The Gestalt of Heat Waves

guest post by Clyde Spencer


Tmax and Tmin time-series are examined to look for historical, empirical evidence to support the claim that heat waves will become more frequent, of longer duration, and with higher temperatures than in the past. The two primary parameters examined are the coefficient of variation and the difference between Tmax and Tmin. There have been periods in the past when heat waves were more common. However, for nearly the last 30 years, there has been a reversal of the correlation of increasing CO2 concentration with the Tmax coefficient of variation. The reversal in differences in Tmax and Tmin indicate something notable happened around 1990.


There was much in the press this Summer about the ‘global’ heat waves, particularly in France and Greenland. For an example of some of the pronouncements, see here. The predictions are that we should expect to see heat waves that are more frequent and more severe because of Anthropogenic Global Warming, now more commonly called “Climate Change.” The basis for the claim is unvalidated Global Climate Models, which are generally accepted to be running to warm. The simplistic rationale is that as the nights cool less, it takes less heating the next day to reach unusually high temperatures. Unfortunately, were that true, that would lead one to conclude that heat waves should never stop.

One agency that hasn’t erased the 1930s heat is the US Environmental Protection Agency. It clearly shows the 1930s with the largest heat wave index values!


Fig. 1. U.S. Annual Heat Wave Index, 1895-2015

If the predictions of worse future heat waves were valid, one might expect to be able to discern a change occurring already, inasmuch as it is commonly accepted that Earth has been warming at least since the beginning of the Industrial Revolution. That is, if the Summer heat waves are occurring more frequently, and they are getting hotter, one might expect that the maximum daily temperatures would exhibit larger statistical variance.

Because humans live on land, and we are concerned about the impact on humans, such as comfort and excess heat-related deaths, it would seem to be most appropriate to look at just air temperatures over land. There is an unfortunate tendency in the climatology community to conflate sea surface temperatures with land air temperatures, which tends to dampen changes because it takes a lot more energy to change the temperature of water than air or even land. Thus, with more than 70% of the surface of the Earth covered with water, small changes or trends in energy will be more difficult to identify in the weighted-averages of water and land temperatures.


To explore the situation, I used the Berkeley Earth Surface Temperature (BEST) project data. I wanted to get as close as reasonable to the raw data, minimizing the averaging that us usually employed in such data, and which raises concerns about subsequent statistical analysis. To that end, I downloaded an experimental data set that has the maximum and minimum daily air temperatures for the land-only surface of Earth. Figure 2, below, shows the coefficient of variation (CoV) of the maximum daily temperatures (Tmax) as a time-series from 1880 through mid-2019. There are two traces shown: 1) a 30-year moving-average* of the standard deviation (SD) divided by the 30-year arithmetic mean, with daily steps; 2) a 1-year moving-average of the SD divided by the annual arithmetic mean, with daily steps. The division by the mean normalizes the SD, creating the CoV.

I converted the BEST temperature anomalies to estimated Celsius temperatures by adding the calculated 1951 through 1980 average Tmax, to avoid an issue of dividing by zero. I then converted the temperatures to the Kelvin scale to allow the Tmax and Tmin CoVs to be comparable. The metadata accompanying the Tmax temperatures shows the estimated uncertainty of the baseline mean (14.41° C) to be ±0.11° C. Strictly speaking, the precision of the calculated anomalies should then be no greater than that value, and probably less, taking into account the uncertainty of individual measurements from which the anomalies are obtained. Nevertheless, BEST reports the anomalies to 3-significant figures to the right of the decimal point, rather than just the one (1) warranted. Ignoring that issue, and moving on ―

The annual moving-average of the Tmax CoV is not particularly informative, other than showing large annual changes in what is essentially the standard deviation. However, the 30-year moving-averages smooths the data considerably, albeit truncating the first and last 15 years of the data. Between about 1895 and 1950, there is no obvious trend. However, after that, the CoV shows a distinct upward trend as might be expected if Summer heat waves were increasing in frequency, duration, and/or temperature. However, something unexpected shows up around 1996 – the CoV starts to decline! The annual CoV values also suggest that there is a decline after about 2000. Surprisingly, the infamous 1930s U.S. heat wave is only weakly reflected in the annual Tmax CoV.


Fig. 2. Tmax Coefficient of Variation Time-Series (


Fig. 3. Tmin Coefficient of Variation Time-Series (

Interestingly, the CoV moving-average time-series for the Tmin looks different, as shown in figure 3, above. As with the Tmax time-series, it is a 30-year moving-average, with daily increments. Essentially, there is a decline in the variance from at least the mid-1890s, to about 1952, followed by an increase until about 1990, and then a return to the decline, at about the initial rate. The declines can be explained in the context of Earth’s radiative cooling being impeded by increasing ‘greenhouse gases.’ That is, it doesn’t get as cold at night, thereby decreasing the diurnal temperature drops, and consequently the Tmin variance. The almost 40-year interruption in the decline is a little more difficult to explain because there has been no similar change in the accumulation of CO2 in the atmosphere! (See Fig. 6, below.) That is a suggestion that any effects of long-lived CO2 are over-ridden easily, possibly by aerosols, short-lived clouds, and water vapor, if not actually being the dominant drivers.

What is happening? To explore further, I examined the most recent Tmax and Tmin data from BEST. Because the daily data are so noisy, I decided to use the monthly time-series to examine the behavior of the temperatures.


Fig. 4. Monthly Averages of High and Low Temperatures Time-Series


[ I have addressed the issue of temperature changes previously, here: ]

Again, Figure 4 is not very informative. It looks as though there may have been a slight increase in the slope of Tmax after about 1975, which is difficult to attribute to the effects of CO2. The rise in Tmin may have decreased slightly after about 1998, with the exception of the 2016 El Niño. Plotting the anomalies [not shown], instead of actual temperatures, accentuates the post-1975 increase in the slope of Tmax; however, the Tmin appears more uniform. Note that the theory of ‘Greenhouse Gas’ warming predicts that the effects should be most apparent in the Tmin. However, neither provides insight on what is happening with the CoVs around 1990!

However, a time-series plot of the monthly data showing the differences between Tmax and Tmin is much more interesting! As reported earlier, the difference has been declining since about the beginning of the 20th Century. However, as with the CoVs, there is a distinct change about 1990! After a decline in the differences for about a century, the differences start to increase. The 3rd-order polynomial regression, shown in purple, is of no particular importance other than to accentuate the change in the difference between the high and low temperatures. Albeit, it is suggestive of a possible 200-year cycle.

It appears that something subtle happened about 30 years ago that isn’t readily apparent in temperature (or temperature-anomaly) data alone. Figure 4 indicates that Tmax and Tmin are both increasing. Inasmuch as Tmax is much larger than Tmin, it will tend to dominate the resulting change, for a similar percentage change in both. It appears that both Tmax and Tmin were impacted similarly by the 2016 El Niño. The slope of Tmax is larger than Tmin between about 1985 and 2005. If Tmax is increasing more than Tmin, then it would argue against CO2 being the primary driver of global warming!


Fig. 5. High and Low-temperature Difference Time Series

The question is, “What is causing the changes around 1990, and is it of any climatological significance?” One possible explanation is that the apparent change in temperature relationships is somehow an artifact of processing; however, I’m not familiar enough with the details of the BEST processing methodology to speculate just how this might occur. Ignoring the rules of precision and error propagation comes to mind though.


Fig. 6. CO2 Concentration from 1958 through 2019


Figure 6 is a plot of the Mauna Loa measurements of CO2, from Scripps Oceanographic Institute. Looking at the figure, there seems to be little to explain the behaviors noted above other than a slight apparent decrease in the growth rate of CO2 after about 1990, for about two or three years. That is, if CO2 is the main driver of temperature changes, there doesn’t seem to be anything in the behavior of the CO2 concentrations that would obviously explain the recent long-term decline in the CoVs or the differences in Tmax and Tmin.

Assuming that the demonstrated CoV behavior is not an artifact and is real, the examined data suggest that if the global-average high temperatures are increasing, a consequence might be increased frequency or severity of heat waves. However, figures 2 and 4 only show the effects of the El Niño phenomenon, at best.


Extrapolations are always fraught with risk. However, based on the behavior of the Tmax CoV, which appears to be declining, there does not seem to be strong empirical support for the prediction that future heat waves will be worse and more frequent than in the recent past.

Clearly, Tmax has increased in the last 40-odd years. However, the CoV peaked about 30 years ago, and appears to still be in decline, based on the annual CoV values. Because Tmax is typically the result of direct solar heating, a decline in the so-called ‘solar constant’ variance could result in a decline in the Tmax CoV.

While Tmin is clearly increasing, almost monotonically, the CoV suggests that the increase is by increasing the floor, or base level, of the minimum temperatures.

One might be tempted to dismiss the effects that I have illustrated as being so small as to be inconsequential. However, the temperature differences, and the 30-year moving-average standard deviations, on which the CoVs are based, have a long-term duration and a magnitude comparable to several decades of average global temperature change. I believe that an explanation is warranted. Do any of the Global Climate Models show these secondary effects?

I would like to invite thoughts on the behavior of the coefficient of variation, and differences in the high and low temperatures, with respect to global temperature changes.


C. D. Keeling, S. C. Piper, R. B. Bacastow, M. Wahlen, T. P. Whorf, M. Heimann, and H. A. Meijer, Exchanges of atmospheric CO2 and 13CO2 with the terrestrial biosphere and oceans from 1978 to 2000. I. Global aspects, SIO Reference Series, No. 01-06, Scripps Institution of Oceanography, San Diego, 88 pages, 2001.

*It is more accurate to refer to the CoVs as a “sliding sample” because only the denominator is actually a moving average.

0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
September 6, 2019 6:29 am

May I ask, if its not co2, or the suns solar cycle, then what is causing the heat, rain, and wind to be stronger or weaker then “normal (I realise that there is no normal) I checked the bbc monthly weather report for september and they said autumn mite be late this year…. It would be nice to have a solid answer as to what is or could be causing the climate to be different, everybody talks about “climate change” in the news and social social media. If it isn’t co2, then everybody buying EV’s or building all the wind farms, solar farm are going to change nothing..

A C Osborn
Reply to  Sunny
September 6, 2019 7:51 am

Sunny, it is not CO2, the news & social media are only following the UN/IPCC agendas to destroy capitalism. They have even acknowledged it as their purpose.
You cannot beleive anything that comes out of the BBC, for examples go to

Reply to  A C Osborn
September 6, 2019 8:19 am

A C Osborn…. I wrote on another post that humanity was better of before mass news media and social media… If it rained or snowed or the winds blew and blew, it was simply down to mother nature. Now if it rains for more then three minutes, its “the worlds going to end”.. I have read some of the ipcc’s statements, and I am shocked that there is no mass reporting of such statements… it seems that they want the new “oil” money for themselves, ie: the green trillions……

Reply to  A C Osborn
September 6, 2019 8:49 am

O S Osborn. I agree, its about money, and the more the better, hence the trillion dollar plans by the usa presidential hopefuls… Ricdre wrote this quote in another post, and I think it speaks volumes… In 2010, Ottmar Edenhofer, co-chair of the IPCC Working Group III, said: “Climate policy has almost nothing to do anymore with environmental protection. The next world climate summit in Cancun is actually an economy summit during which the distribution of the world’s resources will be negotiated. It also explains why the IPCC and other rich climate change people spend millions a d fly all over the world for our future… Google spent millions recently on a party for the likes of obama in Scilly..

Reply to  A C Osborn
September 6, 2019 3:11 pm

Explain this then. Which fake media outlet is this one?
Sunny, if you really want a quick summation of the big picture that is representative of scientists who actually know their stuff, spend a few minutes reading this article. It is from the Royal Society. Hardly the work of quacks or the rabid media.

Clyde Spencer
Reply to  Simon
September 6, 2019 5:42 pm

There are reasons to believe that the formerly ‘sterling’ reputation of the Royal Society is not what it is today. They appear to be less than objective.

Reply to  Clyde Spencer
September 6, 2019 6:49 pm

“There are reasons to believe that the formerly ‘sterling’ reputation of the Royal Society is not what it is today. They appear to be less than objective.”
Translation. They don’t buy into the skeptic nonsense.

Reply to  Clyde Spencer
September 6, 2019 6:50 pm

“There are reasons to believe that the formerly ‘sterling’ reputation of the Royal Society is not what it is today. They appear to be less than objective.”
Be specific, which part is not right? Makes complete and reasonable sense.

Reply to  Simon
September 7, 2019 12:42 am

If you’re not skeptical, you’re not doing science right.

Reply to  Joz Jonlin
September 7, 2019 1:34 pm

Actually I think the work skeptic to describe some who write here is not accurate…. But I’m not allowed to use the “d” word.

Reply to  Sunny
September 6, 2019 8:29 am

What makes you say it’s not the sun causing climate change? There’s more to the sun’s influence upon Earth’s climate than just Total Solar Irradiance (TSI), i.e., W/m^2.

Reply to  icisil
September 6, 2019 8:52 am

Icisil… I would like to see conclusive evidence stated by a lot of respected scientists, that the sun is the cause of earths weather, for me to state that the sun is the cause of our weather.. so far I have not read conclusive evidence

Clyde Spencer
Reply to  Sunny
September 6, 2019 10:00 am

Because the sun is the ultimate source of virtually all the energy received by Earth, it should be a suspect. Indeed, it was looked at early in the research. Basically, the conclusion was that it was thought that the total variation in TSI from sunspot peak to trough was too small to have any significant impact. However, we don’t have quantitative measurements of the TSI (it was long thought to be a constant) before about 45 years ago. We also know almost nothing about how the distribution of energy with wavelength might have varied in the past.

One of the points I tried to make is that there are some subtle changes occurring in temperatures that aren’t obvious from just looking at average annual global temperatures. That suggests to me that there are some unknown or unexplained interactions, which may be part of feedback or amplification loops.

Rhys Jaggar
Reply to  Clyde Spencer
September 7, 2019 5:05 am

In order for TSI to be decisive, all the wavelengths of radiation must be equally important in providing energy to the earth’secsystem.

Has that been demonstrated to be true, or are specific wavelengths particularly critical?

If so, do such critical wavelengths exhibit greater variability when TSI varies e.g. does one critical wavelength vary by >5%?

These are absolutely fundamental questions and in over twenty years of following climate discussions outside academia/learned societies, I have never come across a single discussion of this point.

You see, if you wanted to get a much bigger response to small changes in TSI, you need to identify wavelengths where the amount of incoming energy is limiting the rate of events utilising such input energy (e.g. stratospheric chemical reaction). That is when you get maximal responsiveness to changes in input.

Reply to  Rhys Jaggar
September 7, 2019 10:21 am

True. I am inclined to think the amount of uv is critical as it brings the top layers of water to boiling i.e. evaporation.

Clyde Spencer
Reply to  Rhys Jaggar
September 7, 2019 10:44 am

Your point is well taken. We know that short-wave UV is critical to the formation and maintenance of ozone. That also means the presence of ozone impacts where in the atmosphere UV also cannot have an impact. We do know that the UV fraction of incoming solar energy varies considerably with the sun spot cycle. The strong absorption of UV and IR at the surface of water impacts how and where incoming EM energy can be absorbed. Vegetation and geologic materials have different absorption characteristics. It is commonly overlooked that the reflectance of vegetation and plankton is reduced by photosynthesis. But, the ‘absorbed’ red and blue light doesn’t result in heating, but rather, creating sugars. Clearly, the spectral distribution of energy should be taken into account in any modeling. However, I don’t think that the nuances of these other factors are known that well. For example, there is some disagreement about just how much the UV varies throughout the SS cycle. It isn’t even known if the variance is even linear.

From my viewpoint, with most of the influential climatologists denying that anything but anthropogenic CO2 is influential, it has inhibited scientists from exploring other possibilities and gathering data that might add to our understanding of the whole system.

Reply to  Sunny
September 6, 2019 12:25 pm


Q: If the Sun stopped radiating, how much cooler would be the surface of Earth?

Reply to  brians356
September 8, 2019 11:51 am

Brians356. I’m totally new to all of this, but I would guess that it would get much colder then it is now

Reply to  Sunny
September 6, 2019 8:31 am

You might want to research long-term ocean cycles, like the AMO. The UK weather is particularly dependent upon ocean temps.

I read some research about forty years ago concerning London, specifically. It hinted that the reduction of coal and wood fires used for heating in the city, at one time the only option, resulted in less particulate air pollution. This, in turn, led to a reduction in London’s famous fogs. The clearer air allowed for both warmer and colder episodes of weather.

I suspect there are quite a few possibilities other than CO2 that could result in weather variations.

Reply to  jtom
September 6, 2019 8:55 am

Jtom…. I cant remember the last time we had fog in london, so you may be right… england has weather which changes day it seems.. it can be hot and then rain daily

Clyde Spencer
Reply to  jtom
September 6, 2019 10:04 am

When my grant application is approved, and I’m assigned some graduate student assistance, I’ll get right on looking at ocean cycles. In the mean time, I can probably provide the best service by raising questions about inconsistencies and hope that others will pick up the banner.

John Collis
Reply to  jtom
September 6, 2019 11:37 am

Bad fogs (pea souped) in London were the norm in the 1950’s and only subsided with the introduction of the clean air act. The fogs caused a large number of respiratory diseases due to the smoke mixed in with the fog (smog).

Reply to  Sunny
September 6, 2019 9:20 am

Sunny: “May I ask, … what is causing the heat, rain, and wind to be stronger or weaker then ‘normal’…”

A large part of the answer to that question lies int the IPCC’s statement in AR3 that “The climate system is a coupled nonlinear chaotic system, and therefore the long-term prediction of future climate states is not possible”. A chaotic system will change in seemingly random ways even if there are no changes to any of the “forcings” to the system, but a chaotic system is not random at all, but instead completely deterministic provided you know the exact initial conditions of the system and you model the system perfectly. Even a small error in either the initial conditions or the accuracy of the model will quickly be blown up into large errors in the output of the model. The IPCC acknowledged this in latter part of the above statement.

If you are unfamiliar with chaotic systems, the old Nova program “The strange science of Chaos” is a good place to start learning about it:

Reply to  Sunny
September 6, 2019 9:36 pm

How has it “changed”? Okay, it’s technically constantly changing, but there’s a range and we’re nice and snug in it. 98.6 degrees Fahrenheit is regarded as normal human body temperature. What does it mean if you’re 98.8 or 98.4? Or what if it’s 97.8 or 99.4?

Reply to  Sunny
September 11, 2019 7:59 am

Temperatures, both min & max, are ‘affected by’ changes in the site where the thermometers are. Most official NWS stations are in urban areas and at airports, both of which have grown, raising the background temperature round the stations. The general term for this is UHI (Urban Heat Island).

And, lest you think the stations are immune to local effects, has shown that well over 90% of stations fail the NW own requirements or siting the station.

September 6, 2019 7:02 am

Hi Clyde.
Very interesting post. I have long advocated for people to start looking at Tmax to get clues about inclining or declining energy coming from the sun and to look at Tmin to get clues about the movement of the inside of the earth. The inside of the earth has been moving faster this past century, hence the reason for the warming noticed on the NH. [Magnetic stirrer effect: our core follows that of the sun]

As far as Tmax is concerned, I calculated a decline in Tmax (global) from 1994.5, let us say 1995.

You can simply click on my name to read about the consequences of the decline in Tmax on the world’s climate.

Reply to  henryp
September 6, 2019 7:34 am

I am now very wary of the Tmin figures, having seen the plots of retained heat by cities recently. These showed temperatures in the early morning and clearly showed significant heat retention by structures such as airports, buildings, roads etc. I am used to seeing generalised UHI, but had no idea that the problem extended so much to Tmins.

A C Osborn
Reply to  Phoenix44
September 6, 2019 9:05 am

Totally agree.

September 6, 2019 7:08 am

Interesting that the EPA’s graph doesn’t reflect the severe heat waves of the mid 1960s that the mid-Atlantic states suffered (and I remember as a youngster). Perhaps those events were more localized — I dunno.

Lance Wallace
September 6, 2019 7:20 am

Could the great changes in the number of stations over the years affect your results? What happens to the 1990 effect if you limit your dataset to long-lived continuous measurements?

Clyde Spencer
Reply to  Lance Wallace
September 6, 2019 10:21 am

Lance Wallace
The question you raise has more important implications. Such as, are those tasked with producing, and paid for producing, the data sets, providing the most useful data to the public for examining alternative hypotheses? Or, are they just ‘rubber stamping’ the official position? As an example, if SSTs and air temps are provided separately, it isn’t too difficult to combine them (Although, I don’t recommend it!); however, it is more challenging to try to untangle them once conflated. Similarly, just providing the average of mid-range temperatures destroys a lot of information.

I’m retired and don’t expect to live long enough to explore all the alternative possibilities. What is needed is a return to the former expectation of unbridled curiosity among young scientists, who are willing and able to examine such things and get their results published.

September 6, 2019 7:25 am

“There was much in the press this Summer about the ‘global’ heat waves, particularly in France and Greenland”

Warmer weather this summer in Alaska, and very briefly in Greenland and France, but cold records were broken all over the northern hemisphere. Growing Degree Days (GDD) have plummeted in key US growing regions (15-30% are typical).

“95% of US ZIP codes experienced a colder growing season in 2019 than 2018. On average, a US ZIP code received 87.2% of the heat accumulation as in 2019 — a 12.8% drop!”

Check 2019 decrease in GDD by US state here >

Paul R Johnson
September 6, 2019 7:41 am

The underlying question here is the reliability of the data. Are the stations subject to increasing encroachment and thus UHI effects? Is the recent trending the same if only the Climate Reference Network stations are considered?

Reply to  Paul R Johnson
September 6, 2019 9:18 am

“The underlying question here is the reliability of the data.”

Arguing over interpretations of historical worldwide temperature data strikes me as being like a home seller haggling with a buyer over contract details after the seller learns the buyer intends to pay 90% of the price with Monopoly money. 90% or more of the 1880-2019 worldwide temp dataset is fake; it was pulled out of thin air by climate scientists. Comprehensive reliable data do not exist.

A C Osborn
Reply to  icisil
September 6, 2019 11:11 am

icisil, I disagree, there is very little wrong with the Raw data, but I do mean Raw, not the current adjusted raw we are now being given by the data adjusters.

Reply to  A C Osborn
September 6, 2019 2:10 pm

Relatively speaking, raw data are almost non-existent. There were probably no more than 40 thermometer stations providing continuous coverage in the entire southern hemisphere from 1880-1950 (70 years). That’s half of the world for half of recorded history. It’s impossible to get an accurate measure of a global temp from that. The northern hemisphere was better during that period, but not much except for the US and western Europe. The image below shows 10 recording stations existed continuously from 1880-2019. From 1880-1950 there were probably 20 more, mostly concentrated in SE Australia. Everything else is Monopoly money.

comment image

Matthew Bruha
Reply to  icisil
September 6, 2019 11:40 pm
Reply to  A C Osborn
September 6, 2019 2:17 pm

And furthermore, it’s not that there’s something wrong with the raw data that exists, but that raw data hardly exist for the area under consideration. Fake, extrapolated data are not raw data.

Reply to  icisil
September 6, 2019 2:33 pm

Yes, data is only data once, when it is collected. Once data has been altered it is no longer data, it is an artifact of analysis.

There is no such thing as “adjusted data”. Data is what you get, not what you make.

da·ta ˈdadə,ˈdādə/ noun:
1- facts and statistics collected together for reference or analysis.

Reply to  icisil
September 6, 2019 3:17 pm

Fake data are not adjusted data. Adjusted data are altered raw data, i.e., actual measurements. Most of the temp data record is extrapolated data, i.e., location-specific data conjured out of thin air by magic from raw data from other locations. It’s not adjusted data at all (though it has the same effect); it’s imaginary virtual data.

Reply to  icisil
September 6, 2019 4:05 pm

Data cannot be created, it must be collected. Do not call it data, it is fiction.

September 6, 2019 7:49 am

I suspect it’s mostly in the sensors, Clyde. The more modern sensors sense temperature changes very quickly, have less mass, and are mounted on less heat conductive plastic stand-offs. Consequently, they read the smaller “puffs” of warm air microcells generated at ground level and convected to the 2 meter level much more readily. The result is more and higher high temperature readings, and higher averages of the data as equipment is changed for newer and “better” as decades of weather station maintenance goes by.

Reply to  DMacKenzie
September 6, 2019 9:32 am

Then are their not “puffs” of cold air as well?

Clyde Spencer
Reply to  DHR
September 6, 2019 10:30 am

I suspect that the answer to your question is that, in the absence of convection and turbulence induced by solar heating, it will usually not be as windy at night. Even so, such wind as there might be will mostly be impacting the Tmin. So, yes, the modern sensors with less thermal mass might be recording more extremes. That could be an explanation for the increasing difference in the delta.

Clyde Spencer
Reply to  DMacKenzie
September 6, 2019 9:46 am

How would more sensitivity to gusts of hot air explain an apparent decline in the Tmax CoV since 1996?

A C Osborn
September 6, 2019 7:55 am

Clyde take a look at the anti-corellation between Tropical Cloud Cover and the Global Temperatures, it is a far better fit than CO2.
It can be found here
You used to be able to post the graph, but for some reason it no longer works.

Clyde Spencer
Reply to  A C Osborn
September 6, 2019 10:40 am

A C Osborn
That relationship is exactly what I would expect. I think that the role of cloud cover does not get the attention it deserves. The “albedo” of Earth is commonly shown with 1-significant figure (i.e. 0.3), while the other parameters in energy calculations are more commonly shown with 3 to 5-significant figures. Clearly, clouds are a dynamic, transient event that can blink in and out in less time than the temporal-step times of Global Climate Models. I don’t think that there is good evidence as to whether global cloud cover is increasing, decreasing, or staying the same over long periods of time.

Tom Abbott
Reply to  A C Osborn
September 6, 2019 11:30 am

“You used to be able to post the graph, but for some reason it no longer works.”

The WUWT/Wordpress website is using 20th century comment software, after the website crashed when the software was upgraded. The new software worked great for a while, then something caused it to crash, and WUWT reverted back to a primitive comment software, that only updates once an hour and no longer allows the posting of pictures.

The original commenting software allowed pictures to post, but after the software upgrade and subsequent change, we are left with this rudimentary software fix we are currently suffering uder.

The website would probably get more visits if its comment software was more user friendly.

Hans Erren
September 6, 2019 8:10 am

I see 1950 discontinuity in the difference data, caused by the WMO prescription of Stevenson screen observation huts. Then a 1970 duscontinuity probably coincident with the introduction of automated recording.

Reply to  Hans Erren
September 6, 2019 9:45 am


are bending points
or points where heating changed to warming (1950) or cooling (1995)

Clyde Spencer
Reply to  Hans Erren
September 6, 2019 10:44 am

Can you be more specific as to what changes in the Stevenson screens would account for the increase in the temperature deltas? 1950 is a point of inflection, not a change in slope. 1970 is neither.

Reply to  Clyde Spencer
September 6, 2019 12:33 pm

here is a simple analysis of Tmax data coming from an Alaskan weather station.
[don’t trust Best!!)
You clearly see what happens in 1950, 1971, 1995 and ?
I made a small mistake with 2016 when I assumed the GB cycle to be 88 years, as per available reports when in fact I later found out it is 86.5 years (current) I find the the last bending point is in fact 2014 when there was a double switch over of the poles on the sun.

Clyde Spencer
Reply to  henryp
September 6, 2019 7:53 pm

My apologies for taking so long to respond. Your Elmendorf AFB graph was eye catching because of the close fit between the two curves. I wanted to think about it a bit before responding. I’m still at a loss at what to say. I think that is in part because the annotation needs to be improved. I presume that the horizontal axis is years since 1942. However, I can’t readily do the additions in my head to find calendar years. I’d like to see more than just one example before agreeing that the temperatures match a sinusoidal as a general rule.

September 6, 2019 8:49 am

I suspect the Arctic Temperature Graph might be showing the benign effect of CO2.

No urban heat to confound the results. It has been showing warmer nights, but no change in daytime max, for several years, now (realize that day and night are not 24 hour events there). I’ve read several times that the impact of CO2 should be warmer nights, because of the ‘insulating’ effect, but no change during the day; that CO2 capturing solar energy and radiating it right back into space counters the outgoing radiation from earth that is captured and re-radiated back to the ground.

While warmer nights might impact some of Earth’s biosphere, the range of natural variation is so much greater, that impact may be close to zero.

My apologies for such a simplistic description, but I tend to think in very simple terms. Still, it has served me well.

Robert W Turner
September 6, 2019 9:13 am

We had a day of 17 degree F below average temperature last week. Had we been 17 degrees above average the high temperature would have been 107 F and would have been global news with the typical Chicken Little headlines.

Steven Mosher
September 6, 2019 9:32 am

“I converted the BEST temperature anomalies to estimated Celsius temperatures by adding the calculated 1951 through 1980 average Tmax, to avoid an issue of dividing by zero. I then converted the temperatures to the Kelvin scale to allow the Tmax and Tmin CoVs to be comparable. ”

The daily file has the climatology.
You dont have to estimate anything.
Just add the climatology to the anomaly

“If Tmax is increasing more than Tmin, then it would argue against CO2 being the primary driver of global warming!”


Finally if you are looking for heat waves FORGET DRY BULB TEMPERATURE,
it will just confuse you.

Reply to  Steven Mosher
September 6, 2019 9:48 am

My advice:

Ignore Mosher.

Reply to  henryp
September 6, 2019 11:43 am

Ignore what?

Reply to  Steven Mosher
September 6, 2019 10:28 am

Dry bulb: “it will just confuse you.”

Maybe you explain why you say that. I would be more useful.

Now try to explain that the MSM and all the nat. meteo bureaux which bombard us with heatwave claims and “validate” readings from poor quality sites as GOAT temperatures and extreme heatwaves.

Clyde Spencer
Reply to  Greg
September 6, 2019 10:59 am

And the MSM provides us with the temperatures, NOT the Heat Index.

Clyde Spencer
Reply to  Steven Mosher
September 6, 2019 10:57 am

On the off chance that you will respond to me with something coherent, please explain why my statement is “WRONG.”

While Heat Index is a better measure than temperature alone of the subjective experience of people experiencing hot weather, the humidity is only one factor. In the absence of high temperatures there will be no “heat waves.” The question posed was whether there was empirical evidence of more frequent periods of hotter than normal weather. Dry bulb temperatures should be adequate to address the issue of more frequent high temperatures.

September 6, 2019 9:49 am

My advice:

Ignore Mosher.

September 6, 2019 9:51 am

The author states, to wit:

The question is, “What is causing the (temperature) changes around 1990, and is it of any climatological significance?” One possible explanation is that the apparent change in temperature relationships is somehow an artifact of processing; (nope, no processing error)

Figure 6 is a plot of the Mauna Loa measurements of CO2, from Scripps Oceanographic Institute. Looking at the figure, there seems to be little to explain the behaviors noted above other than a slight apparent decrease in the growth rate of CO2 after about 1990, for about two or three years.

The author needs to be “looking” through a, per se, “dissecting” microscope at said Figure 6 data (Mauna Loa CO2 ppm), in order to comprehend the “reasons” for the aforenoted “after 1990” …. “ slight decrease in the growth rate of CO2 ” and the “decrease in temperature”.

Please note in the following the presence of a “cooling” La Nina (1989) and an erupting Pinatubo (1992), to wit:

Maximum to Minimum yearly CO2 ppm data – 1979 to May 2018
Source: NOAA’s Mauna Loa Monthly Mean CO2 data base

CO2 “Max” ppm Fiscal Year – mid-May to mid-May

year mth “Max” _ yearly increase ____ mth “Min” ppm
1979 _ 6 _ 339.20 …. + …… __________ 9 … 333.93
1980 _ 5 _ 341.47 …. +2.27 _________ 10 … 336.05
1981 _ 5 _ 343.01 …. +1.54 __________ 9 … 336.92
1982 _ 5 _ 344.67 …. +1.66 __________ 9 … 338.32
1983 _ 5 _ 345.96 …. +1.29 El Niño __ 9 … 340.17
1984 _ 5 _ 347.55 …. +1.59 __________ 9 … 341.35
1985 _ 5 _ 348.92 …. +1.37 _________ 10 … 343.08
1986 _ 5 _ 350.53 …. +1.61 _________ 10 … 344.47
1987 _ 5 _ 352.14 …. +1.61 __________ 9 … 346.52
1988 _ 5 _ 354.18 …. +2.04 __________ 9 … 349.03
1989 _ 5 _ 355.89 …. +1.71 La Nina __ 9 … 350.02
1990 _ 5 _ 357.29 …. +1.40 __________ 9 … 351.28
1991 _ 5 _ 359.09 …. +1.80 __________ 9 … 352.30
1992 _ 5 _ 359.55 …. +0.46 Pinatubo _ 9 … 352.93
1993 _ 5 _ 360.19 …. +0.64 __________ 9 … 354.10
1994 _ 5 _ 361.68 …. +1.49 __________ 9 … 355.63
1995 _ 5 _ 363.77 …. +2.09 _________ 10 … 357.97
1996 _ 5 _ 365.16 …. +1.39 _________ 10 … 359.54
1997 _ 5 _ 366.69 …. +1.53 __________ 9 … 360.31
1998 _ 5 _ 369.49 …. +2.80 El Niño __ 9 … 364.01
1999 _ 4 _ 370.96 …. +1.47 La Nina ___ 9 … 364.94
2000 _ 4 _ 371.82 …. +0.86 La Nina ___ 9 … 366.91
2001 _ 5 _ 373.82 …. +2.00 __________ 9 … 368.16
2002 _ 5 _ 375.65 …. +1.83 _________ 10 … 370.51
2003 _ 5 _ 378.50 …. +2.85 _________ 10 … 373.10
2004 _ 5 _ 380.63 …. +2.13 __________ 9 … 374.11
2005 _ 5 _ 382.47 …. +1.84 __________ 9 … 376.66
2006 _ 5 _ 384.98 …. +2.51 __________ 9 … 378.92
2007 _ 5 _ 386.58 …. +1.60 __________ 9 … 380.90
2008 _ 5 _ 388.50 …. +1.92 La Nina _ 10 … 382.99
2009 _ 5 _ 390.19 …. +1.65 _________ 10 … 384.39
2010 _ 5 _ 393.04 …. +2.85 El Niño __ 9 … 386.83
2011 _ 5 _ 394.21 …. +1.17 La Nina _ 10 … 388.96
2012 _ 5 _ 396.78 …. +2.58 _________ 10 … 391.01
2013 _ 5 _ 399.76 …. +2.98 __________ 9 … 393.51
2014 _ 5 _ 401.88 …. +2.12 __________ 9 … 395.35
2015 _ 5 _ 403.94 …. +2.06 __________ 9 … 397.63
2016 _ 5 _ 407.70 …. +3.76 El Niño __ 9 … 401.03
2017 _ 5 _ 409.65 …. +1.95 __________ 9 … 403.38 (lowest CO2 ppm in 2017)

And iffen anyone has a problem “visualizing” the above “dissected data”, a composite graph of the plotted atmospheric temperatures and CO2 ppm quantities can be viewed by simply “clicking” on the following url link, to wit:

Is a picture worth 1,000 words?

Clyde Spencer
Reply to  Samuel C Cogar
September 6, 2019 11:10 am

No serious argument from me. You have basically reinforced my argument that there is an inadequate explanation for the behavior of CoV and temperature delta to be found in the CO2 alone. Things are happening for which CO2 is only a part of, and probably a small part, the drivers.

Reply to  Clyde Spencer
September 7, 2019 9:21 am

Clyde, iffen “CO2 is only a part of, and probably a small part, the drivers”, ….. then its driver part is “the little end of nothing”.

The only thing that the increase in atmospheric CO2 can legitimately be given credit for being a “driver” of is the present day enhanced growth of the “green” growing biomass.

Changes in atmospheric CO2 ppm quantity (as measured @ MLO) always tracks behind changes in temperature, … the temperature of the ocean surface water of the Southern Hemisphere.

The changes in the temperature of the near-surface atmosphere is not effected, affected or driven by the changes in atmospheric CO2 ppm quantities, … a scientific fact that is denoted on this composite graph of CO2 and air temperature.

The Southern Hemisphere ocean water is the only CO2 source/sink that is large enough to effect changes in the atmospheric CO2. Anthropogenic emissions of CO2 has no effect whatsoever on Mauna Loa measurements – 1958-2019 inclusively.

The temperature of the SH ocean water changes ….. 1) on a seasonal (equinox) cycle, …. 2) when an El Nino occurs, …. 3) when a La Nina occurs and/or ….. 4) when an enormous volcanic eruption (Pinatubo) occurs.

The seasonal (equinox) cycle, …. which determines the hemispheric temperature of the ocean water, is the primary driver of atmospheric CO2, as defined on this graph, to wit:

Keeling Curve Graph w/equinox lines

Bob Vislocky
September 6, 2019 10:05 am

Whether heat waves are increasing or not also depends upon the definition of a heat wave. If only looking at maximum temperatures, then heat waves probably are not increasing significantly. If a heat wave is defined to include minimum temperature or dew point, then heat waves probably are increasing. Different countries define heat waves differently. IMO, it doesn’t matter if heat waves are increasing because extreme cold weather is decreasing … and cold kills more than heat, so global warming is a net positive in this regard.

Greg Goodman
September 6, 2019 10:09 am

Thanks Clyde, this is very a interesting analysis, but are the link under figs 2 and 3 correct. They appear to be “anomalies” not Coefficient of Variation promised in the caption.

Clyde Spencer
Reply to  Greg Goodman
September 6, 2019 11:02 am

The links are to the source of the ‘raw’ data. I have done the calculations to derive the CoV.

September 6, 2019 11:17 am

Possibly the change with the Tmin has to do with cars. A rapid increase of the population owning cars after WWII followed by emission standards being implemented. In other words changes in pollution.

Clyde Spencer
Reply to  Bear
September 6, 2019 11:58 am

I’m afraid I don’t get your point. Clearly, smog became a significant issue in many metropolitan areas in the ’70s, before pollution controls were implemented. However, I don’t see anything happening in the graphs in the 60s through 80s. Can you be more specific in how you expect cars to impact the temperature records?

old engineer
September 6, 2019 11:46 am

Thanks for a very interesting post. I have always wondered why there was not more analysis using the approach you used. Until you know the CoV of a variable you can’t say much about the significance of a change.

However, I think that the data you used was too coarse. The WMO definition of a heat wave is “Five or more days during which the daily maximum surpasses the average maximum by 5 degrees C (9 degrees F) or more.”

I am sure there are times and places in the winter when this condition could be meet, For example if, for someplace in the winter, the average Tmax is 5 degrees C and the Tmax for 5 days is 10 degrees C that is technically a heat wave. But the MSM is not going to be screaming about a heat wave.

So why not look at only July or August (for the northern hemisphere) from year to year. And use a number of individual stations before looking at stations averaged together. I think this would have a number of advantages. First, you could choose stations not badly influenced by UHI. And you would be more likely to get days where Tmin is around sunrise and Tmax is in the late afternoon. Where I live it is not unusual in the winter to have a cold front arrive during the early morning, causing Tmax to be shortly after midnight and Tmin to be shortly before midnight that evening.

Just some thoughts. Again, enjoyed your post.

Clyde Spencer
Reply to  old engineer
September 6, 2019 12:07 pm

old engineer
Thank you for your comments. There are many ways in which the available data can be sliced and diced. Unfortunately, there is only so much that I’m willing and able to do for free. I do have a life. My hope was to point out some problems with the data that is available and how it is currently being used (Or NOT used!). Perhaps others will be willing to try out some of the other permutations.

As to being too coarse, if I had found nothing, then I would definitely agree that the data were too coarse to be useful. However, I found two phenomena that beg to be explained. Hopefully, others will pick up where I left off.

old engineer
Reply to  Clyde Spencer
September 6, 2019 1:42 pm

Clyde –

I’m sorry I left the impression that you should follow my suggestions. I hate it when someone says to me ” you should do thus and so.” I’m always tempted to say ” Why don’t YOU do it?” Notice that I didn’t say I would do the analysis I suggested and report back. Which I really should do, but probably won’t.

Again, thanks for the work you put into this. As you say you provided some results that should stir further investigation.

Richard M
September 6, 2019 12:45 pm

An analysis of regional data may add insight. It could be the changes are located on specific areas of the planet such as the Arctic. A warming Arctic is probably more humid which would lead to reductions in the diurnal temperature range. How this gets averaged with other regional changes might explain what has happened.

Reply to  Richard M
September 6, 2019 1:07 pm

Yes. Sometimes local data on Tmax give a clue.
Provided nobody manipulated them.
See my earlier comment.

Clyde Spencer
Reply to  Richard M
September 6, 2019 5:33 pm

Richard M
I have advocated that the various climate zones be examined to see if there are patterns in the zones, rather than lumping everything together as is usually done.

It is commonly acknowledged that the Arctic is experiencing more ‘warming’ than the rest of the world. However, I rarely see a definition of Arctic. Is it only that which is above the Arctic Circle, or does it include high latitudes where permafrost is found? In any event, I think that the relative humidity is playing a significant part in the ‘amplification’ of warming in the Arctic. Now, what about the Antarctic, which doesn’t seem to be experiencing similar warming except where there appears to be volcanic activity?

September 6, 2019 2:04 pm

“Assuming that the demonstrated CoV behavior is not an artifact and is real, the examined data suggest that if the global-average high temperatures are increasing, a consequence might be increased frequency or severity of heat waves.”

It really isn’t telling you anything about heat waves. CoV is a distraction; you might as well have plotted standard deviation. Dividing by the mean in K (I hope it was K) is just scaling by a near constant. It changes the scaling on the y-axis, but won’t change the curve.

It is the standard deviation of a daily global mean. That means that it depends a lot on the number in the sample, varying roughly as 1/sqrt(N). That changes a lot over the period. But it is insensitive to local heatwaves like the ones in France. Being a small area, their fluctuations disappear in the daily global average.

SD is also symmetric, increasing with cold variations just as much as hot. But although sensitive to variations in the tails (extremes), a general shift in temperature variation can also cause change in SD.

Clyde Spencer
Reply to  Nick Stokes
September 6, 2019 4:39 pm

It appears the you either didn’t read my article carefully, or you didn’t understand it. It is NOT derived from the standard deviation of the daily global averages. It is derived from the standard deviation of 1) 365 daily global Tmax values, and 2) 10,950 daily global Tmax values. The division isn’t just a scaling factor, but a correction for an obviously increasing Tmax. That is, if the standard deviation were computed for a sine wave that had a constant amplitude, the standard deviation would be constant. However, if that sine wave were combined with or added to a linear trend, the standard deviation with increase over time. The division by the moving average corrects for the linear trend, leaving the changes in the amplitude of a non-constant sine wave.

Yes, I know that SD is sensitive to decreases in temperature as well as increases, but, obviously the general trend is upward to Tmax. If you have a better metric for quantifying the frequency, duration, and amplitude of heat waves, let’s hear it.

Reply to  Clyde Spencer
September 6, 2019 5:48 pm

“It is derived from the standard deviation of 1) 365 daily global Tmax values, and 2) 10,950 daily global Tmax values.”
As I said, global averages. And the point is that any heat wave information is swamped by taking just that average for one day. Heat waves are local. As you noticed, the US heatwaves of the thirties didn’t show a blip.

“but a correction for an obviously increasing Tmax”
So what was the proportional increase in Tmax, in K (I’m still hoping you used K)? My guess is from about 287 to 288. Pretty small correction.

Clyde Spencer
Reply to  Nick Stokes
September 6, 2019 6:08 pm

I mentioned in the article that I did convert to Kelvin. Are you confirming that you didn’t read carefully, or is there some other excuse? Yes, the corrections were not large. However, I wanted to be sure not to give someone like yourself ammunition to complain that I had overlooked something, albeit a small effect.

Clyde Spencer
Reply to  Nick Stokes
September 6, 2019 7:41 pm

You remarked, “As you noticed, the US heatwaves of the thirties didn’t show a blip.” Actually, the 1930’s heat spell is more evident in the CoV of Tmax than in the raw temperatures (or anomalies). It may have been a regional weather event, or the data processing may have suppressed the older warming. But, I’m trying to deal with a broad brush approach here because the MSM is making the claim that heat waves will become more common around the globe in the future. My question was whether historical evidence existed to support the claim. It appears that the Tmax CoV did increase for more than 50 years. However, that trend appears to have reversed for about the last 23 years. Aren’t you the least bit curious as to why?

Michael Jankowski
Reply to  Nick Stokes
September 6, 2019 5:43 pm

“…CoV is a distraction…”

You have that in common.

September 6, 2019 2:15 pm

Could the drop-off in the last few years be a result of the averaging? Averaging or something data is the same, mathematically, filtering and will cause a delay or phase shift of the data.

Many years ago, in an electronics course on audio amplifiers, I was told that Filters will cause a phase shift of the audio signal. This phenomenon was reinforced in my control systems courses and in practice while trying to tune process control systems actually controlling a process, e.g. steam generator water level or feed-water flow. In process control we could compensate for this phase shift through the use of feedforward and Lead-Lag compensation. This is used in your home thermostat by the use of an anticipation setting.

Clyde Spencer
Reply to  Usurbrain
September 6, 2019 4:43 pm

The CoV is calculated as a centered, sliding sample to avoid ‘edge effects.’ That is why the first and last 15 years are missing.

September 6, 2019 3:09 pm

At arou d 1990 mzny weather stations started to replace mercury in glass themometers with electronic probes. These respond much more quickly to changes in temperature resulting in increases in Tmax and decreaes in Tmin.

Reply to  Gerard
September 6, 2019 3:15 pm

Differing daily measured values from the old and new sensors for temperature measurement spurred the author [Hager] to conduct a comparison spanning from January 1, 1999 to July 31, 2006 at Fliegerhorst Lechfeld (WMO 10856) 8-1/2 years long, daily without interruption, among other comparison tests of mercury maximum glass thermometers in a Stevenson screen and a Pt 100 resistance thermometer inside an aluminum enclosure, both unventilated. The 3144 days yielded a mean difference of +0.93°C; the Pt 100 was higher than the mercury thermometer. The maximum daily difference even reached 6.4°C!

Clyde Spencer
Reply to  Gator
September 6, 2019 5:51 pm

If the German study can be replicated and verified, it would be damning evidence that any claims about long-term changes cannot be supported. It would be unprofessional to continue to make such claims if verified.

Reply to  Clyde Spencer
September 10, 2019 11:28 am

Isn’t it amazing that the serial adjusters are so incurious about such a large problem? LOL

If I had the funds, I would replicate this study, but I don’t. I’m hoping someone out there will pick this up and continue this important work. We know with absolute certainty that the grantologists, the ethically challenged, will never investigate this, as they know it would mean massive adjustments downward for all the changed stations and the end of their imaginary trend.


Reply to  Gator
September 10, 2019 10:04 am

Gator… Why isn’t this public knowledge??? Why does proven science which disproves the ipcc and u.n not get out to the masses? To be honest I am sick and soo tired of not being told the truth.

Reply to  Sunny
September 10, 2019 11:31 am

Why? Do you really need to ask this?

When academia, pop culture, the “news” media, and our public education system are all owned by unhinged leftists, honesty is hard to come by.

Clyde Spencer
Reply to  Gerard
September 6, 2019 5:38 pm

If the change in instrumentation is responsible for the delta T changes, then that basically says that the old data are not comparable to modern data and comparisons are invalid. That would also suggest that the recent rise in Tmax and Tmin are artifacts of the instrumentation and they should be adjusted downward to compensate for the greater sensitivity to transient events.

Stephen W
September 6, 2019 3:24 pm

Since UHI is expected to behave similarly to predicted AGW, Figure 5 appears to show the impact of UHI between 1880 – 1980, with the recent warming since 1980 probably caused by the sun (higher maximums).

Clyde Spencer
Reply to  Stephen W
September 6, 2019 5:55 pm

Stephen W
My previous research also showed global Tmin is increasing faster than Tmax. My impression is that UHI is the opposite, with dark pavement driving up Tmax more than Tmin.

September 6, 2019 5:17 pm

I expect AGW via increase in GHGs to decrease the variance between hot days and cold days in the NH midlatitudes, because the Arctic is warming more than the tropics (and more than almost everywhere else) due to regional positive feedback.

I expect AGW to increase daily minimum temperatures only slightly more than daily maximum temperatures, while warming from growth of urban effects greatly increases daily minimum temperatures and has much less effect on daily maximum temperatures.

There is the matter of weather stations being moved more out-of-town when cities grow, and shifts in urban-vs-rural population of weather stations that are considered in a global land temperature dataset. A shift towards rural consideration will increase the spread between daily maximum and minimum temperatures, and vice versa.

Another factor is a trend that I think started sometime in the 1990s, and that is with airports that deal with a lot of overnight packages. A lot of package shipping industry infrastructure was built in previously rural areas adjacent to many of these airports. One airport where I have seen this cause an increase in daily minimum temperatures and a decrease in late night and early morning relative humidity is PHL airport at Philadelphia. As part of this package shipping industry buildup by PHL airport, the USPS moved their main Philadlphia metro area central operations from a big building by the big almost-Center-City multi-railroad train station to a newly constructed facility by PHL airport. I hope airports where this happened aren’t in the BEST dataset without adjustments that properly correct for this kind of stuff. Thankfully, PHL airport seems to not be in the USHCN or GHCN as of when I checked for that a few years ago.

Clyde Spencer
Reply to  Donald L. Klipstein
September 6, 2019 6:03 pm

Donald L K
You said, “I expect AGW via increase in GHGs to decrease the variance between hot days and cold days in the NH midlatitudes, because the Arctic is warming more than the tropics (and more than almost everywhere else) due to regional positive feedback.”

The data set I was dealing with was for global values, which means that for the NH Summer, most of the Tmax values were coming from the mid-latitude NH, while the Tmin values were probably from the mid-to high southern latitudes.

In any event, the Tmax values were predominantly coming from whichever hemisphere was experiencing Summer.

Reply to  Clyde Spencer
September 7, 2019 7:49 am

If in NH summer, the Tmax values were mostly from NH midlatitudes and Tmin were mostly from mid or high SH latitudes, then how were they only 12 degrees C apart? Or is Figure 4 showing different Tmax and Tmin figures than you’re talking about here?

Clyde Spencer
Reply to  Donald L. Klipstein
September 7, 2019 8:20 am

Donald L K
That is an inconvenient question. Because the individual data points are the global daily averages, there are Tmins and Tmaxes from both hemispheres, sampling opposite seasons. In theory, that should suppress the seasonality of the data. However, because there is more land in the northern hemisphere, the sampling is not symmetrical. What I should have said initially is that the extreme values would correspond to the season. That is, the coldest of the Tmin values would come from the Winter, and the hottest of the Tmax would come from the hemisphere experiencing Summer.

Reply to  Donald L. Klipstein
September 7, 2019 7:54 am

I just thought of another factor to consider: Were there changes in the considered thermometers towards a higher percentage of them in more humid or alternatively more arid parts of the world? That can have an effect similar to the effect of a shift of a higher percentage being in more urban or alternatively more rural places.

Fabio Capezzuoli
September 6, 2019 10:49 pm

Looks like the work I did examining how temperature volatility changes with time.

I have to agree in part with Nick’s critic: heatwaves are local, and considering global averages smooths them out.

But yes, I agree that volatility of temperatures, even at the global scale means something.

Clyde Spencer
Reply to  Fabio Capezzuoli
September 7, 2019 7:59 am

My thinking was that if there were only one or two heat waves, they will have their effects so diluted that one will not see them. HOWEVER, if they become more frequent and intense, then they might show up on a global scale. That was the impetus to explore the data. As it turns out, contrary to what Stokes is claiming, there is something happening that is impacting the coefficient of variance. Unless he or someone else can provide an acceptable alternative explanation, we are stuck with the obvious conclusion that there are ways to tease information out of data that isn’t obvious from just the temperatures. Stokes has a reputation of being a naysayer no matter how much evidence is presented that disagrees with his alarmist ideology.

Fabio Capezzuoli
Reply to  Clyde Spencer
September 12, 2019 9:09 pm

” Stokes has a reputation of being a naysayer no matter how much evidence is presented that disagrees with his alarmist ideology. “. Yes, I did notice that; he still makes good points, at times.

The visible changes in the CoV are there tho. Global temperature seems to have stopped becoming more volatile.

September 7, 2019 1:55 am

looking at the Elmendorf graph, you must remember that I did that investigation in 2013, so it would have included all data from 2012. I just looked at the latest results and indeed I did find Tmax rising again over the past 5 years at a rate of 0.04K/annum. That is not too far off bad from what I could have predicted back in 2014.

Indeed, I am not a mathematician, and so you might battle a bit reading my graphs. The one that might interest you is the first graph in this document here;

it is an estimate of the drop in the speed of warming on Tmax (global)

Important to note here is that this investigation was done in 2015 and included all results up until and including 2014.

Given this information, and solving the equation: 0.039 ln (x) – 0.1112 = 0 you can figure out the point in time where Tmax started declining,

i.e where the energy we get from the sun, that passes through the atmosphere as noticed on earth, started declining.
– do not confuse that with the energy coming from the sun measured TOA or anywhere in between TOA and sea level

Reply to  henryp
September 7, 2019 3:25 am

Looking at the results of my doc. there, some of you might want to ask me: but why is it not cooling?
It is indeed very puzzling to try and figure out what is happening. Remember we (in this case Clyde and me ) are evaluating air temperature and the statistic analysis I did is only one picture, just like any picture, frozen in time.
What actually happened since 1995 is best shown in this graph here:

For some reason, completely different to what I had expected, it appears that especially the sea waters in the NH became much warmer than anyone could have predicted.

IMHO there is no other explanation for this than that of increased volcanic activity and heat coming from the bottom up…… (Remember the eruptions in Iceland, Italy, Hawaii, Indonesia?)

but if someone here has another explanation?

Either way, I don’t think that we will escape the coming droughts. (click on my name to read my full report on that)

Clyde Spencer
Reply to  henryp
September 7, 2019 12:47 pm

You said, “I did find Tmax rising again over the past 5 years at a rate of 0.04K/annum.” I’m not sure how reliable that rate is. BEST says that their baseline ‘climatology’ for computing Tmax anomalies has an uncertainty of 0.11 deg C. There is no associated uncertainty for the daily anomalies. However, the monthly data has anomaly uncertainties typically greater than 0.5 deg C for the early data and ~0.1 deg C for recent data. I can’t vouch for the veracity of the uncertainty data, but I would tend to believe that anything less than 0.1 deg C is of low reliability. That is, only several years of observations would substantiate that the trend is even positive, and not a statistically random trend.

September 7, 2019 3:28 am

Looking at the results of my doc. there, some of you might want to ask me: but why is it not cooling?
It is indeed very puzzling to try and figure out what is happening. Remember we (in this case Clyde and me ) are evaluating air temperature and the statistical analysis I did is only one picture, just like any picture, frozen in time.
What actually happened since 1995 is best shown in this graph here:

For some reason, completely different to what I had expected, it appears that especially the sea waters in the NH became much warmer than anyone could have predicted.

IMHO there is no other explanation for this than that of increased volcanic activity and heat coming from the bottom up…… (Remember the eruptions in Iceland, Italy, Hawaii, Indonesia?)

but if someone here has another explanation?

Either way, I don’t think that we will escape the coming droughts. (click on my name to read my full report on that)

Clyde Spencer
Reply to  henryp
September 7, 2019 12:22 pm

While I suspect that the contribution of CO2 and heat from ~45,000 miles of undersea spreading centers is under appreciated, speaking as a geologist, it is my feeling that the impact on water is going to be too subtle to readily see in a simple temperature graph of surface waters. Typically, up-welling coastal waters from the depths are cold, and temperature readings from submersibles in proximity to Black Smokers continue to show low temperatures except in close proximity to the hydrothermal vents. So, yes, there are submarine heat sources, but I don’t think that there is compelling evidence that it overpowers what is going on at the surface, which is what we monitor.

Reply to  Clyde Spencer
September 7, 2019 1:12 pm

Interesting, I am amazed that most geologists do not want to attribute any extra heat simply coming from below (especially in the NH, especially the past 6 years or so)
yet what is your/ their explanation for the ordinary speed by which our magnetic north pole is shifting, particularly going more northward?
This has no effect on the heat coming from below?
My finding was that Tmin in the SH is going down – here where I live Tmin has dropped by 0.8K over the past 40 years – whereas Tmin in the NH is going up. How do you explain this except for acknowledging the movement of earth’s inner core? Come down two km into a goldmine here, and soon notice the sweat on your face. Then you begin to wonder: how big is this elephant, exactly?

Clyde Spencer
Reply to  henryp
September 7, 2019 6:23 pm

Heat moves through solids as vibrational energy transferred to neighboring molecules — conduction. You would have to demonstrate that non-magnetic silicates, and magnetic metals above their Curie point, can have their vibrations directionalized, or aligned, such that the heat does not propagate isotropically from the hot core.

Just because there is a correlation does not mean that one of the variables is dependent on the other. That gets mentioned frequently here with respect to CO2 and temperature. It is formally called “spurious correlation.” If you can demonstrate some way that heat can be directionalized, then I’ll take notice. Being able to channelize heat with a magnetic field would be immensely useful for, among other things, being able to insulate components from destructive heat.

Jim Ross
Reply to  henryp
September 8, 2019 9:46 am


I don’t have an explanation for the extra warming in the NH, but I think the following observation is important.

If you remove HadCRUT4 from your plot just for clarity, you should see something even more surprising. The reason that HadSST3 NH shows a much higher warming rate than HadSST3 SH is a consequence of excess warming, obviously, but it is only during the summer months. The NH winter months show warming at roughly the same rate as the SH. There are issues with data coverage, but it looks like the divergence (particularly since early 2003) is real. The fact that the anomalies since then are showing major influence of the annual cycle in the NH raises other questions, but the impact is not just in overall temperature trend; it seriously contaminates the short term signal due to ENSO as well.

I will try to link to my plots (with apologies for the ads on Postimage):

All monthly data since 1990 (SH and NH):
comment image
All SH data but only March values for NH:
comment image
All SH data but only August values for NH:
comment image

Reply to  Jim Ross
September 8, 2019 11:24 am


Thanks very much for the reply.
But this leaves me even more puzzled.
I thought it was earth. You say it is the sun.
What happens if you check January (SH – summer)

Reply to  Jim Ross
September 8, 2019 12:17 pm

Jim. Thx so much for ur comment. But nog it leaves me more puzzled. What happens if you look at SH January results to SH or NH?

Jim Ross
Reply to  henryp
September 8, 2019 12:51 pm

Puzzled is good! At the end of the day that is surely why most of us are here. We want to learn and we know that the view that the “science is settled” is rubbish.

I did not say specifically that “it is the sun”, but your inference is valid … what else could it be? I am a bit reluctant to jump in too deep yet as I have been looking at HadSST3 for a while now and there are a lot of “issues” that make me uncomfortable. I will take a closer look at SH January tomorrow, but a quick look suggests that it is more strongly influenced by ENSO, which seems to “peak” (high or low) around January. Look at 2008 for the response to La Niña, for example.

Jim Ross
Reply to  henryp
September 9, 2019 4:47 am


Hopefully, this is what you were looking for. First, SH all months with SH January only:
comment image

Second, NH all months with SH January only:
comment image

TBH, I don’t see this adding much information but would be interested in any feedback. The NH winter months (whether January or March) show the same sort of growth rates as the SH. The January SH data is strongly reflective of ENSO, a well-known correlation.

What is not shown here is the fact that the NH divergence caused by the increased impact of the seasonal cycle beyond that captured in the base period (1961-1990) only dominates from 50/60N northwards. This is exactly in line with the plots shown by Willis here:

He shows both satellite data and HadCRUT, the latter being a combination of HadSST and land data. At least one question must be to what extent is land temperature data driving the increased warming trend in higher latitudes versus the impact of the sea surface data. I believe that the approach of disaggregating global trends into land/sea data, seasons and latitude bands is hugely important in trying to understand the process(es) behind the warming.

Reply to  Jim Ross
September 9, 2019 8:38 am

Hi Jim

Very interesting!!
At this stage one has to realize that there are 4000 stations sampled in the NH and only 400 in the SH. Never mind the fact that in this 4400 sample there obviously is no balance in latitude, like I had in my sample;
I also suspect that there are very few stations sampled at the high latitude in the SH (or should I say: very low latitude; very confusing\ with the + and -)
Assuming my results are correct, (I showed them somewhere earlier up the thread), we are currently already in a period of global cooling. If so, it follows that as the temperature differential between the poles and equator grows larger due to the cooling from the top, very likely something will also change on earth. Predictably, there would be a small (?) shift of cloud formation and precipitation, more towards the equator, on average. At the equator insolation is 684 W/m2 whereas on average it is 342 W/m2. So, if there are more clouds & rain in and around the equator, this will amplify the cooling effect due to less direct natural insolation of earth (clouds deflect a lot of radiation). Furthermore, in a cooling world there is more likely less moisture in the air, but even assuming equal amounts of water vapour available in the air, a lesser amount of clouds and precipitation will be available for spreading to higher latitudes. So, a natural consequence of global cooling is that at the higher latitudes it will become cooler in winter and warmer and drier in summers.

Hence, somehow your results do make sense to me now. In summer there will be more sun hours at the [higher] latitudes and because we have many more stations recording at the higher lats in the NH it explains the results you are getting now….

You get it it, too?

Again, I say, that all of this will not free us of the natural climate change, i.e big droughts coming up, at the high lats, just about now… (Click on my name to read my report on that)

Jim Ross
Reply to  Jim Ross
September 10, 2019 3:49 am


Your reference to the distribution of stations between the northern and southern hemispheres indicates that you are discussing land-based observations. Sea surface temperature data do not have quite the same distribution problem, but they certainly do suffer from a number of sources of potential bias.

Reply to  Jim Ross
September 10, 2019 4:38 am

Sorry Jim. Yes I was still referring to HadCrut4 which, in any case, was showing an extraordinary correlation with the NH SST, albeit not exactly in tandem. I thought the heat might come from below but now you have convinced me otherwise.
It is just such a pity that nobody realizes that all their estimates of energy coming in are wrong.


HadCrut4 is wrong because the stations sampled are not balanced to zero latitude & chosen 70% @sea and 30% @in-land. The sats are probably also wrong. I think they have underestimated what is coming down from the sun at the moment and what is busy deteriorating their instruments. Obviously the lower the solar polar magnetic field strengths, the more of the most energetic particles are coming lose from the sun. Earth is defending us from these particles by forming ozone, N-oxides and peroxides. In its turn, this changes the atmosphere, TOA. And hence it is globally cooling.

It is all one big paradox, is it not? The sun gets hotter and we are getting cooler. But that is the reason why we exist…..the earth’s atmosphere is our first defense.

Hence, do not go to Mars before you have created some kind of atmosphere.

Jim Ross
Reply to  Jim Ross
September 10, 2019 7:07 am

No problem, Henry.

I have only investigated HadSST3 in any detail and I have no idea how they merge the land and sea surface data to get HadCRUT4. When I get some time …

Anyway, thanks for your feedback.

michael hart
September 7, 2019 10:29 am

“It appears that something subtle happened about 30 years ago that isn’t readily apparent in temperature (or temperature-anomaly) data alone. [..]
The question is, “What is causing the changes around 1990, and is it of any climatological significance?” One possible explanation is that the apparent change in temperature relationships is somehow an artifact of processing; however, I’m not familiar enough with the details of the BEST processing methodology to speculate just how this might occur.”

Or something more fundamentally physically related to the collection and selection of the data to be processed.

Clyde Spencer
Reply to  michael hart
September 7, 2019 10:56 am

There are certainly many issues with climatological data, most of which have been addressed here on WUWT. However, the person most acquainted with the BEST data isn’t defending it. In his typical manner of ‘drive-by’ criticisms, he drops in to make an unsupported claim of “Wrong,” and then disappears into the ether.

My coefficient of variation plots show an abrupt offset in the Tmax and Tmin around 1908 that I strongly suspect is evidence of a bad splice in data sets. Not a peep out of the person who should be quickest to notice that. Might there be other more subtle errors in the data processing? The science can only be as good as the data available to work with.

Ulric Lyons
September 8, 2019 10:32 am

1934 had the same solar driver type as the heatwaves of 1948-49, 1976, 2003, and 2017-18.

September 10, 2019 5:04 am

Note that as Tmax goes down, Ozone (& others) goes up.
1990? 1995?

comment image

Reply to  HenryP
September 10, 2019 8:10 am

@ Jim Ross
thanks for a very illuminating discussion.

God bless you all.

September 12, 2019 5:15 pm

Hi Clyde
I was impressed by your post.
I downloaded the data sets and did some initial data exploration in Python.
If you are interested, you can find my musings at

Clyde Spencer
Reply to  Finn McCool
September 12, 2019 9:14 pm


Thank you for sharing. It was interesting to see what you have done. Throughout my career I have programmed in Fortran, several dialects of Basic, Lisp, and Turtle Graphics. However, once I started using spreadsheets, I just walked away from the other languages. I’ve thought about learning Python, but I’m reluctant to invest the time. I don’t have many years left to get proficient.

The K-Means clustering was interesting in that it suggests there are distinct periods of time in which the temperatures are different from the other periods.

You asked some questions, which I’ll try to answer quickly.
1) For the more aggregated sets, Mosher averages (arithmetic mean) the Tmax and Tmin daily values for monthly and longer periods of time. However, the base period is actually a mid-range value of Tmax and Tmin, not a true mean; the mid-range values are then averaged to create a 30-year average. The various anomalies available in other BEST data sets are the difference between the baseline average mid-range values and the monthly (or annual) Tmax and Tmin.
2) There are historical data sets going back well before 1880. I think that Mosher’s claim to fame is finding and splicing these data sets together. He shows much larger uncertainties for the early data. However, I don’t know how he determines the uncertainties.
3) The temperatures have been converted to ‘anomalies’ for purposes of interpolation and infilling of missing data. It is one of the more controversial practices of climatology.
4) I don’t know the reason for the change in slope around 2000. The increase in CO2 has been a fairly smooth exponential growth, so jumps must be attributed to something else. This is one of the issues that I was exploring. If CO2 is the ‘control knob’ on temperature, how does one explain abrupt changes?

As to the confidence interval for 1883, the best I can suggest for you is to look at one of the monthly data sets to at least get a feeling for the magnitude of uncertainties he is assigning. Some of the data sets have been plotted, saving you the trouble of doing it yourself.

Clyde Spencer
Reply to  Clyde Spencer
September 15, 2019 8:35 am

You asked about the calculation of uncertainties in the BEST data sets. Take a look at this:

Reply to  Clyde Spencer
September 16, 2019 5:12 pm

Hi Clyde,
Thanks for that. I haven’t had time to go through the paper in detail, but using KNA to analyse the data is an interesting concept. I first used it in the early 90’s in geological modelling to create block models of coal seams. I preferred Delauney triangulation mainly because coal is fairly straight forward to model. In highly faulted areas, kriging was slightly better in calculating coal to overburden rations for mine planning.
In the real world, confidence intervals can be highly misleading and misused. I tried to explain to my son, an AGW believer, that, although I have 95% confidence that some bloke at the bar is between 45 and 55, I have only 5% confidence that he is 51.
I also noticed a paper in WUWT today which is relevant to your reply above.
As an aside, I understand your use of spreadsheets. Excel and the likes of Tableau and Power BI are wonderful tools. But..
It is more easy to make mistakes in a spreadsheet.
Sometimes impossible to find the mistakes.
Calculations are horrendously tedious to construct.
Calculations can be obfuscated by using undefined variables.
Difficult to run ‘while’ loops without adhoc Basic code.
Calculations and results are hard to critique.
Limited data analysis methods.
At the end it boils down to what you are comfortable with. Try Jamovi. It’s a spreadsheet style interface to R.
Many thanks for the link, Clyde.

%d bloggers like this:
Verified by MonsterInsights