Scafetta on UHI

Guest post by Rud Istvan

Charles the Moderator asked via email whether the following Scafetta paper had any merit? I opined yes, so here is another possible guest post.

This new paper’s abstract, to be published in Global and Planetary Change 10/19, (not a prestigious journal) reads as follows:

Global and Planetary Change

Volume 181, October 2019, 102989


Detection of UHI bias in China climate network using Tmin and Tmax surface temperature divergence

Nicola Scafetta a, Shenghui Ouyang b

Abstract Near-surface temperature records show that China warmed by about 0.8 °C from 1950 to 2010. However, there exists an ongoing debate about whether this warming might have been partially due to urbanization bias. In fact, homogenization approaches may be inefficient in densely populated provinces that have experienced a significant urban development since the 1940s. This paper aims to complement previous research on the topic by showing that an alternative approach based on the analysis of the divergence between the minimum (Tmin) and maximum (Tmax) near-surface temperature records since the 1940s could be useful to clarify the issue because urban heat island (UHI) effects stress the warming of nocturnal temperatures more than the diurnal ones. Then, the significance of the divergence observed in the data could be evaluated against the expectations produced by the CMIP5 general circulation model simulations. From 1945–1954 to 2005–2014, on average and over China, these models predict that Tmin had to warm 0.19 ± 0.06 °C more than Tmax. However, during the same period, the climatic records show that Tmin warmed 0.83 ± 0.15 °C more than Tmax. A similar analysis demonstrates that the effect is more pronounced during the colder months from November–April than during the warmer ones from May to October. A comparison versus China urbanization records demonstrates that the regions characterized by a large Tmin-Tmax divergence are also the most densely populated ones, such as north-east China, that have experienced a diffused and fast urbanization since the 1940s. The results are significant and may indicate the presence of a substantial uncorrected urbanization bias in the Chinese climate records. Under the hypothesis that Tmax is a better metric for studying climatic changes than Tmean or Tmin, we conclude that about 50% of the recorded warming of China since the 1940s could be due to uncorrected urbanization bias. In addition, we also find that the Tmax record from May to October over China shows the 1940s and the 2000s equally warm, in contrast to the 1 °C warming predicted by the CMIP5 models.


The time period suffices. Per published CMIP5 ‘experimental protocol’ documentation[1], the mandatory 30 year best hindcast is back from 2005 YE plus a forecast.

The diurnal heat difference between Tmin and Tmax has been discussed here many times. No WUWT explanation should be necessary. UHI is conclusively shown by the night-time and seasonal differences.

Chinese Tmin warmed mostly per UHI, not GW. So the Chinese can build as many USC/HELE scrubbed coal plants as they want with no significant climate impact per their ridiculous Obama agreement.

The only viable criticism of this new paper is the well-known problem that global climate models do NOT reliably regionally downscale (here, to China). For those references, see footnotes to essays ‘No Bodies’ and ‘Last Cup of Coffee’ in my ebook Blowing Smoke, or the review summary by Pielke and Wilby, Regional Climate Downscaling, What’s the Point? in EOS 93:52-53 (2012).

[update 9:45 pm Pacific Time.  There’s a much more comprehensive discussion of this over at JoNova]

0 0 vote
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
See - owe to Rich
July 25, 2019 11:09 am

Nice to see Scafetta diversifying from solely solar studies.

Tom Halla
July 25, 2019 11:37 am

It does appear that Tmin rises without equal Tmax rises would reflect UHI influences, and gives a reasonable way to quantify that influence.
It looks like Scafetta is on to something here.

Nick Schroeder
July 25, 2019 11:55 am

(not a prestigious journal)

Just who gets to decide?? The true believer boys in the smoke filled back room?

Why publish at all in some backwater rag that a few hundred might see and few dozen read?

Publish your work everywhere for the entire world to critique not just a privileged cabal of pay-to-play good-old-boys.

And that include WUWT.

Reply to  Nick Schroeder
July 25, 2019 12:19 pm

The best journals publish studies that are widely cited. I guess that’s sort of democratic. Scientists cite studies that seem to lead somewhere. You could say that each citation is like a vote for the journal.

Reply to  commieBob
July 25, 2019 3:29 pm

That’s the way it should work.
These days, all to often, author’s hunt for papers that can be interpreted as supporting theirs, then cite them.

Reply to  commieBob
July 25, 2019 3:29 pm

In theory, Bob, yes. Unfortunately, the “best” journals have drunk the cool aide and prevent any research not consistent with the ‘agenda’ from being published within their covers.

HD Hoese
Reply to  commieBob
July 25, 2019 5:36 pm

“ Scientists cite studies that seem to lead somewhere.” That’s a nice idealistic view, but my impression is that many (most?) journals are too far into impact factors, which apparently are more quantitative than qualitative. To where does it lead? Is it better to get a good idea, that might be against the current popular, grantable paradigm (fad?), published in a less read and cited journal than rejected by the “best”? It is not the more ideal world where the science I read left several decades ago. Languish a work might, but the ultimate worth is not necessarily the number of citations. Science is not democratic and the best is not necessarily judged by numbers. My conclusion is that the current judgement system we have now is new although perhaps with older more worthless elements.

There is also the impending question of the still apparently exponential increase in the number of journals driving librarians, and their supporters, crazy.

Reply to  Nick Schroeder
July 25, 2019 12:36 pm

Somehow I have a hard time seeing AGW “true believers” sitting in a smoke filled room. The accumulation of particulates would negatively affect the ‘snowflakes’.
My suspicion is that they promulgate their “fatwas” while sitting in rooms filled with nitrous oxide (laughing gas).

Reply to  Nick Schroeder
July 25, 2019 9:41 pm

Once upon a time, Nature and Science were two of the top-ranked scientific journals in the world.

Now, they have published so much nonsense, especially on global warming hysterics, that they have little remaining credibility.

A prime example of this climate nonsense was MBH98, a great steaming pile of horse pucks published in Nature and the centerpiece of the 2001 IPCC Third Assessment Report (TAR),

I say published on the internet, provide all your calculations in an Excel workbook or other to prove your work, and let everyone take potshots at you. That is real peer review, not the PAL review that has characterized the printed journals like Nature and Science for several decades.

Erik Pedersen
July 25, 2019 12:23 pm

Stop the world, I want to get off… Somebody in obscure are deciding above my head and I don’t like it…

Rud Istvan
July 25, 2019 12:45 pm

CtM was going to point out something we also discussed. JoNova posted a much longer analysis of this new paper on her blog yesterday, July 24. It includes a number of the excellent figures and heat island maps not mentioned in the abstract. Well worth a read at if you cannot access the full preprint.

July 25, 2019 1:25 pm

The only viable criticism of this new paper is the well-known problem that global climate models do NOT reliably regionally downscale (here, to China).

If the artificial CO2 emission “knob” applied in GCM simulations happens to be in the means of global instead of regional or local. then the regional downscale is impossible, even to contemplate it…!


Reply to  whiten
July 25, 2019 3:31 pm

well-known problem that global climate models do NOT reliably regionally downscale

Only in the world of clmatology can one be wrong everywhere, but correct on average.

July 25, 2019 1:37 pm

It’s remarkable how the steep decline in Tmax from ca. 1940 to ca. 1975 evident in Chinese measurements disappears in the model projections and in BEST’s regional “homogenization.” Is there an Amalgamated Data Fabricators union?

Robert of Texas
July 25, 2019 4:03 pm

So we are to just accept that TMAX is right? I personally believe it is also affected by UHI, just not as much. So looking at the delta between TMAX and TMIN will give you a better estimate, but it still will not be an accurate estimate.

Once again, we need to discard the urban based data and use well sited rural-only sites if we want to see the natural trend, as opposed to the urban-heat addition.

michael hart
July 25, 2019 4:11 pm

Richard Nixon understood China a lot better than climatologists ever will.

Reply to  michael hart
July 25, 2019 6:12 pm

Sadly he didn’t understand being a crook would cost him the presidency.

Samuel C Cogar
Reply to  Simon
July 26, 2019 3:54 am

Being a crook has never cost anyone the job of POTUS.

But getting the crooks in Congress and government Agencies really PO’ed at you will get you jailed, fired or terminated.

Tom Abbott
July 26, 2019 8:37 am

From the article: “The only viable criticism of this new paper is the well-known problem that global climate models do NOT reliably regionally downscale (here, to China).”

Why do we even need to downscale the global climate models? If we have regional data that coincides with other regional data from all over the world, that will give us the same results: a global profile. If the regional data from all over the world shows the same temperature profile, i.e., temperatures were just as warm in the 1930’s as they are today, then the regional data can substitute for the global climate model.

If every regional weather reporting station is reporting the very same data, i.e., that it was just as warm in the recent past as in the present, then who needs global models? We already have all the information we need. We don’t need GIGO computer models feeding us lies.

But since all we have now are GIGO computer models feeding us lies, our alternative is to switch to the more reliable data of the regional weather reporting.

Global computer climate models are FAKE SCIENCE concocted to defraud the People. If some people like dealing in science fiction then keep using them.

July 26, 2019 9:00 am

The term “heat island” describes built up areas that are hotter than nearby rural areas. The annual mean air temperature of a city with 1 million people or more can be 1.8–5.4°F (1–3°C) warmer than its surroundings. In the evening, the difference can be as high as 22°F (12°C). Heat islands can affect communities by increasing summertime peak energy demand, air conditioning costs, air pollution and greenhouse gas emissions, heat-related illness and mortality, and water pollution.

Clearly TPTB are not adjusting modern temps down sufficiently to counter act this massive influence.

Glen Martin
July 26, 2019 12:26 pm
Kevin kilty
July 27, 2019 7:15 am

“The only viable criticism of this new paper is the well-known problem that global climate models do NOT reliably regionally downscale (here, to China).”

Not only models, but that appears to be how the world operates. The Little Ice age was global, but it operated over slightly different time spans and with greater depth in different places. I would wager the same is true of every major climatic episode during the Holocene.

Howard Dewhirst
July 28, 2019 2:22 am

Please define ‘a prestigious journal’

%d bloggers like this: