From the “decades late and a dollar short” department comes the peer-reviewed exercise in the obvious. Of course, UEA was always a little bit slow. If they weren’t, we’d never have been given the gift of “Climategate.” – Anthony
Via Eurekalert:
Temperature of some cities could rise faster than expected under 2°C warming
New research led by the University of East Anglia (UEA) shows how many tropical cities are predicted to warm faster than expected under 2°C of global warming.
Cities are often warmer than rural areas due to a phenomenon known as the urban heat island, which can be influenced by various factors, such as regional climate and vegetation cover. This can lead to increased heat-related health risks for some urban populations.
Published in Proceedings of the National Academy of Sciences (PNAS), the study combined state-of-the-art climate change projections with machine learning models to show how these urban heat islands can be amplified in many tropical and subtropical cities under climate change – mostly in monsoon regions such as India, China and Western Africa.
The researchers produced projections for 104 medium-sized cities with populations ranging between 300,000 and one million.
Their results show the day-time land surface temperatures in 81 per cent of these cities are predicted to warm more than surrounding rural areas. In 16 per cent, they may rise between approximately 50 to 100 per cent higher than surrounding areas under 2°C of global warming, a benchmark likely to be reached in the second half of this century.
The cities studied are in the warmer parts of the world, which the authors say makes these increases even more significant for human health and the urban environment. Medium-sized cities also represent a large proportion of global cities, with more than 2.5 times as many in this category than those with a population over one million.
Lead author Dr Sarah Berk, who did the work while a PhD student in UEA’s School of Environmental Sciences, said: “Under climate change, cities face not only the challenge of increasing temperatures in their surrounding areas, but also the challenge of potential changes in their heat islands.
“However, while global climate models are essential for projecting future temperature changes, they are limited in their ability to capture the trends of smaller cities. Even high-resolution global models can only predict changes for the largest urban areas or megacities.
“To bridge this gap, in our study we projected changes in land surface temperature in medium-sized cities, showing that in many of them, the urban warming rate is faster than rural surroundings,” added Dr Berk, now at the University of North Carolina at Chapel Hill.
Co-author Prof Manoj Joshi, from the Climatic Research Unit at UEA, said: “Urban heat stress under climate change is an increasing concern, as many cities in the tropics and subtropics can be warmer than their rural surroundings, heightening their vulnerability to rising temperatures.
“This analysis shows even state-of-the-art projections likely underestimate future urban warming. For example, our results suggest that several cities in North-East China and northern India are projected to warm by 3°C, despite Earth System Model projections of their hinterlands showing a warming of 1.5-2°C.
“Our research enables more informed planning for the future risks to human health and the urban environment, highlighting the need to complement conventional climate modelling with approaches such as machine learning and AI.”
Prof Joshi added: “These findings also underscore the importance of investigating the effects of climate change on urban heat exposure, since climate change results in an increased frequency of extreme heat events, which can have severe human health impacts including increased mortality.”
The team excluded cities in mountain and coastal regions to remove influences of features such as hills, lakes and oceans, to ensure they captured relationships based on physical processes related to climate, rather than other differences.
In the five largest cities by population, the greatest changes are seen in Jalandhar (India), Fuyang (China) and Kirkuk (Iraq), which experience 0.7-0.8°C additional change in temperature compared to their rural surroundings.
The remaining two, Marrakech (Morocco) and Campo Grande (Brazil), see negligible differences between urban warming and that of their surroundings.
However, other cities experience significantly greater warming, for example Asyut (Egypt), Patiala (India) and Shangqui (China), which experience 1.5-2°C additional change, which is up to 100 per cent more than their hinterlands.
This work was supported by the Natural Environment Research Council and the ARIES Doctoral Training Partnership. It also involved researchers now at the Karlsruhe Institute of Technology.
‘Amplified warming in tropical and subtropical cities at 2°C climate change’, by Sarah Berk, Manoj Joshi, Peer Nowack, and Clare Goodess was published in PNAS on February 3.
Journal
Proceedings of the National Academy of Sciences
I didn’t realize it had been lost. Is this like Columbus discovering America and the natives not realizing they’d been lost?
“Backwater” best describes UEA…take from one who lives a mere 50 miles away…
After UEA was established in the 1960’s, locals in Norwich observed that the Fine City had lost a first-class golf course and replaced it with a third-class University.
Meaningless word salad. Climate is the statistics of weather observations. All cities are either tropical or subtropical, aren’t they?
Oh, “That’s not what they meant!”, is it? Maybe their grasp of English is as feeble as their knowledge of science.
Take this piece of pseudoscientific folderol – “Under climate change, cities face not only the challenge of increasing temperatures in their surrounding areas, but also the challenge of potential changes in their heat islands.”
Ah, the rich and fertile soil of “climate science” creating its own fertiliser!
I am constantly perplexed by the common practice by many supposed climate scientists to ignore the NOAA Climate Reference System results when evaluating urban surface temperature heat island effects. The Climate Reference System was designed to provide a heat-island-immune surface temperature record (and rainfall, sunshine and other weather measures) by distributing weather stations far distant from human-caused heat such as cities and airports. For the lower 48 there are 114 weather stations distributed uniformly throughout. The January 2026 temperature average of these stations is slightly lower than the temperature reported for January 2005 when the system was first enabled. Throughout the 21 year record, my Mark I eyeball cannot perceive any temperature trend up or down. The lower 48 is not the world but it is a good chunk of it and it has an excellent group of well-cited weather stations showing no change for the past 21 years. Why is this not a headline? Why are such stations not being deployed everywhere to get an unassailable measure of the worlds temperature? Why do climate scientists so rarely if ever use this information? Tis a puzzlement.
If you scroll down, there is chart with a plot of CRN temperature data from 2005 to Jan 2026.
. The lower 48 is not the world but it is a good chunk of it and it has an excellent group of well-cited weather stations showing no change for the past 21 years. Why is this not a headline?
See Richard Verney July 8, 2017 at 6:24 am
NOAA agrees
Because it puts a wrench in the gears of their attempts to blinker the plebes, and saddle them with carbon taxes and a social credit scoring system ala the ChiComs!? Just who do you suppose the media work for, the public!?
“The lower 48 is not the world but it is a good chunk of it”
About 1.6% of the Earth’s surface, hardly a “good chunk”.
1.6% is certainly sufficient to reflect “global warming” if it is actually happening. It’s not just a matter of total land mass but also a matter of climate zones represented by that land mass. From coastal to inland, from sub-tropical to moderate, across three separate time zones, etc …If “global warming” is truly GLOBAL, the US represents a good sampling of the possible effects.
Land and ocean have totally different profiles. What % of land mass is the North American continent.
The original post referred to the Lower 48, that is about 5% of the Earth’s landmass.
The original post referred to the Lower 48, that is about 5% of the Earth’s landmass.
Actually, the continental United States makes up roughly 10% of all land located in Earth’s temperate zones (both hemispheres).
Your eyes have deceived you.
The warming trend from 2005/01 to 2026/01 from nClimDiv is +0.71 F.decade-1.
The warming trend from 2005/01 to 2026/01 from USCRN is +0.85 F.decade-1.
What this suggests is that nClimDiv (old USHCN) may still be biased too low.
Tell us about the changes in Tmax and Tmin. How did each affect the average?
Have you done any time series analysis on the trends? Does auto-correlation have any effects. How about changing means and variance? Does seasonality have any effects?
Are summers or winters increasing?
We for once have some good data to analyze, simply doing linear regression on averages is not scientific.
Like I said below I don’t mind having a discussion with you, but since this particular topic requires an acceptance of averaging temperatures and since you don’t accept average temperatures as being meaningful or useful it would be pointless for me to engage with you here.
It’s not a matter of “accepting”. It’s a matter of physical science and basic statistics – averaging temperatures of differing variances with different distributions (e.g. sinusoidal vs exponential decay) is a perfect example of GIGO, garbage in/garbage out and, when coupled with the idiotic assumption that temperature = heat, gives a meaningless result.
The *real* reason you don’t want to discuss the issue isn’t “acceptance”, it’s the fact that you simply can’t justify the results based on either physical science or basic statistical principles.
You didn’t answer any of my points or questions. Your reason is that I will not accept averages of averages of averages as scientific proof of anything.
I do not support mathematical procedures that supposedly increase accuracy, reduce precision uncertainty, and invents resolution beyond what was measured. If that is your complaint, then so be it.
I have asked you multiple times to provide university physical science lab requirements that allow one to use averaging to accomplish these things. I have given you multiple examples of lab courses that explicitly deny this in measurement results. You have provided not one example otherwise. That is telling.
Here is a paper on using time series analysis for trends. It even specifically mentions temperature time series.
https://www.tandfonline.com/doi/full/10.1080/00273171.2024.2436413#abstract
It even mentions using time series analysis for global climate research.
Funny how monthly and annual averages along with linear regression trends don’t fit into this.
I’m not talking about propagation uncertainty with a measurement model that uses averaging.
I’m talking about averaging intensive properties and the rejection of it.
If you have change your mind and decided that averaging intensive properties like temperature is meaningful and useful afterall then great. Just say so and then we can discuss the differences in the averages of nClimDiv and USCRN.
No change. The use of temperature in climate science came about because it is a factor in the SB radiation equation. Average temperatures and insolation and bingo you got a solution to warming.
When I studied themodynamics, heat sinks and cooling towers you needed calculus to solve the dynamic gradients. When was the last time you saw time based gradients in a radiation diagram?
Averages may be fine for back of the envelope thinking. They are not adequate for studying the heat mechanisms of the earth. After 50 years of climate science, the study of the thermodynamics of the earth should have progressed further.
Someday I wish I could get an answer to this question from someone defending the global average temperature.
If I put two rocks in your right hand, one at 80F and one at 70F, do you have a temperature of 150F in your hand?
If no, then how do you calculate an average temperature. If yes, then how do you keep from getting burned?
I’ve asked you before and you just ignored the question.
If I put two rocks in your right hand, one at 80F and one at 70F, do you have 150F in your hand?
If no, then how do you calculate an average temperature? If yes, then how did you manage to not get burned?
It suggests NOTHING of the sort.. ClimDiv is purely an FAKE series, adjusted in an attempt to match USCRN.
Any difference is purely down to how it has been “adjusted”
Before 2005, they did not have USCRN to adjust to
ClimDiv started above USCRN in 2005, so their initial guess was too high.
The only warming in USCRN comes as a small step change at the 2016 El Nino, with zero trend either side.
That’s why there are such things as trend lines. There is a clear warming trend in USCRN; faster, in fact, than the adjusted ClimDiv. Data here – check it for yourself!
(Be nice if you self-proclaimed ‘skeptics’ would try and check things for yourselves once in a while.)
The question a scientist would be examining is what causes the trend to change? Is Tmax increasing? Is Tmin increasing? Both? Are there regional differences? Are seasonal changes occurring that raise the average? Are summers warming? Are winters warming?
Your obsession with trends of averages blinds you to researching what is actually occurring.
Here are three graphs from NOAA. Do you see any large trends?
And just for grins, I am going to show you a trend for the Ithica, NY USCRN station that has auto-correlation and seasonality removed. Look closely. Does the trend line of a stationary graph of the data show a trend? That should be your starting point to analyze what causes an increase.
Step 1. And then their modelling takes its primary inputs from a mix of urban and rural weather stations to get the 2C increase, and then they back feed it to amplify the urban predictions.
Step 2. ???
Step 3. Profit!!!
State of the Art projections? I thought it was State of the Ark.
“mostly in monsoon regions such as India, China and Western Africa.”
This is curious for a few reasons. Why the focus on places that have a monsoon pattern? Or are they saying that their models are suggesting that the heating will supposedly be greater there? If so, how? Wouldn’t a (presumably summertime) monsoon mitigate heating there? As for those cities having the largest UHI warming, isn’t UHI most prominent in colder climates, during the colder months, and at night when the residual warmth is released in town? Wouldn’t that reduce the demand for nighttime heating, and thus be a good thing?
I thought something similar, have they chosen those because of the temperatures involved, 2 – 3C, to align with the crisis narrative?
I still would like to have the full contents of the “climategate” archive.
I downloaded Eric Worral’s Climategate app:
Download Climategate Latest Version 1.01 Android APK File
It requires self-installation, including potentially turning on the ability to install apps from other than the AppStore. The web page has some install instructions.
(Android version – no idea whether an Apple version is available).
The university of easy access is famous for more than its climate efforts and it explains their style
UEA pioneered the teaching of Creative Writing more than 50 years ago and is globally regarded as a leader of the subject.
https://www.idp.com/universities-and-colleges/university-of-east-anglia-uea/ma-in-creative-writing/PRG-UK-00501540/
Fiction.
As one might expect from East Anglia University, they have wildly underestimated the magnitude of the Urban Heat Island. They probably should spent a few days taking data rather than using models.
“Cities are often warmer than rural areas due to a phenomenon known as the urban heat island,”
Well, who knew !!!
I bet this phenomenon is 97% unprecedented.
( Ed uk shun is u wunduful fing ) (:-))
So, the university has discovered the heat island effect which is expected to kick in any day now.
Two things
Next?
Now they have to explain it to the Met Office in simple enough terms that they can understand.
Shouldn’t climate alarmists be particularly careful not to produce hot air?
Published in Proceedings of the National Academy of Sciences (PNAS), the study combined state-of-the-art climate change projections with machine learning models
and what about crystal ball gazing’s and Ouija board insights?
Chicken entrails would be more apt.
Nice pick up Anthony.
Did the study not examine how weather stations sited in and close to urban areas will affect the overall temperature trend? Maybe this is the first step in admitting that the GCM’s have a major problem predicting anything. If you remove the UHI effects, it will make the models even more wrong.
Some ‘discovery’ indeed. But the penny has not dropped yet. They finally recognise the UH effect but have still to discover that it has corrupted the measured temperature record. Or perhaps they just don’t want to.
I think some are beginning to recognize that too much data and too many independent people analyzing the data is eating away at the foundations of CAGW.
Many of us here have spent time analyzing random locations that have little to no warming. These are small chinks in the armour of current climate science that need explanations which never come. Over time these build into a cohesive whole illustrating that errors exist.
The first paper I found that dealt seriously with Tmax and Tmin separately was from agri-science that showed warmer night time temperatures and longer growing seasons.
My research showed similar results. An increasing Tmin overall and moderating winter temperatures. Somehow climate science has used the coldest temperatures to illustrate that the earth is burning up.
Is this the beginning of the gargantuan a$$covering needed to explain how “97% of experts” could get the whole Global Warming thingy so wrong? “The UHI Effect ate my temperature data, I swear!”
The UK continues to be a fascinating case study, in real time!
The government is in disarray and very unpopular. Local elections are coming up in May which the voters are expected to use as a referendum on it. The general expectation is that it will lose by a landslide. A key element in the discontent is cost of living, and disillusion with the Net Zero drive is rising.
There is only one party in the UK with unequivocal opposition to Net Zero, and that is Reform, which is leading in the polls and threatens to eject Labour from many councils.
The Government’s reaction was to allow local councils to cancel the elections, and many who were afraid of Reform did so. However this was pretty evidently unconstitutional – the UK has an unwritten constitution, but its nonetheless real and based on legal precedent and a body of oridinary legislation and case law. The result of the cancellations was an application by Reform for judicial review, which was granted, and due to be heard on Friday of this week.
Faced with an almost certain loss in the case, the Government did yet another U-turn and the elections will now go ahead. Its hard to say how important Net Zero and energy will be in them. It probably won’t dominate as a direct issue, but its effects in raising the price of energy and the cost of living generally, which are very important direct issues, will be considerable.
So the May elections will be one test. A by-election for Westminster at the end of this month, in a northern seat, will be another test.
A test of what, exactly, you ask. Simple: it and every other election between now and the General Election are a test of whether, in a western democracy, a woke political and media elite can impose net zero with all its costs on a population whose view of it varies from the indifferent to the hostile. Can they do it, or do they get kicked out at the next election. And can they get away with cancelling elections, if this is the only way to stay in office and keep on with the agenda?
Worth watching.
“The University of East Anglia Discovers the Urban Heat Island Effect”
A rather provocative title which is obviously not a correct description of the study!
Did she take into account UHI increases due to projected population growth / urbanization in those cities over the study period? That alone could swamp any meager effect of global average temperature.
This “research” reads like another high school science fair project. 1. Child fed a constant diet of climate doomism from pre-K through high school. 2. Mentally sensitive, anguished HS graduate who is good with math wants to major in college in something “environmental” . 3. Socially misfit doomer college graduate can’t get a real job so goes for a PhD in “climate” studies. 4. Government-funded diploma mill “climate” professor must help the candidate come up with a dissertation topic and plucks one out of the climate scare menu. 5. Grad student, lacking any worldly experience or historical perspective, churns out a high school science fair quality demonstration, calls it groundbreaking dissertation research. 6. Professor notches another “peer-reviewed” publication on his CV. 7. Newly-minted PhD leverages her “intersectionality” (female x climate change) into a low-paying DEI post-doc:
Sarah Berk, a Postdoctoral Research Associate at the Data-Driven EnviroLab at University of North Carolina at Chapel Hill
A key part of the project “Exploring Disparities and Solutions to U.S. Urban Heat Stress and Climate Justice” is recognising the importance of integrating community in our research.
So, now her work is centering on social justice (so-called UHI disparities). Post-docs famously get very low wages, little better than an undocumented illegal alien lawn service worker.
As opposed to a maths graduate who worked for a couple of years in finance and then decided she wanted to switch to a field where she could use her maths skills to make an environmental impact and pursued a PhD in Climate Research. Post-doc salaries of about $70,000 pa seem a bit higher than “an undocumented lawn service worker”.
Urban heat islands were discovered in the 1800s but research into them only started in earnest a couple of decades ago.… probably because it was inconvenient
This paper can be used to promote the development of reliable electricity needed for air conditioning in these at-risk places. For the poorly housed, set up large swamp-coolers (a wet membrane and a fan) and then collect the off-product minerals to be sold and used accordingly. Sarah and friends are just the ones to make this argument.
Anthony,
In reading the comments it looks as if this article has been taken by some to mean that UEA and/or CRU has only just now discovered the UHI effect. I’m assuming this was unintentional, but it might help if you added a note to the article mentioning the following citations relevant to the topic.
[Howard 1818]
[Lamb 1959]
[Lamb 1969]
[Jones et al. 1986]
[Jones et al. 1990]
Basically, the UHI has been known to science since at least the early 1800’s. And CRU has understood the importance of the effect since its inception in 1972 by Hubert Lamb and has taken steps to account for it when producing gridded temperature products.
Are you saying that they have REMOVED the effects of UHI from the temperature data series on a global basis?
Your last Jones reference paper indicates that the effect is less than 0.05°C on a global basis. Not having access to the paper, I can’t evaluate the study, but it is hard to believe.
AW’s study of siting problems in the U.S. and the recent investigations in the U.K. would also indicate that UHI is more than just a minute part of the contamination of the temperature data.
I don’t think anyone expects the USCRN network to show a hockey stick increase over the next 10 years which means UHI is not a small matter.
I would like to have a discussion with you about this, but you’ve been steadfast in your belief that average temperatures are useless and meaningless. It would be pointless for me to engage with you on this topic as a result.
Average temperatures are useless in the way they are analyzed. They have one hypothesis at the start, CO2 is raising temperatures, summer/winter, day/night, rural/urban. Then see, the averages are increasing just like we said!
Why did it take agri-science to sit down and discover that night time temperatures were increasing causing longer growing seasons? This should have been recognized and trumpeted at the very start of climate science in the 1970s.
The pointless problem exists because you and climate science refuse to move on from traditional Tavg. You would lose the precious “long records” that allow simple linear regression on Tavg.
Not surprising considering Anthony’s provocative title for his post!
https://wattsupwiththat.com/2026/02/16/the-university-of-east-anglia-discovers-the-urban-heat-island-effect/#comment-4166540
It does explain why basically ALL adjustments of raw data …
COOL the past and WARM the present.. doesn’t it. 😉
And the sentence at the top: “From the “decades late and a dollar short” department comes the peer-reviewed exercise in the obvious. Of course, UEA was always a little bit slow. If they weren’t, we’d never have been given the gift of “Climategate.” – Anthony”
the study combined state-of-the-art climate change projections with machine learning models to show ……
Very impressive science … for the KIndergarten.
Even the IEA recognises that air conditioning in cities in China, India, SE Asia and parts of Africa is going to be a major cause of electricity use rise in the near future.
IEA ‘World Energy Outlook 2025’ (Nov 2025)