Urbanization Effects on GHCN Temperature Trends, Part I: The Urbanization Characteristics of the GHCN Stations

From Dr. Roy Spencer’s Global Warming Blog

January 14th, 2023 by Roy W. Spencer, Ph. D.

I’ve previously posted a variety of articles (e.g. here and here) where I address the evidence that land surface temperature trends from existing homogenized datasets have some level of spurious warming due to urban heat island (UHI) effects. While it is widely believed that homogenization techniques remove UHI effects on trends, this is unlikely because UHI effects on trends are largely indistinguishable from global warming. Current homogenization techniques can remove abrupt changes in station data, but cannot correct for any sources of slowly-increasing spurious warming.

Anthony Watts has approached this problem for the U.S. temperature monitoring stations by physically visiting the sites and documenting the exposure of the thermometers to spurious heat sources (active and passive), and comparing trends from well-sited instruments to trends from poorly sited instruments. He found that stations with good siting characteristics showed, on average, cooler temperature trends than both the poorly-sited locations and the official “adjusted” temperature data from NOAA.

I’ve taken a different approach by using global datasets of population density and, more recently, analysis of high-resolution Landsat satellite based measurements of Global Human Settlements “Built-Up” areas. I have also started analyzing weather station data (mostly from airports) which have hourly time resolution, instead of the usual daily maximum and minimum temperature data (Tmax, Tmin) measurements that make up current global land temperature datasets. The hourly data stations are, unfortunately, fewer in number but have the advantage of better maintenance since they support aviation safety and allow examination of how UHI effects vary throughout the day and night.

In this two-part series, I’m going to look at the latest official global GHCN thermometer (Tmax, Tmin) dataset (Version 4) to see if there is evidence of spurious warming from increasing urbanization effects over time. In the latest GHCN dataset version Tmax and Tmin are no longer provided separately, only their average (Tavg) is available.

Based upon what I’ve seen so far, I’m convinced that there is spurious warming remaining in the GHCN-based temperature data. The only question is, how much? That will be addressed in Part II.

The issue is important (obviously) because if observed warming trends have been overstated, then any deductions about the sensitivity of the climate system to anthropogenic greenhouse gas emissions are also overstated. (Here I am not going to go into the possibility that some portion of recent warming is due to natural effects, that’s a very different discussion for another day).

What I am going to show is based upon the global stations in the GHCN monthly dataset (downloaded January, 2023) which had sufficient data to produce at least 45 years of July data during the 50 year period, 1973-2022. The start years of 1973 is chosen for two reasons: (1) it’s when the separate dataset with hourly time resolution I’m analyzing had a large increase in the number of digitized records (remember, weather recording used to be a manual process onto paper forms, which someone has to digitize), and (2) the global Landsat-based urbanization data starts in 1975, which is close enough to 1973.

Because the Landsat measurements of urbanization are very high resolution, one must decide what spatial resolution should be used to relate to potential UHI effects. I have (somewhat arbitrarily) chosen averaging grid sizes of 3×3 km, 9×9 km, 21×21 km, and 45 x 45 km. In the global dataset I am getting the best results with the 21 x 21 km averaging of the urbanization data, and all results here will be shown for that resolution.

The resulting distribution of 4,232 stations (Fig. 1) shows that only a few countries have good coverage, especially the United States, Russia, Japan, and many European countries. Africa is poorly represented, as is most of South America.

I’ve analyzed the corresponding Landsat-based urban settlement diagnoses for all of these stations, which is shown in Fig. 2. That dataset covers a 40 year period, from 1975 to 2014. Here I’ve plotted the 40-year average level of urbanization versus the 40-year trend in urbanization.

There are a few important and interesting things to note from Fig. 2.

  1. Few GHCN station locations are truly rural: 13.2% are less than 5% urbanized, while 68.4% are less than 10% urbanized.
  2. Virtually all station locations have experienced an increase in building, and none have decreased (which would require a net destruction of buildings, returning the land to its natural state).
  3. Greatest growth has been in areas not completely rural and not already heavily urbanized (see the curve fitted to the data). That is, very rural locations stay rural, and heavily urbanized locations have little room to grow anyway.

One might think that since the majority of stations are less than 10% urbanized that UHI effects should be negligible. But the seminal study by Oke (1973) showed that UHI warming is non-linear, with the most rapid warming occurring at the lowest population densities, with an eventual saturation of the warming at high population densities. I have previously showed evidence supporting this based upon updated global population density data that the greatest rate of spurious warming (comparing neighboring stations with differing populations) occurs at the lowest population densities. It remains to be seen whether this is also true of “built-up” measurements of human settlements (buildings rather than population density).

Average Urbanization or Urbanization Growth?

One interesting question is whether it is the trend in urbanization (growing amounts of infrastructure), or just the average urbanization that has the largest impact on temperature trends? Obviously, growth will have an impact. But what about towns and cities where there have been no increases in building, but still have had growth in energy use (which generates waste heat)? As people increasingly move from rural areas to cities, the population density can increase much faster than the number of buildings, as people live in smaller spaces and apartment and office buildings grow vertically without increasing their footprint on the landscape. There are also increases in wealth, automobile usage, economic productivity and consumption, air conditioning, etc., all of which can cause more waste heat production without an increase in population or urbanization.

In Part II I will examine how GHCN station temperature trends relate to station urbanization for a variety of countries, in both the raw (unadjusted) temperature data and in the homogenized (adjusted) data, and also look at how growth in urbanization compares to average urbanization.

5 21 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
January 17, 2023 2:17 pm

Sounds like a pandemic.

January 17, 2023 2:38 pm

The question is what we’ll be able to do about it. We have a USCRN but the rest of the world doesn’t. Until then, the discussion about climate science really goes nowhere.

Richard Greene
Reply to  Walter
January 18, 2023 1:05 am

There are very few discussions about climate science. Most discussions are about the fantasy CAGW boogeyman, completely unrelated to science, and the Nut Zero panic reaction to the fake CAGW boogeyman. This is climate politics, not climate science and grid engineering.

Reply to  Walter
January 18, 2023 9:49 am

nope but if you

  1. use roys land class dataset.
  2. use uscrn as a standard!
  3. classify all global stations using #2

youll find that 50% of all stations look like CRN

Tom Halla
January 17, 2023 2:48 pm

Which is why I would trust UAH satellite numbers more than surface stations as a measure of warming. Most of the world has very few ground stations, so claiming “global warming” from mostly urban heat island contaminated sites that are not randomly distributed is silly.

Rud Istvan
Reply to  Tom Halla
January 17, 2023 3:20 pm

Yes. It is silly. And one of the general climate model problems is that they are unavoidably parameter tuned to best hindcast 30 years (an official CMIP required output) against a surface record we already know is too hot due to UHI. So of course they then run hot in their forward projections. By a factor of about 2x.

Based on such provably faulty models:

  1. Summer Arctic sea ice should have disappeared. It didn’t.
  2. Sea level rise should have accelerated. It didn’t.
  3. Glacier National Park glaciers should have disappeared. They didn’t.

The whole thing would have collapsed long ago but for the enormous money momentum the thing has also ginned up.

Reply to  Rud Istvan
January 17, 2023 3:59 pm

Summer Arctic sea ice should have disappeared. It didn’t

No CMIP climate model I can find predicts that Arctic summer sea ice would be gone this soon. The IPCC currently projects “practically ice-free conditions at [Arctic] summer minimum at least once before 2050”.

Sea level rise should have accelerated. It didn’t.

It did.

Glacier National Park glaciers should have disappeared. They didn’t.

Which CMIP model makes this forecast? I can’t find any that project Glacier National Park glaciers should have disappeared by 2023. I stand to be corrected.

Rud Istvan
Reply to  TheFinalNail
January 17, 2023 5:15 pm

Let’s play attribution. You lose.

  1. Prof. Wadhams of the UK said it would be gone by 2014. His expertise was self admittedly informed by climate models.
  2. It didn’t. For proof details see my long ago post here ‘Sea Level Rise, Acceleration, and Closure’. You need to study upon basic facts alarmists assert that are just wrong. Manufactured deliberately wrong. As explained in the post.
  3. it wasn’t a CMIP model. It was official National Park Service posters plastered all over the Park until embarrassingly taken down during the park closure 2020. So, trust the NPS?
Reply to  Rud Istvan
January 17, 2023 11:35 pm

Nailed !

Richard Greene
Reply to  Rud Istvan
January 18, 2023 1:14 am

Lets be more specific:
Every environmental prediction of doom in the past 100 years (at least) has been wrong — 100% wrong — that takes great talent. And a compliant mass media to hide that fact.

It is not fair to criticize the accuracy of climate models. They are not programmed for accuracy. Accuracy is not a goal. It never was. The goal is to scare people and generally support the CAGW prediction in the 1979 Charney Report, revised by the IPCC a few years ago.

If accuracy was a goal:
CMIP models would be getting more accurate over time, rather than less accurate, and

The Russian INM model, which least overpredicts the global warming rate, would get 99% of the attention rather than 1%

When people criticize the climate computer games for inaccuracy, it makes me laugh — they are deliberately inaccurate for use as climate propaganda.

That’s why I have called them then climate computer games since 1997, and the people who abuse them for propaganda are Climate Howlers, Green Dreamers and Nut Zero Zealots.

Reply to  TheFinalNail
January 17, 2023 5:39 pm

Sea level ris should have accelerated. It didn’t.

It did

Not according to a 2019 study in Nature: “A revised acceleration rate from the altimetry-derived global mean sea level record” which Willis Eschenbach mentioned here a couple weeks ago.


It’s always been a conundrum that individual tide gauge station data show no acceleration in what should be a phenomenon measurable globally even accounting for local land subsidence or post-glacial rebound.

Richard Greene
Reply to  stinkerp
January 18, 2023 1:16 am

You can’t measure 1 mm changes in absolute sea level from a satellite that drifts up to 10 feet in orbit — it is a con job.

Relative sea level is much more important and tide gauges measure that well.

Steve Case
Reply to  TheFinalNail
January 17, 2023 7:09 pm

Sea level rise should have accelerated. It didn’t.
It did.

These 65 tide gauges from the Permanent Service for Sea Level (PSMSL):

 SanFrancisco, Fernandia, Honolulu, New York, Key West, Fremantle, Sydney,  Brest, Seattle, Helsinki, Baltimore, Balboa, Boston, Philadelphia, Los  Angeles, Pensacola, Sewells Pt, Galveston, Stockholm, Portland ME,  Marseille, Oslo, San Diego, Ketchikan, Victoria, Trieste, Charelston I, Astoria,  Newlyn, Trois-Rivieres, Poti, Slipshavn, Frederikshavn, Hirtshals, Aarhus,  Travemunde, Turku/abo, Korsor, Pietarsaari/Jakobstad, Kobenhavn,  Mantyluoto, Hornbaek, Gedser, Frederica, Esbjerg, W Terschelling, Tuapse,  Furuogrund, Visby, Ratan, Warnemunde 2, Wismar 2, Cuxhaven 2, Smogen,  Kungsholmsfort, Olands Norra Udde, Ijmuiden, Harlingen, Delfzijl, Den  Helder, Hoek Van Holland, Vlissingen, Maassluis, Galveston an La Jolla, 

when analysed for acceleration in mm/yr², a very tight distribution centered between 0.0 and 0.1 mm/yr² is produced as follows;

-.07> 1
-.01> 2
0.02> 5
0.03> 1

Any rational person would say that’s no acceleration.

The -.07 mm/yr² outlier is Trois-Rivieres on the St Lawrence in Canada.

Last edited 19 days ago by Steve Case
Reply to  TheFinalNail
January 17, 2023 11:38 pm

nasa insists on using satellite data beginning in 1992 even tho they know full well it is junk , as shown by one of their recent papers detailing problems syncing new satellite data with previous satellite data . satellites can’t measure sea level height along coastlines , where it matters and can be checked by observations , yet they use satellite data instead of empirical tide gauges . anyone with a functioning brain should question nasa junk sea level measurements.

Reply to  TheFinalNail
January 18, 2023 8:18 am

The so called acceleration only occurs because satellite data was spliced onto the end of earth based tidal data.
The earth based data shows no acceleration. Satellite data is less accurate than earth based data.

The data shows no acceleration.

Reply to  TheFinalNail
January 18, 2023 11:26 am

Well, the National Park Service used the term “climate models” or “current climate models”, or “computer models”. Do you think they built their own?

Reply to  BobM
January 18, 2023 11:29 am

another one

Reply to  BobM
January 18, 2023 11:29 am

All from my 2017 trip to Glacier.

Reply to  Rud Istvan
January 18, 2023 10:50 am

tuned to pre urban records, if tuned at all

Reply to  Tom Halla
January 17, 2023 3:29 pm

Which is why I would trust UAH satellite numbers more than surface stations as a measure of warming.

Why UAH and not RSS, or the other satellite data producers who broadly agree with the surface data?

Tom Halla
Reply to  TheFinalNail
January 17, 2023 3:35 pm

Because UAH is calibrated to radiosonde data, not reconciled with surface stations, like RSS.

Reply to  Tom Halla
January 17, 2023 3:40 pm

Can you substantiate that, or is it just something you heard?

Any time I’ve seen the Christy chart with the supposed radiosonde data no details are provided. What radiosonde data? Have you found it and checked it yourself?

As far as I know that old chart Christy keeps rolling out has never been published in any peer reviewed paper, which means he never has to supply the credentials.

Are you skeptical? I am.

Last edited 19 days ago by TheFinalNail
Tom Halla
Reply to  TheFinalNail
January 17, 2023 3:49 pm

UAH and RSS were formerly in agreement, until RSS drew criticism for their data. So, if your hypothesis is correct, either RSS was once providing bad data then, or is providing “corrected” data now. RSS changed, not UAH. QED.

Reply to  Tom Halla
January 17, 2023 4:13 pm

UAH and RSS have rarely been in agreement.

Formerly, RSS was the data set of choice for the likes of Christopher Monckon because it was running much cooler than UAH. It was the data set of choice for Lord M’s first run at ‘the pause’ here at WUWT.

UAH v6 cooled their warming trend, breifly bringing it into line with the v3 RSS trend. A short time later RSS v4 was released and its warming trend increased, being more closely aligned with the surface data.

As far as I know, both are currently accepted by the US Met Society, despite their sharp differences. Needless to say, Lord M has abandoned RSS and now uses UAH for his latest ‘pause’ series.

Last edited 19 days ago by TheFinalNail
Reply to  Tom Halla
January 17, 2023 5:29 pm

I prefer UAH, but I thought a reason for the difference with RSS was that UAH accounted for satellite orbital drift which affects temperature measurements while RSS does not. Please correct me if I’m wrong. I did not know about RSS being calibrated to surface stations, which sounds like a monumentally bad idea given the UHI biases of ground stations.

Reply to  stinkerp
January 17, 2023 5:55 pm

Oh, I see. The satellite drift error was corrected by UAH a couple decades ago. The use of ground station data is used by RSS to extrapolate missing areas of global coverage. Once again a really bad idea for a couple reasons. “Infilling” data that doesn’t exist is simply making up numbers that aren’t actual measurements and subject to a plethora of errors in estimation. Using ground station measurements demonstrated to be biased by urban heating effects inevitably introduces artificial warming into the data.

Last edited 19 days ago by stinkerp
Reply to  Tom Halla
January 17, 2023 7:31 pm

RSS panicked an mixed in model data to create a warming trend that didn’t exist in the raw data. UAH kept following real scientific principles.

Reply to  Tom Halla
January 18, 2023 5:51 am

UAH has changed their dataset numerous times over the years, sometimes substantially, often in response to significant errors being found in it:


Christy and Spencer have managed to reclaim some of their lower trend estimate in v6 that they lost due to errors being discovered in earlier versions. Time will tell if it lasts.

Anyone who thinks that the surface indexes are uncertain but places unrestrained faith in a single satellite temperature series is deluding themselves.

Tom Halla
Reply to  AlanJ
January 18, 2023 6:43 am

But if you trust in the good faith of GISS, BoM, or Hadcrut, you are either naive or delusional.

Rud Istvan
Reply to  TheFinalNail
January 17, 2023 5:06 pm

Yes. I did. There are 4 radiosonde sets Christy uses, plus two ERA reanalysis datasets that partly rely on radiosondes. Christy has previously published references to all six in footnotes to his Congressional briefing charts. You should do more of your own homework.

Reply to  Rud Istvan
January 17, 2023 5:13 pm

If you could link to them then I would be most grateful.

I have asked numerous times and never once been directed to the source.

It seems most odd that if Christy has this damning evidence he hasn’t bothered to put it to peer review.

Most odd.

Philip CM
Reply to  TheFinalNail
January 17, 2023 5:54 pm

Peer review wouldn’t be sanctioning any work not confirming the CAGW model.

Reply to  TheFinalNail
January 17, 2023 8:28 pm

Coffin nail’s comments show how rusty it got being away from WUWT for so long.

Find your own links bent nail.

Not odd at all.
Radiosonde data is collected by government. Let government submit the data to the recently proven corrupt backstabber pal review malfeasance process.

You can find that link too. Just use your nail.

Tom Abbott
Reply to  TheFinalNail
January 18, 2023 3:37 am
Chris Hanley
Reply to  TheFinalNail
January 17, 2023 4:34 pm

If Dr Mears does not trust his own product why should anyone else?
Carl Mears the principal at RSS: “… stronger case can be made using surface temperature datasets, which I consider to be more reliable than satellite datasets (they certainly agree with each other better than the various satellite datasets do!)”.
That is an irrational even bizarre explanation for his preference, Dr Mears has unwittingly discredited his own satellite data set while he has no authority to speak on behalf on Dr Christy and DR Spencer who I am sure stand by the quality of their product.

Last edited 19 days ago by Chris Hanley
Reply to  Chris Hanley
January 17, 2023 4:52 pm

That’s a great source.

Dr Mears was recognising the fact that temperatures derived from lower troposphere satellite data (microwave soundings of oxygen isotopes) are, pretty obviously, nowhere near as reliable as temperatures derived from surface thermometer sources. There are many fewer processing steps in the surface data for a start.

For years Dr Mears cautioned against using the RSS v3 to suggest that warming had paused (a caution blissfully ignored by Lord M here at WUWT). Mairs always said RSSv3 contained a known cooling bias and their V4 confirmed this.

Regarding which scientist I would trust more; the one who is most happy to say that he/she might be wrong is the one I would lean towards.

Last edited 19 days ago by TheFinalNail
Chris Hanley
Reply to  TheFinalNail
January 17, 2023 5:18 pm

All the global temperature data sets are estimates that all authors would concede, Dr Mears is not saying his product might be wrong but is wrong and he thinks the surface data is more accurate.
Drs Christy and Spencer have not to my knowledge made a similar statement.

Reply to  TheFinalNail
January 17, 2023 6:14 pm

Why would you trust what any scientist said? It’s the observational data that that does all the talking. Look at the data and draw your own conclusions. Or not. Your choice.

Reply to  TheFinalNail
January 17, 2023 10:13 pm

You can’t trust the surface data sets. At best, they’re badly corrupted with UHI. At worst, they’re badly corrupted with ‘homogenisation’ errors. We don’t talk about fudging the data anymore.

Gary Pearse
Reply to  TheFinalNail
January 17, 2023 7:36 pm

Nails, so, you stand corrected on sealevel acceleration and the the RSS satellite temperature having been degraded by tying it to the (worst) fiddled surface temperature set (RSS).

J. Hansen, the Father of Crisis Anthropo Global Warming, first head of the RSS team and the inventor of temperature fiddling to fit T models after the super el Niño of 1998 failed to set a new global T record (ie failed to break the late 1930s – early
1940s century highs, now pushed down over 0.5C, goalposts, shifted, etc.).

Regarding disappearance of summer Arctic Ice, as a long time commenter here at WUWT, surely you won’t claim to be unaware of the ‘scientific’ certainty of Arctic Ice disappearance called for about
over a decade ago (hey, nevermind Arctic ice, top clisci practitioners warned 2 decades ago that children in Europe would not know what snow was in a few years!). Yeah, you guessed it, they pushed the funeral for Arctic Ice (and presumably Arctic creatures) out of sight to 2050.

Come back in a few years and all the things they have you believing right now they’ll have to change again. You apparently weren’t here when the Father of CAGW predicted the West Side Highway (near his offices in NYC) would be under water by 2000, when reminded, he said he meant 2020! Except for an inch or two, the highway is still ~ 10ft above the water.

Tell me, what kind of person hangs on and even raises a defence for such a moble load of bull.

B Zipperer
Reply to  Gary Pearse
January 19, 2023 8:10 pm

The poster child for who doubles-down after being wrong, multiple times over decades is… wait for it… Paul Ehrlich. And he just wrote another book confirming that he has not learned anything.
James Hansen is an amateur by comparison. LOL

Reply to  Tom Halla
January 18, 2023 9:04 am

Dr. Spencer brought up a good point about the ground stations being mostly in urban locations, where they read high, which supports the 100 or so, hot running, politics-inspired, computer “temperature prediction” programs

After balloons and satellites, there is no need for ground stations.
Their readings should be ignored. The stations should be demolished

Tom Halla
Reply to  wilpost
January 18, 2023 9:10 am

I would advocate keeping ground stations, but with the caution they are subject to UHI and other siting issues.

Reply to  Tom Halla
January 18, 2023 1:19 pm

If the ground station are retired/demolished, then the computer program artist will not have the corrupt data to play more games

Tom Halla
Reply to  wilpost
January 18, 2023 6:20 pm

RSS has been corrected, so playing games with data does not involve particular sources.

Reply to  Tom Halla
January 19, 2023 6:31 am

Why include incomplete “worldwide” data known to be biased to the upside by urban conditions?

Ground stations are antiques, outdated.

Each time I see “corrected”, I get angry.
I am a scientist/engineer, who does not “correct” data.

I like the voluminous raw data, covering the whole world, including the southern hemisphere, as with balloon and satellites, that read scientifically realistic temperatures, compared to fantasy, politics-inspired computer programs

Last edited 17 days ago by wilpost
Tom Halla
Reply to  wilpost
January 19, 2023 6:49 am

NASA GISS is at the Sir Cyril Burt level of fabrication.

January 17, 2023 3:19 pm

In the global dataset I am getting the best results with the 21 x 21 km averaging of the urbanization data, and all results here will be shown for that resolution.“. What do you mean by “best”? Hopefully not the best-looking cherries.

January 17, 2023 3:27 pm

The thing I find odd about these claims, that UHI is the cause of global warming trends, is that the biggest warming trend even in Roy Spencer’s UAH data set, is in the Arctic.

Urbanisation in the Arctic?

Rud Istvan
Reply to  TheFinalNail
January 17, 2023 3:45 pm

No. Think about it a bit. The UHI heat rise is ‘real’, but not caused by CO2. The ‘Arctic amplification’ effect is also real; Lindzen explained it many years ago, So more real heat amplifying UAH heat measurements in the Arctic resulting in differentially more Arctic warming is an expected result. Just not because of CO2 induced global warming.

Reply to  Rud Istvan
January 17, 2023 4:19 pm

So more real heat amplifying UAH heat measurements in the Arctic resulting in differentially more Arctic warming is an expected result.

What does this sentence mean? What do you mean by “more real heat”? Lindzen explained what, exactly?

This is arm-waving.

What is causing the heat over the Arctic, over the oceans, over vast areas of desert?

Rud Istvan
Reply to  TheFinalNail
January 17, 2023 5:24 pm

You sir are probably a lazy failing troll. You know not. I know.

UHI causes more real atmospheric heat. It is not climate change related, it is surface absorption related.

You can easily look up Prof. Lindzen’s explanation of Arctic amplification. My own edification on it was in his office the day he kindly reviewed not just the climate chapter as I had requested, but the entire book Arts of Truth. (Look up the long Spitzbergen footnote in the chapter on continental drift to see his influence from a one day session.) I bought him lunch at the MIT faculty club. You?

Reply to  Rud Istvan
January 17, 2023 7:37 pm

If you watch the recent YouTube video of Drs Jordan Peterson and Richard Lindzen, the latter broadly states that early mechanisms for faster Arctic warming are now not favoured. Not much explanation is given.
Geoff S

Reply to  Rud Istvan
January 18, 2023 6:20 am

In the energy problems of the 70’s and early 80’s the term “thermal mass” was big. It seems as though it has transitioned to UHI. Bricks, etc warm in the sun and take awhile to come back to ambient.

Reply to  TheFinalNail
January 18, 2023 6:22 am

What he is saying is that UHI causes local heating. This heating is then amplified in the Polar readings. There is a real local increase in heat, it does then cause real Polar heatings. Its just that the local rise in heat is caused by the urban areas heating up as a result of building, not by CO2 driven warming. And this then causes Polar warming, via the amplficiation effect, which also has nothing to do with CO2 emissions.

I don’t know if he is right or wrong, but this is what he is saying, and its perfectly logical. It is just arguing from the premiss that urban warming is not CO2 driven.

As to vast areas of desert, I don’t know what you are referring to. Is there any evidence that the temperature is rising across them? Which ones? Measured when?

Reply to  Rud Istvan
January 17, 2023 4:40 pm

It is intersting that so called “polar amplification” works the opposite way in the SH to the NH. What is Lizden’s explanation for that observation?

Arctic warming about 4 times the global average (but mostly in winter) and Antarctic cooling. So {polar amplification) is an inappropriate name. Maybe North Pole amplification and South Pole reversal.

Or could simply be that peak solar intensity has been increasing over the NH for at least 1000 years and peak reducing over the SH for the same period.

Chris Nisbet
Reply to  RickWill
January 17, 2023 5:18 pm

He has talked about the difference between the two poles. The thing with Antarctica is that ground level is a couple of km above sea level, and the Arctic is not. From memory, I seem to remember him saying something like “it’s complicated” (wrt the poles, and ‘climate’ generally).
I think this might be where I saw him talk about this…

Reply to  Chris Nisbet
January 17, 2023 7:20 pm

Greenland and Antarctica are very similar blocks of ice. Both have elevation to 4000m and both are surrounded by water and sea ice. Greenland plateau has warmed 10C in January while Antarctica has cooled a little.

Rud Istvan
Reply to  RickWill
January 17, 2023 5:39 pm

RW, please up your game.
The Arctic is an ocean surrounded by continent. The Antarctic is a continent surrounded by ocean. Therefore the poleward heat transport dynamics are very different. (Hint, continents facilitate, oceans impede, poleward heat transport. Reason is basic—oceans have more latent heat capacity by a large factor.) That answer is trivial to anyone who knows anything about climate dynamics. Which apparently excludes you.
An Army recommendation. You should now follow the first rule of holes. When in one wanting out, first stop digging.

Reply to  Rud Istvan
January 17, 2023 6:56 pm

Greenland is an ice block surrounded by water the same as Antarctica. Greenland has an elevation of 4000m. This is similar elevation to the highest point in Antarctica.

The Greenland plateau has warmed 10C in winter. Antarctic plateau has cooled a little.

So both poles are behaving very differently. Polar amplification does not exist. Someone might claim North Pole amplification but that simply reflects the warming of all the northern oceans including the Arctic. The Southern Ocean is cooling.

So Greenland is doing the opposite of Antarctica. Both big blocks of ice surrounded by water and sea ice. One block experiences “polar” amplification while the other doesn’t. So it is not “polar” amplification. Just the North Pole is warming and the South Pole is cooling – what could be causing that? Surely anyone with an ounce of understanding about heat on Earth would look to the solar intensity changing over the two regions.

Richard Greene
Reply to  RickWill
January 18, 2023 1:26 am

Anarctica’s permanent temperature inversion, especially at the higher elevations, leads to a permanent temperature inversion and cooling by an increase in greenhouse gases.

It is obvious Antarctica is not warming from greenhouse gases, at least since the 1970s

The explanation is obvious — increasing greenhouse gases do not cause warming of most of Antarctica.

It does not matter why — it just matters that we observe the trend there and not expect it to change.

The world is at no risk from Antarctica ice melting. If there was any risk, we would have observed melting in the past 50 years. But we didn’t.

Reply to  Richard Greene
January 18, 2023 1:54 am

There is no temperature inversion over the Southern Ocean but it is cooling.

The reason the Southern Ocean and Antarctica are cooling is because the peak solar intensity has been trending down for about 1000 years. The same driver involved in the Arctic warming due to summer solar intensity increasing for the past 1000 years.

Richard Greene
Reply to  RickWill
January 18, 2023 2:08 am

You have no idea what the solar intensity was until the satellite age when it was measured at the top of the atmosphere by NASA satellites. Those data claim a small reduction of solar intensity TOA, barely enough to mention. The almost constant TOA solar intensity can not explain the lack of warming over Antarctica. So find another theory.

Reply to  RickWill
January 18, 2023 8:29 am

Greenland is surrounded by water, barely. On the west side there is not much water, and a continent on the other side.
There are two points of land that come close to Antarctica, S. America and Africa. Both are relatively small pieces of land, and both are much further from Antarctica than Canada is from Greenland.

Beyond that, is the entirety of Greenland included when calculating Arctic temperatures? If not, your point is meaningless.

Reply to  Rud Istvan
January 17, 2023 7:39 pm

Ric is discussing Greenland, not the Arctic.
Geoff S

John Hultquist
Reply to  RickWill
January 17, 2023 9:37 pm
Reply to  Rud Istvan
January 17, 2023 10:14 pm

But it’s still unbearably cold. Don’t pack your swimsuit yet.

Richard Greene
Reply to  Hivemind
January 18, 2023 1:30 am

Hold on: The world is getting warmer, and women are wearing fewer clothes to beat the heat. Yet another advantage of global warming!

For reference, here is my latest January 2023 Global Warming Fashion Show, which absorbs 97% of my climate research time, and may lead to a Nobel Prize, or at least a Nobel Prize Participation Trophy:

:Global Warming Fashion Show (onionbloggle2012.blogspot.com)

Reply to  TheFinalNail
January 17, 2023 3:48 pm

biggest warming trend even in Roy Spencer’s UAH data set, is in the Arctic.

Yep – the Arctic land has most warming in winter and the Arctic Ocean most warming in summer.

January temperature on the Greenland plateau up almost 10C in the past 70 years – now that is the front runner in the “global” warming stakes.

The Arctic Ocean is the front runner in the ocean “global” warming stakes with August temperature up 2C over the past 40 years.

By contrast both Antarctic plateau and Southern Ocean have long standing cooling trends.

There is NO “global” warming. Some place are warming at certain times, some places are cooling at certain times and some places have no long term trend. All due to the fantastic CO2 molecule of course. And people still believe this GHE garbage.

Last edited 19 days ago by RickWill
Reply to  RickWill
January 17, 2023 4:24 pm

By contrast both Antarctic plateau and Southern Ocean have long standing cooling trends. There is NO “global” warming. Some place are warming at certain times, some places are cooling at certain times and some places have no long term trend.

But there certainly is a global warming trend. Even UAH has a statistically significant one.

You pick out two specific places that you say have a cooling trend and deduce from that that there is no global warming trend; even when all the data sets, surface and satellite, UAH et al, say that there is.

Does not follow…

Reply to  TheFinalNail
January 17, 2023 4:55 pm

Does not follow…

Yes it does. Unless you use a qualifier of “average” there is NO global warming. Sure the “average global surface temperature” is increasing. However, as I wrote, some places are warming at certain times. Some places are cooling at certain times and some places have no trend over long periods.

I have attached the trend for the Arctic sea surface temperature. As you can see, there is near zero trend in the minimum value but the maximum value has increased by almost 2C in 60 years. The reason for that is obvious if you give it a moments thought.

By contrast the Arctic land temperature has near constant maximum of 5C but the minimum has increased from -30C to -20C in 70 years. So a 10C rise in winter when there is no sun. It takes a bit more thinking why that happens but the facts that Greenland and Iceland are both gaining permanent ice and NH snow extent are trending up is a solid clue for the reason.

Screen Shot 2023-01-18 at 11.50.05 am.png
Ron Long
Reply to  RickWill
January 17, 2023 5:19 pm

Guys, I think you should consider the probability that the “Arctic Warming” is mostly due to ocean current changes, wherein Japan Current/Northern Pacific sea water now has a tendency to divert more of the current to the north, passing between Alaska and Russia and entering the Arctic Ocean. This type of change is a re-distribution of heat, not a gain of additional heat. Sure looks like the Earth is recovering from the Little Ice Age, and we get a break before plunging into the next glacial cycle of the Ice Age we live in.

Reply to  Ron Long
January 17, 2023 7:12 pm

not a gain of additional heat. 

All oceans surface north of 20N is warming strongly in summer. August surface temperature up from 18.8C to 19.5C in the 40 years of the satellite era.

This is not due to minor changes in ocean currents. It follows the increasing summer solar intensity over northern latitudes and gradual loss of permanent summer snow cover and reduced summer moisture that allows land to get hotter in summer. More open ocean water is reaching the 30C limit in summer. . Semi-enclosed bodies of water like the Gulf of California are exceeding 30C in August.

Screen Shot 2023-01-18 at 1.59.26 pm.png
Last edited 19 days ago by RickWill
Reply to  RickWill
January 17, 2023 8:10 pm

Yes, there are many examples of large regions of Earth failing to follow an overall ‘global warming’ pattern. If we accept that early temperatures are best regarded as having uncertainty too high to make them reliable, there are recent examples, say in the satellite era from 1980 onwards.
Here is a straight plot of monthly UAH Australian lower troposphere temperature, showing 10 years 7 months of no warming trend to now, using customary (but debatable) linear least squares fit. That in itself deviates from an intuitive pattern of a steady, global T change related to a steady, global CO2 change.
Take a smaller part of this, from the 1.32C anomaly in May 2016 to the -0.56C value in November 2022. That is a cooling of 1.88C in 78 months that mathematically converts to MINUS 28.9C per century.
Yes, there is deliberate cherry picking with these numbers, but they are valid.
One has to ask why a cooling spell of 30C per century equivalent can in any reasonable sense be ignored when alarm is generated from an alleged century long warming of the globe of between 1 and 1.5C. Why do expert climate researchers avoid admission and discussion of such strong departures from their favoured mechanisms?
Remember as well that the alleged warming over that century is near certainly too high because if urban effects that Dr Spencer is validly demonstrating. Take away people and you can find flat responses such as this old graph from remote Macquarie Island ground measurements.
Geoff S

Richard Greene
Reply to  RickWill
January 18, 2023 2:14 am

“There is NO “global” warming.”

That is a ridiculous statement.

The average surface temperature of our planet is always warming or cooling, and has been doing so for 4.5 billion years\

The change in temperature at every location is a result of the net change in all the local, regional and global climate change variables.

Those variables are not all global and do not all have even effects everywhere.

There is no logical reason to expect every location to have the exact same climate change.

But there is every reason to expect constant changes in the average temperature of our planet.

There was global cooling from 1940 to 1975

Global warming from 1975 to 2015

And close to a flat trend (a slight decline) so far from 2015 to 2023

Reply to  TheFinalNail
January 17, 2023 6:18 pm

That’s not what is claimed in this article, you are just trolling. The exact claim is that some UHI effect remains in the “homogenised” surface data adding to the warming. At no stage was it claimed that it is the cause of the warming. Crawl back into your unscientific hole.

Reply to  TheFinalNail
January 18, 2023 8:24 am

Is mischaracterizing the arguments of others, really your best defense?

January 17, 2023 3:31 pm

The discipline of climate science has been positioned to live and die based on the temperature trends, not theoretical foundation. If there was a suitable theory which exists for climate gas forcing, the urbanization effects would be inconsequential. In fact, the level of urbanization could practically be deduced. But instead, the atmospheric models rely on statistical fitting to match surface data. This empirical inference is the achilles heel of the discipline, and it will ultimately lead to its demise. The hypothesis testing is a sham, because the models are calibrated to the ever changing properties of the surface. Many of the changes observed and captured into the parameterization have nothing to do with climate gas. This is pseudoscience, an epistemology based on belief and circular reasoning. The evidence stated is “we can’t think of anything else”, we tried, and we are the authority. This is a mockery of science. The classic philosophers would be ashamed.

Reply to  JCM
January 17, 2023 4:13 pm

Very well stated. On this point:

This is pseudoscience, an epistemology based on belief and circular reasoning. 

One of the most circular aspects of reasoning that few recognise is the use of ocean heat content to calibrate the ToA energy budget. The ocean heat retention over the last 70 years averages to around 1W/m^2. This is then equated to the ECS contribution of CO2. That requires the wild assumption that only CO2 has caused that heat retention.

The silliness in all this is that anyone with a iQ above the level of a gnat knows that it is impossible to heat water from heat input at the surface when the bottom water is constantly replenished with chilled water. Warming the surface increases evaporation and the water column cools down on average. In fact the only way oceans can warm at depth in decades is by reduced net evaporation as is being observed with reducing river runoff.

A recent paper provides useful insight into the time constants involved in heat transport from ocean surface into deep oceans:

 The mean transit times Γfrom the North Atlantic surface into the PSZ are about 1700 years, roughly 300 years older than the mean transit time from the North Atlantic into the PAZ, which underscores the fact that (as quantified below) roughly half the water last ventilated in the North Atlantic that enters the PSZ does so by diffusing up into the PSZ while traversing the PAZ. By the same token, the mean transit time from the Antarctic margin, where AABW forms, to the PAZ is around 1100 years, but to the PSZ it is 1400–1500 years. Mode/intermediate waters from the subpolar Southern Ocean reach the mid-depth PSZ on average in 1000–1200 years, which is roughly 200–400 years more quickly than their mean transit into the deeper PAZ 

Any useful climate model; needs to look at the physics of heat transport into deep oceans rather than relying on incompetent assumptions about all heat increase in deep oceans getting there in a matter of decades from surface energy balance. This is a classic case of the wrong assumption reinforcing the wrong belief.

Richard Greene
Reply to  JCM
January 18, 2023 2:22 am

” the atmospheric models rely on statistical fitting to match surface data.”

The climate computer games make predictions of future global warming trends that are unrelated to any past temperature trend observations. The games predict 2x to 3x faster warming than ever observed. Your statement is wrong. If the ECS predictions matched surface data since 1975, they would not scare anyone, and that is not allowed.

TCS predictions using RCP 4.5 seem to match surface data for the 1975 to 2015 global warming trend. They do not match the 1940 to 1975 cooling trend or the 2015 to 2023 nearly flat trend.

Reply to  JCM
January 18, 2023 3:09 am

The discipline of climate science has been positioned to live and die based on the temperature trends“.

I think this is incorrect. Climate ‘science’ lives on the availability of easy money obtained by scaring the sheep into bleating ‘save me, save me’.

January 17, 2023 5:22 pm

the latest GHCN dataset version Tmax and Tmin are no longer provided separately, only their average (Tavg) is available

Why? Weather stations measure max and min temperatures daily and provide the data. Why was it removed? My inner skeptic thinks it’s nefarious. The trend difference in daily high (slow increase) and low temperatures (rapid increase) matter and removing them from GHCN impedes understanding of global temperature trends and the mechanisms driving them. Some suggest increasing nighttime temperatures indicate that CO2 is reducing radiative transfer to space while others say it’s more from increasing latent heat absorbed by growing infrastructure affecting temperatures at the ground. Whatever the dominant mechanism, blending the data by averaging them obscures important differences, blurring our view. It seems like it mostly serves to amplify the narrative of alarming human-caused warming via CO2 and methane.

Richard Greene
Reply to  stinkerp
January 18, 2023 2:26 am

“Why? Weather stations measure max and min temperatures daily and provide the data. Why was it removed?”

For national security.
Keep asking questions and the Bidet FBI will raid your home.

January 17, 2023 6:24 pm

In the latest GHCN dataset version Tmax and Tmin are no longer provided separately, only their average (Tavg) is available.

Why would that be the case? The data is the data, not a function of it.

Clyde Spencer
Reply to  doonman
January 17, 2023 8:38 pm

This is disappointing to read because there is less information in Tavg than in Tmax and Tmin together. The daily mid-range averages were problematic for understanding what is happening. They have further degraded the quality of the data. I can’t help but wonder what the rationalization for this is.

Jim Gorman
Reply to  Clyde Spencer
January 18, 2023 4:00 am

There can really only be one reason. Independent Tmax and Tmin gives evidence that we are not going to “burn up” due to high temperatures. A less important reason is that models deal in average temps and not Tmax and Tmin forecasts.

Clyde Spencer
Reply to  Jim Gorman
January 18, 2023 12:52 pm

… models deal in average temps and not Tmax and Tmin forecasts.

Which is another short-sighted decision because if the raw data are Tmax and Tmin, that is really what they should be attempting to forecast. That is the best way to compare available empirical data with model outputs.

Reply to  doonman
January 18, 2023 6:57 am

It is invalid to add Tmax to Tmin then divide by 2 for a Taverage.
Tmin and T max are caused by different mechanisms. They come from radically different populations. The populations have different distributions.
They cannot be combined in a statistical sense. If they are, notably, the uncertainty of Taverage is so large that it is hard to claim any change over the 20th century because it all sits inside the noise envelope. The Tmin and Tmax numbers are not IID, not independent, not identically distributed.
Geoff S

Clyde Spencer
Reply to  sherro01
January 18, 2023 12:54 pm

Gorman has remarked to me that Tmax is sinusoidal, while Tmin is the result of an exponential decay.

January 17, 2023 8:13 pm

The Guardians (extraterrestrial chosen ones) are afraid Gaia is pulling the dust over our eyes here folks-

Peta of Newark
January 17, 2023 8:45 pm

We all enjoy a good puzzle – here’s one:

Attached is a screenshot of the (5-minute data-point interval) record from a weather station near me in South Norfolk from about 24 hours ago
i.e. Midnight to noon on Jan 17th

What is anyone’s explanation for what happened to the Temp/Dew point graph between 03:00 and 06:00 UTC

Why, when the wind dropped, did the temp graph go haywire?
The town of Wisbech is not very big, its a long way from anywhere that is ‘big’ and all the other local stations showed the same or very similar anomaly – whether they there were in the town, rural upwind or downwind.

The wind dropped and coincidentally, so did temperature.
When the breeze picked up again, from same direction and about same strength as before, the temp came back up with it

So what happened or especially and as it was before sunrise, how did CO2 do that

edit to PS
See also, from the very instant of sunrise at around 08:15, the temperature graph skyrockets – are you **really** sure Earth’s atmosphere is transparent to solar radiations?
I mean, really really sure?

Wisbe 17 Jan 23.PNG
Last edited 18 days ago by Peta of Newark
Dodgy Geezer
January 17, 2023 9:36 pm

What is the point of proving that warming contains an element of UHI?

If anyone repeats this in a technical paper they will have their grant withdrawn. Under such circumstances; the science HAS to say that temperatures are continuing to rise entirely due to CO2. There is no other option – if scientists want to continue to eat. ….

Richard Greene
Reply to  Dodgy Geezer
January 18, 2023 3:17 am

NASA-GISS “adjusts” for UHI globally.

I investigated and wrote an article for my blog a few years ago. The total adjustment was about -0.05 degrees in a century — a very small decline to eliminate UHI. The temperatures were slightly reduced to eliminate warming caused by the CHANGE of UHI over time.

Just enough of an adjustment for NASA-GISS to say they made an adjustment. The other temperature compilations did not bother with their own UHI adjustments, after seeing the NASA-GISS UHI adjustment was so small.

When I examined why the NASA-GISS numbers were so small, I discovered the puzzling reason:

There were plenty of weather station downward adjustments, as expected to reduce the effect of increasing UHI from economic growth,

It was assumed that most temperature stations had some UHI that should be removed, reducing their measured temperature slightly.

The surprise was that almost as many stations had their temperatures adjusted upward for UHI, offsetting almost all the downward revisions, and resulting in a tiny net change over a century.

Why would any stations need an upward temperature adjustment to account for changes of UHI:

That says UHI declined as a result of economic growth or a move of the weather station.

Maybe when an urban weather station is moved to an airport in the suburbs, the UHI effect was claimed to be reduced? i can’t imagine that an new airport weather station surrounded by concrete, asphalt and hot jet exhaust, with air traffic increasing over time, has less UHI than the old city weather station location. But it appeared that NASA-GISS assumed that.

A further complication is that NASA- GISS did not lower the current temperature station numbers to eliminate UHI — they raised historical temperatures for their UHI adjustments. That was confusing, maybe deliberately. but he adjustment was so small it barely changed the global average temperature. The wild guess temperature numbers infilling is potentially a much bigger problem than a suspiciously small UHI adjustment.

My conclusion:
This UHI adjustment was NASA-GISS junk science, done only so they could claim “we adjust for UHI”. I suspect the almost insignificant effect of their UHI adjustment on the global average temperature was decided BEFORE creating their UHI adjustment methodology.

My article from 2019 is at the link below — the longest article I’ve ever written for a blog. Unfortunately, due to a serious vision problem at the time, the formatting after the Summary is unconventional narrow columns that may not be readable on a smart phone: I reduced the very large original font just now but don’t have the energy to reformat the entire article.

Honest global warming chart Blog: Urbanization bias adjustments are tiny, and they are science fraud, partially caused by the lack of rural weather stations with long, continuous records, located outside the US, needed to determine the correct adjustments (elonionbloggle.blogspot.com)

Last edited 18 days ago by Richard Greene
Pat from Kerbob
January 17, 2023 10:23 pm

This seems like the money shot

“the latest GHCN dataset version Tmax and Tmin are no longer provided separately, only their average (Tavg) is available.”.

Just another way to hide data, especially if Tmax shows a decrease.

Can’t have that

Richard Greene
January 18, 2023 1:00 am

It’s obvious economic growth will cause some of the global warming trend. Not that anyone could measure the effect. It’s also obvious that 70% of the planet are not affected — the oceans.

So even if it was possible to get real raw temperature data for each US weather station, and we could somehow determine the effect of economic growth, how would that change climate scaremongering? It would not change the climate scaremongering at all

The climate scaremongering began in 1979 with the Charney Report, just 4 years into a new warming trend that started in 1975. Which followed global cooling from about 1940 to 1975. The prediction of ECS in 1979 had nothing to do with the actual warming trend from 1975 to 2015 — that was not known in 1979. The Charney Report ECS was used by the IPCC until a few years ago, when the IPCC arbitrarily narrowed the ECS range, which was a wild guess to begin with, and apparently was not scaring people enough with the +1.5 degrees C. lower limit.

No matter how accurate, or inaccurate, the historical temperature data are, they have almost no effect on climate scaremongering. Scaremongering is not based on any past climate trend. It’s a boogeyman unrelated to past climate observations, scaring people effectively since 1979, and revised a few years ago.!

Last edited 18 days ago by Richard Greene
January 18, 2023 5:44 am

For the US, at least, the USCRN is identical to the full CLimDiv network. Where is the evidence of urbanization bias in the full network?

comment image

Last edited 18 days ago by AlanJ
Bob Rogers
January 18, 2023 8:16 am

You know something I never hear about? Here in the USA (according to the EIA) we burn 369 million gallons of gasoline a day in cars. That means we release about 44,280 million million (44^15) BTUs of heat. If I’ve done the math right, that’s about 200 BTU per hour per square yard. That’s gotta have some impact on the surface temperature. I mean, that’s a little less than 10% of what the sun provides on average. And I didn’t even count diesel.

Maybe it’s just the act of burning fuel that’s causing warming.

January 18, 2023 9:47 am

I have (somewhat arbitrarily) chosen averaging grid sizes of 3×3 km, 9×9 km, 21×21 km, and 45 x 45 km. In the global dataset I am getting the best results with the 21 x 21 km averaging of the urbanization data, and all results here will be shown for that resolution.

well thats a flawed approach. ive been working with that land class dataset for years

  1. use a circular buffer!!!!!!!!!!!!!!!!!!!
  2. use crn stations as a standard.

UHI effects can occur from as little as 1sq km of paved surface….. see basset.

if you have urban classifier you need to show the verification stats

that is false positive and false negative

D Boss
January 19, 2023 5:14 am

I do not doubt that UHI exists. Nor do I question Roy Spencer’s data analysis skill. However there is a simpler way to determine what’s going on: Plot the value of the official “adjustments” to the temperature record, against CO2 concentration:


A graph of NOAA USHCN adjustments vs CO2 concentration between 1960 and 2015 is presented in the link above. The fact the adjustments are congruent with the the CO2 rise, shows the adjustments have nothing to do with “correcting” UHI. It is plain and simple – make the data conform to the warming hypothesis, which is of course not “scientific”.

Rather it is fraud, and propaganda!

Or as Spencer correctly points out, METAR data* is reliable – and upon which aircraft safety is dependent, hence it cannot be “adjusted” falsely and you end up with the following:


  • *for those who do not grasp what a METAR is, or why it cannot be adulterated – it is the hourly readings at airports worldwide. Why it’s important is planes generate lift based on the quality of the air – primarily it’s density, and that is dependent upon the barometric pressure and physical altitude and temperature/dewpoints. A false reading of any of these data can and has caused devastating air crashes at the most critical times for planes – taking off and landing. (well I suppose you could adjust the METAR data, but at great risk and outcry from the crashes that would occur – and temperature.global grabs this data in real time, before any manipulations can be applied to historical data)
  • And yes, airports do potentially suffer from UHI, however that is irrelevant to their purpose, which is to provide critical safety data for the planes taking off and landing in those UHI conditions!
%d bloggers like this:
Verified by MonsterInsights