeal to Authority because you don’t know enough to argue the facts.
Defended their claims they have done.
Successfully defended their claims, they have not.
It really is amazing how unscientific and illogical Alan is.
Yeah, about that null hypothesis –
does it exist?
Is it in the Epstein files?
AlanJ is an insider with all things climate “science” –
maybe he knows where the null hypothesis is for anthropogenic CO2 sole cause of global warming.
Duh
I suppose the GISS headquarters staff was too busy spinning and socializing to notice.
Very interesting that authors Kara and Yavuz, as seen in the above article, have difficulty bringing themselves to use the term “urban heat island” (UHI), instead referring to:
— non-urban areas cooling
— cold spots characterized by dense vegetation showing a notable cooling effect
— proximity to water bodies contributing to temperature mitigation.
A word search of the full abstract of their referenced 2025 paper and its listed keywords does not reveal “urban heat island” or “UHI”, only references to “urbanization”. Admittedly, you do find these terms in lower sections of their paper.
Thus, they appear to be sensitive as to where they may be getting their next funding increment.
/sarc
I was surprised that one of the Met Office mentioned the UHI effect in one of their videos. This in relation to a possible coming heatwave. But then, the UK says a heatwave is 3 consecutive days of temperatures between 24 and 28 degrees C.
Anyone living in a non temperate zone would scuff at this. It would be considered a cold day for summer..
I was surprised that one of the Met Office mentioned the UHI effect in one of their videos. This in relation to a possible coming heatwave. But then, the UK says a heatwave is 3 consecutive days of temperatures between 24 and 28 degrees C.
Anyone living in a non temperate zone would scuff at this. It would be considered a cold day for summer..
Anyone who’s ever ridden a motorcycle has probably encountered hot, warm, and cold pockets of air while riding, and has noticed the temperature difference between urban and rural riding that can happen in just seconds.
Glad to hear science is catching up.
i used to ride, and you are 100% spot on.
WUWT readers aren’t surprised. In the late 1990’s or early 2000s, Roger Pilke, Sr pushed the National Academy Panel that was reviewing global warming (the term In then used most often) to pay close attention to land use and its influence on reported temperatures. Embracing his advice could have promoted a much better understanding of the basic issues and provided for better mitigation and adaptation strategies.
NASA GISS and NCAR have claimed to adjust for the UHI effect, but their publications almost always attribute the reported warming only to Climate Change. Anthony has been diligent in documenting the combined influence of siting for weather stations and the change from mercury thermometers to electronic instruments on reported temperatures. That led to the identification of 100 reference stations that are properly sited and housed, which document a lower temperature change than the averages typically reported On top of that, we know that the MET in the UK guesstimates 1/3 or more of their temperature inputs for nonexistent locales. UAH has provided satellite data that seems to parallel the reference stations, and he has recently been publishing estimates of the UHI contribution of UHI to reported GMTs.
The globe is in a warming phase (perhaps partly due to GHGs), but the Climate Alarmists prefer to highlight exaggerated measurements from heat islands, “sea Level rise” from costal areas that are sinking, and hyping the latest extreme weather event to create panic among the general public.
Good science can be used to identify better ways to mitigate localized problems such as the UHI problem and siting of towns and homes in flood planes rather than wasting vast amounts of money on pie-in-the-sky technologies.
For decades winter weather forecasts in the U.K. have had the disclaimer “these are temperatures in urban areas, rural temperatures will be a few degrees cooler”
I grew up in Virginia but moved to the Philadelphia area in the 1977-1978 timeframe. At that time there was a local weather man named Jim O’Brien who used to talk about the heat dome or heat island of Philly and how it would be several degrees warmer in Philly than in the surrounding areas. First time I heard reference to heat islands being from VA, but long before “global warming” became a thing.
Not surprising. At least to those of us who are paying attention, or who drive from the middle of a city to the countryside regularly.
I noticed this one summer 20 years ago when did drives to the court houses and fairgrounds of very county in the state. On hot summer days, I was surprised at the temperature differential even in very small towns.