UAH Global Temperature Update for June 2021: -0.01 deg. C

From Dr. Roy Spencer’s Global Warming Blog

July 2nd, 2021 by Roy W. Spencer, Ph. D.

The Version 6.0 global average lower tropospheric temperature (LT) anomaly for June, 2021 was -0.01 deg. C, down from the May, 2021 value of +0.08 deg. C.

REMINDER: We have changed the 30-year averaging period from which we compute anomalies to 1991-2020, from the old period 1981-2010. This change does not affect the temperature trends.

The linear warming trend since January, 1979 remains at +0.14 C/decade (+0.12 C/decade over the global-averaged oceans, and +0.18 C/decade over global-averaged land).

Various regional LT departures from the 30-year (1991-2020) average for the last 18 months are:

YEAR MO GLOBE NHEM. SHEM. TROPIC USA48 ARCTIC AUST 
2020 01 0.42 0.44 0.40 0.52 0.57 -0.22 0.41
2020 02 0.59 0.74 0.45 0.63 0.17 -0.27 0.20
2020 03 0.35 0.42 0.27 0.53 0.81 -0.96 -0.04
2020 04 0.26 0.26 0.25 0.35 -0.70 0.63 0.78
2020 05 0.42 0.43 0.41 0.53 0.07 0.83 -0.20
2020 06 0.30 0.29 0.30 0.31 0.26 0.54 0.97
2020 07 0.31 0.31 0.31 0.28 0.44 0.27 0.26
2020 08 0.30 0.34 0.26 0.45 0.35 0.30 0.24
2020 09 0.40 0.41 0.39 0.29 0.69 0.24 0.64
2020 10 0.38 0.53 0.22 0.24 0.86 0.95 -0.01
2020 11 0.40 0.52 0.27 0.17 1.45 1.09 1.28
2020 12 0.15 0.08 0.22 -0.07 0.29 0.44 0.13
2021 01 0.12 0.34 -0.09 -0.08 0.36 0.49 -0.52
2021 02 0.20 0.32 0.08 -0.14 -0.65 0.07 -0.27
2021 03 -0.01 0.13 -0.14 -0.29 0.59 -0.78 -0.79
2021 04 -0.05 0.05 -0.15 -0.28 -0.02 0.02 0.29
2021 05 0.08 0.14 0.03 0.06 -0.41 -0.04 0.02
2021 06 -0.01 0.30 -0.32 -0.14 1.44 +0.63 -0.76

Despite the near-normal global average temperatures, the USA Lower 48 temperature anomaly of +1.44 deg. C was the warmest in the 43 year satellite record, ahead of +1.15 deg. C in 1988. In contrast, the Antarctic region (poleward of 60 S latitude) experienced its 2nd coldest June (-1.25 deg. C), behind -1.34 deg. C in June, 2017.

The full UAH Global Temperature Report, along with the LT global gridpoint anomaly image for June, 2021 should be available within the next few days here.

The global and regional monthly anomalies for the various atmospheric layers we monitor should be available in the next few days at the following locations:

Lower Troposphere: http://vortex.nsstc.uah.edu/data/msu/v6.0/tlt/uahncdc_lt_6.0.txt
Mid-Troposphere: http://vortex.nsstc.uah.edu/data/msu/v6.0/tmt/uahncdc_mt_6.0.txt
Tropopause: http://vortex.nsstc.uah.edu/data/msu/v6.0/ttp/uahncdc_tp_6.0.txt
Lower Stratosphere: http://vortex.nsstc.uah.edu/data/msu/v6.0/tls/uahncdc_ls_6.0.txt

5 17 votes
Article Rating
258 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
July 2, 2021 6:12 pm

It looks like June was a cool month at the surface too. Much like February.

John Tillman
Reply to  Nick Stokes
July 2, 2021 6:57 pm

March, April and June were negative anomalies. May was barely positive. The over five year cooling phase is still very much intact, thanks to La Nina.

Reply to  Nick Stokes
July 2, 2021 9:44 pm

Like I said, Earth is cooling. UAH LT for June 2021 is -0.01C.

And the planet is much bigger than just North America – something the alarmist mainstream media would be astonished to learn:
USA plus 1,44C – be very afraid – you’re all gonna burn,
Australia minus 0,76C – be very afraid – you’re gonna freeze off your billabongs.

How are the crops doing across the Great Plains? Not so good, but it’s still early.

mark from the midwest
Reply to  ALLAN MACRAE
July 3, 2021 3:36 am

Crops across the plains states are doing fine. Parts of Kansas, Iowa and Southern Minnesota are dry, but Nebraska, Illinois, Indiana, and Ohio are doing fine. In central Michigan we’re expecting bean yields at 50+ bu/acre, and corn could be well over 225.

Tim Gorman
Reply to  mark from the midwest
July 3, 2021 7:11 am

Same here in eastern Kansas.

Reply to  mark from the midwest
July 3, 2021 10:22 pm

https://wattsupwiththat.com/2021/07/01/claim-nasa-satellites-see-upper-atmosphere-cooling-and-contracting-due-to-climate-change/#comment-3282818

LATEST USDA CROP FIGURES SOUND THE ALARM BELLS
Consistent cold isn’t good for crops. According to latest USDA figures, the situation in 2021 is looking far worse than their original projections foresaw, and the commodity markets are climbing as a result.
Grains have “exploded higher” after the initial release of the USDA Stocks and Acreage Reports this week.
There were some big surprises in the report.
Arlan Suderman, Chief Commodities Economist at of StoneX gives us the rundown:
“The prices literally exploded higher after the reports’ release,” said Suderman — this was in response to a “smaller acreage than expected for corn and soybeans … The corn acreage came in at 92.7 million acres … that was about 1.1 million acres below what the trade expected (which was already low).”
Soybean acres were an even bigger surprise, continued Suderman, which came in at 87.55 million acres with the trade expecting 88.95 million acres.
“Stocks being less than expected for corn, soybeans and wheat (have sent) the markets off to the races,” he said.
Corn stocks are currently estimated to be at 4.11 billion bushels, which is down a whopping 18% when compared to the same time of last year — this is despite a 2% increase in planting acreage on 2020.
Looking forward, Suderman sees roll-on implications for July’s crop reports.
Expect higher prices moving ahead.

Reply to  ALLAN MACRAE
July 6, 2021 6:44 pm

USDA regularly bad mouths crops and predicted yields. Then once the Fall harvest begins and actual yield numbers are available, USDA suddenly brings the predicted yields closer to actual yields, always higher than predicted.

92.7 + 1.1 = 93.8.
1.1 / 93.8 = 1.2%

Looks to be well within estimate error bounds. It’s not like they actually measure every ear of corn in every field or even every field that is planted in corn.

Futures movements are dangerous grounds for ordinary investors.

Last edited 2 months ago by ATheoK
Geoff Sherrington
Reply to  Nick Stokes
July 3, 2021 3:50 am

If only we could find the mechanism. Geoff S

Sara
Reply to  Nick Stokes
July 3, 2021 10:10 am

Well, it IS and has been cool here in my AO, and well to the south of me, and when I say “cool”, I mean below “normal” or average temperatures for March, April, May and June. The few “warm” days we’ve had have consistently not had warm nights.

I cannot emphasize “not had warm nights” enough. Daytime temperatures are normally higher than night-time temps, no matter what the season. It is now July, the normal daytime temp should be above 75F and the normal overnight temp mid to upper 60s and mid-70s, but is not going in that direction.

I do not mind the cooler weather. Plenty of sweaters in my possession. But this also applies to farmlands well to the south, east and west of me, and to the north, along with ‘tourist-attractive” locales. Cooler weather did not bring more rain, either.

When I watch the Iceland live volcano cams, I frequently see the volcano tourists wearing cold weather clothing. Without a reference for their “normal” weather patterns (which I don’t have), they appear to be stuck in late winter weather. I do know that the Icelanders normally have a high-volume tourist trade, and that their summers have been warm enough to swim in hotel pools, but that seems to not be going on this year.

So is it just flakey weather, the effect of high-end volcanism, or a change in the weather patterns that is going on? That’s a question that should be addressed, but unfortunately, it may be a short-term event, maybe a single year or a few years, instead of a long-term event.

Tom Abbott
Reply to  Nick Stokes
July 3, 2021 12:48 pm

Yes, June was cool.

The globe has gone from being 0.6C cooler than the 2016 highpoint two months ago, to being 0.7C cooler than the 2016 highpoint last month.

Gordon A. Dressler
July 2, 2021 6:21 pm

Looks like . . . oh, no! . . . not another pause (continuation of such) developing. There will be the gnashing of teeth and the crying in the streets by the AGW/CAGW alarmists . . . well, perhaps just those, like Joe Biden, who promised to follow the science.

Greg
Reply to  Gordon A. Dressler
July 3, 2021 1:33 am

Don’t worry, he has short term memory problems and a history of not keeping his “promises”.

Bill Everett
Reply to  Gordon A. Dressler
July 3, 2021 7:50 am

Most of the temperature levels since 2004, which seems to be the end of the thirty year warming period that started around 1974 or 1975, have been below that of 2004.

Lance Wallace
July 2, 2021 6:21 pm

Did you miss the +1.45 in November 2020 for the lower 48?

Bellman
Reply to  Lance Wallace
July 2, 2021 6:36 pm

I think he means warmest June anomaly.

Bryan A
Reply to  Lance Wallace
July 2, 2021 7:52 pm

The Dim-O-Crats in the lower 48 are just Hot under the Collar to hang Trump before he can run again

pochas94
Reply to  Bryan A
July 3, 2021 3:40 am

DeSantis should run and pick Trump for his Sec of State.

July 2, 2021 6:22 pm

“the USA Lower 48 temperature anomaly of +1.44 deg. C was the warmest in the 43 year satellite record, ahead of +1.15 deg. C in 1988”

2020 11 @ 1.45 ??? or maybe Roy just means for month of June? Didn’t specify.

July 2, 2021 6:33 pm

More cooling … another fracture in the fraudulent climate consensus … https://www.youtube.com/watch?v=O1-FxwVkQ60

Bellman
July 2, 2021 6:42 pm

Pause now starts March 2015.
Warming rate of over 3°C / century since February 2007.
Warming rate over 4°C / century since October 2010.

Gordon A. Dressler
Reply to  Bellman
July 2, 2021 6:54 pm

Bellman, did you even read the above article?

Clearly stated: “The linear warming trend since January, 1979 remains at +0.14 C/decade (+0.12 C/decade over the global-averaged oceans, and +0.18 C/decade over global-averaged land).”

That is equivalent to, worst case over land, 1.8 °C/century since February 2007 and likewise since October 2010.

You have a lot of explaining to do as to how your stated warming rates are so much higher than what the UAH data actually shows.

bdgwx
Reply to  Gordon A. Dressler
July 2, 2021 7:00 pm

I just did a LINEST in Excel. I get the same figures as Bellman. Note that the 0.18C/decade figure is for the period 1979-present. 2007-present and 2010-present are different periods.

Gordon A. Dressler
Reply to  bdgwx
July 2, 2021 7:16 pm

So, bdgwx, you agree with Bellman that short term intervals (e.g., 2021-2007 = 14 years, and 2021 – 2010 = 11 years) are much better than longer intervals (e.g., 2021- 1979 = 42 years) in establishing overall trendline slopes?

Go figure.

BTW, both NASA and NOAA define “climate” as weather averaged over a specified geographic area for an interval of 30 years or longer.

Bellman
Reply to  Gordon A. Dressler
July 2, 2021 7:24 pm

No. I do not think these are better intervals. I think they are terrible intervals. I think it’s ridiculous to base a trend on 5 or 6 years as Monckton does, and even worse to select the start date to give your desired result. Choosing 11 and 14 year periods is almost as bad, and suffers from the same issues as the Pause.

Maybe I should have made it clear in my post, but I fell into the trap of assuming everyone knows the backstory. Normally I make it clear that my trends are cherry-picks.

Bryan A
Reply to  Bellman
July 2, 2021 8:00 pm

Bellman
July 2, 2021 6:42 pm
Pause now starts March 2015.
Warming rate of over 3°C / century since February 2007.
Warming rate over 4°C / century since October 2010.

And a Warming rate of over 350C / Year since 07:00
Only shows the Shorter the Period, the less accurate the long term trend

Mike
Reply to  Bellman
July 2, 2021 11:47 pm

I think it’s ridiculous to base a trend on 5 or 6 years as Monckton does,”
Not what he’s doing though is it. In fact he’s doing the opposite by saying there is no trend during this 5 or 6 year (or whatever) period.
Now, it’s simply up to you to prove him wrong.

Bellman
Reply to  Mike
July 3, 2021 4:06 am

A trend of zero is still a trend.

Monckton defined the old pause as “the farthest back one can go in the RSS satellite temperature record and still show a sub-zero trend.”, and I see nothing, apart from no change for the definition of the new pause, apart from changing the data set.

We don;t know what the actual trend was over the pause period because of the inherent uncertainty of linear regression on noisy data. For the last 6 years, depending on exactly how you adjust for different factors the uncertainty could be up to ±9 °C / century.

Now, it’s simply up to you to prove him wrong.

Not how this works. Monckton makes a specific claim – there is a pause, he needs to show statistical justification for the claim.

Jim Gorman
Reply to  Bellman
July 3, 2021 5:18 am

We don;t know what the actual trend was over the pause period because of the inherent uncertainty of linear regression on noisy data.

The data is NOT noise. Noise would be extraneous information that is not temperature yet appears to be temperature. In other words, it would be interference to knowing what the actual temperature is. The data you are looking at IS the signal.

What you are seeing is natural variation. Hmmmm, see that word variation, it is what is known in statistics as variance. It is a reason that linear regression of a continuous, varying, time series can not adequately and reliably project what the future holds. It’s what Fourier or wavelet analysis was really designed to do properly.

Bellman
Reply to  Jim Gorman
July 3, 2021 8:01 am

Noise is variation on the trend, not variation from temperature. The trend is a possible correlation between two variables, in this case between time and temperature. If the correlation was perfect the anomaly for each month could be predicted by time. But most months are not exactly predicted by time, the difference is variation. It doesn’t matter at this point what caused the variation, we just need to know it exists and what it looks like. From this we can estimate how much confidence to put in any trend. The variation means we cannot assume any trend is exactly correct because data might just happen to be above or below the trend at different parts of the line. In general the more data points we have and the less variation relative to the trend, the more confidence we can have in the trend.

If we do have an idea what caused some of the variation, we can also try to account for that using a multi-variant regression, but nobody is doing that here. We are just talking about the trend compared with time.

Clyde Spencer
Reply to  Bellman
July 3, 2021 10:08 am

The trend is a possible correlation between two variables, in this case between time and temperature.

When the slope of the regression is effectively zero, there is no correlation. That is, the independent variable has no predictive power. The dependent variable varies randomly about some constant value.

Jim Gorman
Reply to  Bellman
July 4, 2021 5:43 am

Noise IS NOT variation on a trend. Please find a reference and show it that describes noise in that fashion.

If you are trying to show a correlation between time and another variable you are barking up the wrong tree. Time is not a causative factor in the variance of temperature.

Please google how time series analysis is performed and what it requires.

Bellman
Reply to  Jim Gorman
July 4, 2021 2:49 pm

Noise IS NOT variation on a trend. Please find a reference and show it that describes noise in that fashion.

https://www.investopedia.com/terms/n/noise.asp

Time is not a causative factor in the variance of temperature.

Of course it’s not. It’s only a proxy for something else that is changing over time. But if you want to talk about the rate of change it’s reasonable to look for the correlation between time and temperature. If I’m looking at a thermometer in my garden and I say it’s risen 5°C in the last two hours, I’m not saying time caused the warming, I’m just saying how quickly the warming occurred.

Clyde Spencer
Reply to  Bellman
July 3, 2021 10:03 am

A trend of zero is still a trend.

However, it is not a warming trend, which is the concern.

Carlo, Monte
Reply to  Mike
July 3, 2021 6:49 am

Bellman still has a burr under his saddle about Monckton? At this point he’s sounding like a skipping record.

Bellman
Reply to  Carlo, Monte
July 3, 2021 8:02 am

Given all the things he’s called me, is that surprising?

Leo Smith
Reply to  Bellman
July 3, 2021 1:38 am

I think it’s ridiculous to base a trend on 5 or 6 years

BUT if the dominant cause of modern climate change is CO2 increase, then it doesnt matter how short the period is, it should be in lock-step with CO2 increase…what you are saying however is that the signal is so weak and pathetic that we need 30 years or more…to detect it.

You can’t have it both ways. Either the data is so strongly correlated with CO2 that the ‘science is settled’ or its so weak that its not in the least bit alarming.

I see exactly the same bait and switch in nuclear radiation alarmism. The undoubtedly dangerous and cancer inducing Iodine 131, has a half life of a (bit over a) week only!

We then conflate the ‘dangers of radiation’ with ‘radioactive isotopes that last a billion years’. Like natural and depleted uranium? Which already exist in such enormous quantities in the world that we could run the world off nuclear power for 10,000 years without exhausting them?

What all the data suggest is that even in the last 70 years weather effects and natural climate events completely dominate the cliamate, and that if you pick one decade its scary warming whilst another gives you alarming cooling!

The myth is busted! I am sure CO2 has some effect on climate, but it absolutely cannot be the single dominant effect.

Why then does no one dare admit this? Because they have all used it for commercial, political or personal advantage?

The tragedy of the commons: Everybody is making money and careers out of climate change, but everybody’s civilisation is being destroyed by the inffective measures that purport to combat it.

Bellman
Reply to  Leo Smith
July 3, 2021 4:11 am

BUT if the dominant cause of modern climate change is CO2 increase, then it doesnt matter how short the period is, it should be in lock-step with CO2 increase...”

This is complete nonsense. Every month shows random fluctuations that have nothing to do with CO2, on larger scales of a few years there are known causes that effect temperature such as El Niños and La Niñas. All of this affects the short term trend.

Jim Gorman
Reply to  Bellman
July 3, 2021 5:24 am

The fluctuations you are describing are not “random”. They are part of the natural variations that occur on this planet. That they appear “random” to you simply means you, and even most climate scientists, are unaware of the how the different factors combine together to give the earth its own climate. IOW, the science is NOT settled.

Clyde Spencer
Reply to  Bellman
July 3, 2021 10:10 am

… on larger scales of a few years there are known causes that effect temperature such as El Niños and La Niñas.

So, you are acknowledging that CO2 is not the dominant driver in temperature changes!

Bob boder
Reply to  Bellman
July 3, 2021 6:09 pm

5or6 years is not every month

Carlo, Monte
Reply to  Leo Smith
July 3, 2021 6:55 am

Excellent; another way of pointing this out is to simply ask: “where is the hockey stick?”

Jim Gorman
Reply to  Bellman
July 3, 2021 4:53 am

Maybe instead of criticizing others, you could provide a rational explanation of why, over even 5 or 6 years, the temperatures are decreasing in the face of ever increasing CO2. That might even go a long way toward explaining natural variation causing both heating and cooling and why CO2 is a bit player in the overall scheme of things.

Bellman
Reply to  Jim Gorman
July 3, 2021 5:08 am

A rational explanation is that 2016 was a very big El Niño year, and we’ve just had a La Niña.

Bellman
Reply to  Jim Gorman
July 3, 2021 5:19 am

How much warming do you expect CO2 to cause in half a decade? Warming in UAH is around 0.13 – 0.14°C / decade. Most other data sets put it around 0.17 – 0.2. Even if the warming rate was twice that, you would only expect about 0.2°C back ground warming over the pause period. The difference between a strong El Niño and La Niña could be more than twice that.

Jim Gorman
Reply to  Bellman
July 4, 2021 5:53 am

I expect CO2 to be the driving factor in warming at all periods of time if warmists are to be believed. You can’t expect the radiation “heat trapping” effects to change over any time period, however short. The only reason for temps to fall with rising CO2 would be variation in the sun’s incoming radiation or cloud reflections. If that is true, whoops, CO2 isn’t the main driver.

Bellman
Reply to  Jim Gorman
July 4, 2021 8:44 am

Then your expectations will be disappointed. Of course CO2 is not the only reason for a given monthly temperature. The data sets, either surface or satellite are not measuring total heat content, but a limited part of the earth, surface or troposphere. Heat moves between them for a variety of reasons.

But let’s assume that CO2 has no effect on temperature. You still have all your problems about what’s causing the short te fluctuations, but also need to explain what causes the longer term warming.

Jim Gorman
Reply to  Bellman
July 4, 2021 10:09 am

I’m not the one making the assertion that CO2 is the control knob for temperature. The folks who believe in the GHE (Green House Effect) are the ones that need to provide the experimental data that show that.

As I said about fluctuations:

That they appear “random” to you simply means you, and even most climate scientists, are unaware of the how the different factors combine together to give the earth its own climate. IOW, the science is NOT settled.

I meant what I said. The science is not settled. Can I provide the proof of how the varying factors interrelate on a time basis to change temperature? Heck no! That doesn’t mean I can’t criticize and point out errors to those that think they have the bull by the horns and laying on the ground.

Bellman
Reply to  Jim Gorman
July 4, 2021 2:52 pm

CO2 is the principle control knob does not mean that every short term fluctuation in temperature is caused by the knob.

mario lento
Reply to  Gordon A. Dressler
July 2, 2021 8:04 pm

I got a warming line of +50F, starting 5am and going through 5pm two weeks ago. That is a gazillion degrees by 2100 if I extrapolate the new trend. We’re going to die!

Loydo
Reply to  mario lento
July 2, 2021 8:47 pm

You’ll be ok mario, at least in the short term; you’re talking about weather, they’re talking about climate.

Mike
Reply to  Loydo
July 2, 2021 11:10 pm

No. Everyone is talking about weather. The sat data above just shows (one aspect of) the weather. Nothing to do with climate.

Last edited 2 months ago by Mike
Chaswarnertoo
Reply to  Loydo
July 3, 2021 2:45 am

I know you’re fick, loydo, but it was a joke.

Graemethecat
Reply to  Chaswarnertoo
July 3, 2021 6:58 am

Appreciation of irony requires a certain level of intelligence, which is clearly beyond Loydo.

John Hultquist
Reply to  Gordon A. Dressler
July 2, 2021 8:38 pm

This is not a definition of climate.
“Climate normals” do not define climate — they produce a generational comparison of weather data — exactly what the members of the 1935 meeting wanted.

mario lento
Reply to  John Hultquist
July 2, 2021 10:24 pm

Lloydo does not understand climate, weather or sarcasm.

Loydo
Reply to  mario lento
July 3, 2021 1:20 am

Whoa, I thought you were being serious. /sarc.

Gordon A. Dressler
Reply to  John Hultquist
July 3, 2021 8:38 am

In reference to my earlier comment “BTW, both NASA and NOAA define ‘climate’ as weather averaged over a specified geographic area for an interval of 30 years or longer”, John Hultquist replied, apparently without any research on his part: “This is not the definition of climate.”

Your honor, I would like to present to the court Exhibit A:
Climate is the average weather conditions in a place over 30 years or more.”
—source https://climatekids.nasa.gov/menu/weather-and-climate/

and Exhibit B:
Scientists traditionally define a Climate Normal as an average over a recent 30-year period . . . Why 30 years? Close to a century ago, the International Meteorological Organization—now known as the World Metrological Organization—instructed member nations to calculate Climate Normals using 30-year periods, beginning with 1901–1930. Also, a general rule in statistics says that you need at least 30 numbers to get a reliable estimate of their mean or average. So, our scientists have traditionally defined Normals as averages over 30 years simply because that is the accepted convention—not because a 30-year average is the only logical or “right” way to define a Climate Normal.”
— source https://www.ncdc.noaa.gov/news/defining-climate-normals-new-ways

I rest my case.

John Hultquist
Reply to  Gordon A. Dressler
July 3, 2021 9:06 am

 Consider 3 places:
Central Washington State** – mostly dry, hot summers, cold winters
Singapore – no distinctive seasons, uniform temperature pressure, high humidity, and abundant rainfall;
Bamako – hot, rainy during the high sun season, otherwise dry

Now consider you have one foot in a bucket of very hot water, the other in a bucket of ice water:
On average you will be fine.

“Climate Normals” are human defined numbers. They are not climate.

If, using the above 3 places, average temperature goes up (or down) by 1.5 Celsius degrees – – who would notice?

** Remove the Cascade Mountains and the climate will change.

Gordon A. Dressler
Reply to  John Hultquist
July 3, 2021 9:16 am

?

pHil R
Reply to  John Hultquist
July 3, 2021 6:05 pm

John,

If, using the above 3 places, average temperature goes up (or down) by 1.5 Celsius degrees – – who would notice?

Try growing a palm tree in Alaska or a Redwood in the Sahara. By your analogy, the palm tree should grow if someone plugs in a heat lamp. Plants would notice. long-term averages are are meaningful and useful.

Dave Fair
Reply to  Gordon A. Dressler
July 3, 2021 12:46 pm

Agreed, Gordon. I would point out, however, we are experiencing cyclic variations in climate metrics on decadal timescales, one of the major ones is of approximately 60 years duration. I wish it were possible to get reliable climate data back about 120 years. Oh, well … by the late 2020s we might know better the actual ECS of CO2.

Reply to  bdgwx
July 6, 2021 6:48 pm

Using a magic weather is actually climate wand?

All you are accomplishing is proving that tracking warming by decades is a false representation of climate.

Bellman
Reply to  Gordon A. Dressler
July 2, 2021 7:05 pm

You obviously have never seen Lord Monckton’s articles cherry-picking a start date to show a negative trend. My comment was one of a series parodying it.

My stated warming rates are obtained by choosing a start date that is far back as I can go, that will give me the desired trend from that date to present.

MarkW
Reply to  Bellman
July 2, 2021 7:27 pm

How the heck is picking the present as the start date, cherry picking.
The end date of the trend is determined by the math, and the math alone.

Bellman
Reply to  MarkW
July 2, 2021 7:59 pm

I’m not picking the present as the start date. I’m picking either of the dates I specified as the start date. The trend is calculated from that date to the present, which I consider the current end date. I’m sorry if the order of chronological time confuses you.

Bellman
Reply to  Bellman
July 2, 2021 8:01 pm

And my choice of start dates are chosen by maths and maths alone.

Last edited 2 months ago by Bellman
pHil R
Reply to  Bellman
July 2, 2021 8:26 pm

You’re guilty of reification. Math(s) does not and cannot choose anything.

Reply to  pHil R
July 2, 2021 9:06 pm
Bellman
Reply to  pHil R
July 3, 2021 4:12 am

Read the post I was responding to. It says of Monckton’s pause “The end date of the trend is determined by the math, and the math alone.”.

MarkW
Reply to  Bellman
July 3, 2021 7:29 am

The post in question was in response to your lie that Monckton picked the start date.

Dave Fair
Reply to  Bellman
July 3, 2021 12:50 pm

A little intelligence is required to understand that the ‘end date’ is the chronological beginning point of the series looking back from the present date.

Peter K
Reply to  Bellman
July 3, 2021 4:07 pm

And whatever keeps the dream alive.

Bellman
Reply to  Peter K
July 3, 2021 4:25 pm

What dream. I don;t know how many times I have to say this – when I point out that you can choose a start date to show a warming rate of over 4° / century, I am not saying this to prove that warming is accelerating or that this is a meaningful rate of warming. I am demonstrating the absurdity of cherry-picking a date that gives you such a warming rate. It’s entirely my argument that, as Monckton says, it’s a statistical abuse.

MarkW
Reply to  Bellman
July 3, 2021 7:29 am

In other words, you are admitting you are lying about what Monckton is doing.

The math starts at today and works backwards.
There is zero cherry picking going on.
You are so desperate to try and disprove the fact that there is no weather crisis that you are willing to throw away what little intellectual integrity you may have once had.

So sad.

Bellman
Reply to  MarkW
July 3, 2021 8:28 am

Here’s how I do it. I get the time series of monthly anomalies. I feed them into an r function whish generates the trend for each month to the present. I look through all the results to find the earliest month that gives a negative trend and claim this will be Monckton’s new pause start date.

I do the same thing for my other trends, for example looking for the earliest date that will give me trend greater than a specific rate.

Whether you consider this as being me or the math choosing the start date is semantics. There’s only one date that meets the criteria. I could quite easily add a function to find the start date for me.

When people here insist that Monckton works backwards from the current date I’m not sure what they think he’s doing. I suspect that some people think he looks back from the current until he finds the first month that gives him a negative trend, but that’s just wrong. If that was the case his pauses would be much shorter. I’ve also seen it suggested that he stops as soon as he finds a positive trend and counts the start of the pause as being the following month, which is just as wrong.

The fact is, the only way Monckton gets the start date he does is to go back looking at every possible start date, and selecting the earliest month giving a negative trend. It doesn’t matter if traces backwards from the current date, he still has to look at every month. Doing it backwards is just inefficient. If he started with the earliest month he could stop as soon as he found a month that gives a negative trend.

If you think Monckton does something different, spell out what the algorithm is, and test it before he published his next article.

Bob boder
Reply to  Bellman
July 3, 2021 6:14 pm

So you cherry pick, got it

Bellman
Reply to  Bob boder
July 3, 2021 6:40 pm

Yes, now how do you think Lord Monckton does it?

Clyde Spencer
Reply to  Bellman
July 3, 2021 10:23 am

You fail to understand the difference between characterizing past behavior versus attempting to discern a change in that past behavior.

It is not unlike being engaged in an aerial dogfight. The pilot of the pursuit plane doesn’t know where the enemy will be minutes in the future, but needs to estimate where to aim and fire in the next few seconds. The best estimate for the position of the target is an extrapolation of the recent past trajectory. Extrapolations are often wrong. However, that is all the pilot has to go on. Sometimes, the extrapolation is correct. That is what justifies the process.

Bob boder
Reply to  Bellman
July 3, 2021 6:12 pm

Bellman
Your just full of nonsense

Clyde Spencer
Reply to  Bob boder
July 4, 2021 11:12 am

Bats in his belfry.

Reply to  MarkW
July 2, 2021 8:14 pm

MarkW, you cannot pick the present as the start date because time doesn’t run backwards. The “cherry pick” is using the present as the END date.

Reply to  nicholas tesdorf
July 2, 2021 8:20 pm

Furthermore MarkW if the current month is abnormally cool, it alters the start date of the trend. When this happens then Monckton is showing us how a change in the present causes a change in the past.

Reply to  nicholas tesdorf
July 2, 2021 8:24 pm

And lastly MarkW, when Monckton determines that his “pause” started X months ago, and then the following month he finds the length of his pause is still X months long and has not changed, then his determination of the start date from the prior iteration was WRONG. In this case his procedure doesn’t consistently determine the true start date making it worthless.

Last edited 2 months ago by nicholas tesdorf
Mike
Reply to  nicholas tesdorf
July 2, 2021 11:23 pm

Gobbledigook.
The fact is it has not warmed (overall) for at least 20 years according to the above graph and probably not (overall) for 100 years.

co2 journal of geophysical research..GIF
Last edited 2 months ago by Mike
Coeur de Lion
Reply to  Mike
July 2, 2021 11:53 pm

Talk of warming rates? The globe is back to where it was in 1987 or have I missed something? That’s climate’ not weather ’

TheFinalNail
Reply to  Coeur de Lion
July 3, 2021 1:52 am

Not quite. As you can see, we were also in negative anomaly territory as recently as 2018. That doesn’t mean that all the accrued warming has just vanished. Each monthly anomaly is just one datum point in the series as a whole and it’s the series as a whole from which the linear trend is derived. Or you can select any consecutive period within the series and obtain the trend for that (a la Lord M).

For instance, you mention that the current monthly anomaly in UAH is the same as some months in 1987, and that’s true. But if you use a spreadsheet to calculate the linear trend from 1987 (Jan), you will see that there has been warming at a rate of +0.13 C/decade; or a total warming of +0.45 C.

To see the negative effect of the -0.01 C June anomaly on the total warming value in UAH since 1987, you have to go to 3 decimal places. Up to May 2021 total warming since 1987 stood at +0.452 C. As of the June update it now stands at +0.450 C. In other words, the negative June 2021 anomaly has reduced the total warming seen in UAH since 1987 by two one-thousandths of a degree C.

Dave Fair
Reply to  TheFinalNail
July 3, 2021 1:04 pm

The UAH6 estimates start at the beginning of a cyclical upswing in global temperatures. It continues through to a peak and the beginnings of a reasonably expected downswing of the identified cycle(s). Color me unimpressed with a 0.14 C/decade trend over a cyclical upswing. Everything else is nitpicking B.S.

Bob boder
Reply to  Coeur de Lion
July 3, 2021 6:16 pm

And where it was a 1000 years ago for that matter.

Anthony Banton
Reply to  Mike
July 3, 2021 1:31 am

How about we update that by the 19 years that it is missing …..
comment image

Bruce Cobb
Reply to  nicholas tesdorf
July 3, 2021 6:38 am

You obviously don’t have a clue as to what a pause is, let alone what it means. Get a clue.

Carlo, Monte
Reply to  nicholas tesdorf
July 3, 2021 6:59 am

So tell me, exactly where is the CO2-caused hockey stick?

Trying to Play Nice
Reply to  nicholas tesdorf
July 3, 2021 5:31 am

Sort of like how the alarmists recalculate past temperatures every month to keep making them lower.

Clyde Spencer
Reply to  nicholas tesdorf
July 5, 2021 11:24 am

Will the real nicholas tesdorf please stand up.

TimTheToolMan
Reply to  nicholas tesdorf
July 3, 2021 6:13 am

MarkW, you cannot pick the present as the start date because time doesn’t run backwards. The “cherry pick” is using the present as the END date.

Absolutely you can. “Today” is the only non-cherry picked and defendable date. Looking backwards is perfectly valid.

Furthermore MarkW if the current month is abnormally cool, it alters the start date of the trend. 

Only if the duration is short. If there is an actual warming trend then you’re right and the duration looking backwards will only be short before the trend dominates. On the other hand if there is no trend for a long time, then the current month doesn’t have such a large impact whether it be a warm or cold month.

The important thing is how long one can look backwards before seeing the trend. If its a few years then its not so convincing. If its 17 years then its very important considering GCMs dont produce warming or cooling trends for that long without a forcing as per Santer’s paper.

In this case his procedure doesn’t consistently determine the true start date making it worthless.

No. Claiming there IS a start date is worthless.

Clyde Spencer
Reply to  TimTheToolMan
July 3, 2021 10:31 am

Absolutely you can. “Today” is the only non-cherry picked and defendable date. Looking backwards is perfectly valid.

Looking backwards in time is like asking the question “When did the first COVID-19 case occur?” Alternatively, it is perfectly valid to ask, “How many years has it been since Pearl Harbor was attacked?”

Bellman
Reply to  Clyde Spencer
July 3, 2021 12:37 pm

Does the date of Pearl Harbor change from month to month?

Clyde Spencer
Reply to  Bellman
July 3, 2021 2:54 pm

Obviously not! I didn’t realize you were that dense. However, the length of time does change from month to month. Probably the COVID situation is a better analogy because they keep finding cases that are older.

Bellman
Reply to  Clyde Spencer
July 3, 2021 4:17 pm

However, the length of time does change from month to month.

Yes it changes by precisely one month every month. The pause length has been changing by several months each month.

Probably the COVID situation is a better analogy because they keep finding cases that are older.

Nobody’s finding an earlier case of the pause. The temperatures are not changing in 2015.

MarkW
Reply to  nicholas tesdorf
July 3, 2021 7:30 am

Time does not run backwards, but math can.

Gordon A. Dressler
Reply to  nicholas tesdorf
July 3, 2021 8:10 am

“The “cherry pick” is using the present as the END date.”

Not so, if you are discussing a curve fit (e.g., linear regression) to a set of data that includes the “present” data point(s). There is NO requirement, or even expectation, that a linear regression line pass through either the first data point in a time series or the last (i.e., most recent) point in a time series . . . such curve fits rarely do either.

Consequently, there is no such thing as “cherry picking” via inclusion of the end point, with the understanding that the data set contains enough samples to be sufficiently representative of the physical/mathematical process being analyzed.

How does one determine if a given sample of data points is representative of the whole set (assuming one does want to do a regression analysis on the entire data set)? A simple direct approach is to first sample, say, every third data point to get Sample Set #1 and then to sample, say, every fifth data point to get Sample Set #2. If regression analyses performed on Sample Set #1 and Sample Set #2 yield essentially the same results, then either is representative of the whole.

Likewise, it is easy enough to determine the effect of either a given start point value or a given end point value in a linear regression analysis . . . just eliminate either data point from sample set being analyzed and re-run the regressions analysis to see if the results (linear equation and r^2 value) change significantly. Assuming the sample set being analyzed has a fairly large amount of data points (say, over 20) and that neither start point nor end point are anomalous with respect to values of the other data points in the set, eliminating either point will not significantly change the slope and y-intercept of the linear fit.

Jim Gorman
Reply to  Gordon A. Dressler
July 4, 2021 6:22 am

The worst assumption is that a linear regression will give an adequate representation of a trend to begin with. Temperature is a continuous function made up of various components with varying phase and amplitudes.

These create a function that simply can not be adequately represented by a simple linear function and probably not even by a single complex function. The proof is that climate models are fubar.

Breaking the temperature trend into shorter and shorter periods is very similar to differentiation. It does indicate various discontinuities. Discontinuities tell me that there are phase and amplitude crossovers occurring that have not been adequately addressed by mathematics at this time.

Gordon A. Dressler
Reply to  Jim Gorman
July 4, 2021 1:05 pm

Jim stated: “The worst assumption is that a linear regression will give an adequate representation of a trend to begin with.”

I just observe that linear regression analyses (linear curve fitting) has served scientist, engineers, biologists, medical researchers, etc.—thereby all of humanity—extraordinary well over the course of human civilization since first principles of statistics were used.

Almost everyone knowledgeable about linear regression knows from the get-go that the resulting linear equation is only an approximation of how a given data set is actually varying. That is, being an approximation in no way diminishes its practical usefulness.

But I do welcome your suggestion for better method of estimating how any arbitrary set of one-dimensional measurements may vary over time. Do you yourself assume exponential, logarithmic, trigonometric, polynomial, infinite-series, etc., “curve-fits” would be better . . . and if so, why?

Jim Gorman
Reply to  Gordon A. Dressler
July 5, 2021 8:56 am

Gordon,

There are so many climate variables with periodic variance to consider ranging from the sun to ocean currents to clouds that it is probably impossible to adequately assume any one curve fit will work to forecast the future.

I spent 30 years with a major Bell Telephone Company trying to forecast and provision common equipment in central offices, pbx’s, people (operators), and equipment/people budgets. I am very familiar with forecasts for one, two, and 5 years plans. I quickly learned that once you leave the present and begin trudging into the future that plans never work out.

Unless you can adequately define the actions of each and every climate variable, which is impossible, trying to address what to use to to generate a forecast is also impossible. A regression of any type on past data only allows pretty curve to be drawn but its usefulness for telling you what the future will bring is extremely doubtful.

Gordon A. Dressler
Reply to  Jim Gorman
July 5, 2021 9:38 am

“A regression of any type on past data only allows pretty curve to be drawn but its usefulness for telling you what the future will bring is extremely doubtful.”

Hmmm . . . an extremely high number of professional stock/equity analysts and financial advisors, as well as millions of savy individual investors, would strongly disagree with that statement. Many trillions of dollars are invested on the basis of past trends being indicative of the future to some limited duration (a week, a month, a year, whatever) while at the same time cautioning (as a CYA statement) that “past performance is no guarantee of future results)”.

But then, maybe you’ve never invested in any equities of any sort.

Last edited 2 months ago by Gordon A. Dressler
Clyde Spencer
Reply to  Gordon A. Dressler
July 7, 2021 8:02 am

The methodology of stock pickers works — until it doesn’t. It may just be luck. In any event, investors know that their extrapolations don’t always work, and hedges, such as putting in stop-loss orders were developed to compensate for their imperfect crystal balls.

One needs to consider such nuances as autocorrelation and ‘runs’ as observed in gambling.

Bob boder
Reply to  nicholas tesdorf
July 3, 2021 6:15 pm

Nicholas, MarkW isn’t picking any thing lordM is and looking at the present compared to the past is not Cherry picking.

John Howe
Reply to  Bellman
July 3, 2021 12:13 am

The alarmists on the ABC use 1970 and 1910 as the start dates for their propaganda.

Bellman
Reply to  John Howe
July 3, 2021 4:26 am

No idea what ABC says, but one obvious question is how much of a difference do you get if you start in 1969 or 1971, rather than 1970? With the short term pause you only have to change the start date by a few months to gets a very different trend.

Using GISS for various start dates I get
1968: +0.187°C / decade
1969: +0.187°C / decade
1970: +0.190°C / decade
1971: +0.192°C / decade
1972: +0.191°C / decade

Very little difference as to what year you pick.

Now contrast this with the different start dates for UAH

March 2013: +0.298°C / decade
March 2014: +0.168°C / decade
March 2015: -0.023°C / decade
March 2016: -0.157°C / decade
March 2017: +0.109°C / decade

Last edited 2 months ago by Bellman
Gordon A. Dressler
Reply to  Bellman
July 3, 2021 11:15 am

Bellman,

And the specific reason you chose to go back to the late 1960’s thru early 1970’s for GISS data, but instead to go back only to the mid-2010’s for UAH data is . . . ???

It’s an apples-to-oranges comparison, and therefore the comparison of temperature rise rates that you posted is meaningless.

Moreover, most people skilled in math will know one is susceptible to spurious statistical error the fewer the number of data points included in the total data set (in this case, the fewer the number of years) being analyzed.

Bellman
Reply to  Gordon A. Dressler
July 3, 2021 12:43 pm

And the specific reason you chose to go back to the late 1960’s thru early 1970’s for GISS data, but instead to go back only to the mid-2010’s for UAH data is . . . ???

Reading the comment I’m responding to would help.

Someone suggested ABC were cherry-picking by starting in 1970, and implied that was as much of a cherry-pick as Monckton starting in the mid 2010s. I showed the difference between the two. It is an apples to oranges comparison, that was the point.

Moreover, most people skilled in math will know one is susceptible to spurious statistical error the fewer the number of data points included in the total data set (in this case, the fewer the number of years) being analyzed.

Yes. That’s what I’m saying.

clarence.t
Reply to  Bellman
July 3, 2021 12:50 am

Seems that Bellman doesn’t understand the methodology.

It’s pointless trying to explain to someone who’s ideology won’t let him understand.

Bellman
Reply to  clarence.t
July 3, 2021 3:54 am

Lets see what Monckton says. If he doesn’t think the new pause now starts in March 2015, then we can agree I don’t understand his method. But so far I’ve predicted every pause start date.

Loydo
Reply to  Bellman
July 3, 2021 5:37 am

I’d say he’s reading this with dismay.

Carlo, Monte
Reply to  Loydo
July 3, 2021 7:02 am

Really? Are you a psychic remote viewing mind reader?

Gordon A. Dressler
Reply to  Bellman
July 3, 2021 11:17 am

“But so far I’ve predicted every pause start date.”

Ummm . . . got any objective evidence to support that statement?

Gordon A. Dressler
Reply to  Bellman
July 3, 2021 7:18 pm

Sorry, Bellman, I don’t do random.

TheFinalNail
Reply to  Gordon A. Dressler
July 4, 2021 3:33 am

Bellman also called this one right. Apparently he dosen’t do random either.

Gordon A. Dressler
Reply to  TheFinalNail
July 5, 2021 9:53 am

TFN,
Since the current pause is stated to have begun 6 years and 10 months ago based on the UAH data presented graphically in the above article, for Bellman to have “called this one right” (your words) I would be looking for objective evidence that he published such a prediction earlier than 6 years and 10 months ago.

Bellman only offered up, as evidence of his claim, four “random comments” (his words) made during 2021 in his post above.

So, where’s your supporting evidence for Bellman calling the start of the current pause correctly earlier than 6 years 10 months ago?

Or is you post itself just a random comment?

Bellman
Reply to  Gordon A. Dressler
July 5, 2021 10:36 am

Whatever Monckton says the current pause did not start 6 years and 10 months ago. It started in March 2015, just as Monckton says, he simple can’t count how many months that was ago.

But I think you are misunderstanding how this works. When Monckton says after the release of UAH each month, the current pause started pon a specific month, that is just the month the pause starts for this month. Next month it may start on a different date. I’m not claiming that I predicted a pause back in March 2015, just that knowing what the data was up to June 2021 I can now predict in what month Monckton will now claim the pause started. When I know Julies figures I can predict what Monckton will say the new start of the pause was.

My only reason for doing these “predictions” is to demonstrate I do understand Monckton’s method, after some here vociferously insisted I had no idea how Monckton calculated the changing pause start date. I’ve wanted to compare these predictions with those of people claiming to understand the method better than me, especially those who insist on the importance of working backwards, but so far no-one has taking the challenge.

Bellman
Reply to  Gordon A. Dressler
July 5, 2021 11:04 am

Predicting the Pause start date – Part 1

Monckton’s first announcement of the new pause was in January 2021, when he said it started in September 2015. Though he wasn’t certain it actually was a pause at the time. Searching my name may give you some idea of why I wanted to prove I understood Monckton’s method.

When the next UAH update was published at the start of February 2021, I posted this comment saying “Lord Monckton’s pause now starts a month earlier, in August 2015, which is also the last month of the old pause.” A couple of hours later Monckton posted his second new pause update saying “The least-squares trend on the latest UAH data show no global warming for 5 years 6 months from August 2015 to January 2021 inclusive”

Bellman
Reply to  Bellman
July 5, 2021 11:14 am

Part 2

Looks like I forgot to post my prediction for the next update, but when the March update was announced I said “The Monckton Pause will now start in June 2015, making this a 5 year 10 month pause.”

Two days later Monckton announced “Now that the small La Niña that has recently ended has begun to have its effect on global temperatures, the UAH monthly global mean lower-troposphere anomalies now show a further sharp drop, lengthening the New Pause by three months, from 5 years 7 months last month to 5 years 10 months this month.” With a graph showing the pause starting in June 2015.

Last edited 2 months ago by Bellman
Bellman
Reply to  Bellman
July 5, 2021 11:24 am

Part 3

In Moncktons April post I presented a look up table for what next month’s pause would like depending on the UAH update for April

April Range     Pause Start  Pause Length
----------------------------------------------------------
+0.40 - +0.65   Aug 2015     5 Years 9 Months
+0.32 - +0.39   Jul 2015     5 Years 10 Months 
+0.20 - +0.31   Jun 2015     5 Years 11 Months
-0.09 - +0.19   May 2015     6 Years 0 Months
-0.28 - -0.09   Apr 2015     6 Years 1 Month
-0.49 - -0.29   Mar 2015     6 Years 2 Months
-0.57 - -0.50   Feb 2015     6 Years 3 Months
-0.71 - -0.58   Jan 2015     6 Years 4 Months

I also said I expected the pause to be at least 6 years old.

April 2021 was -0.05°C in UAH, Monckton said the pause now started in May 2015 and was 6 years old.

Bellman
Reply to  Bellman
July 5, 2021 11:30 am

Part the Last

1st of June say the May update to UAH, in which I said “The Monckton Pause now starts in April 2015, making it 6 years and 2 month old.”

Do you want to take a guess as to when Monckton said the pause started that month?

With this month’s correct prediction I make that 5 correct predictions, with one month where I forgot.

Gordon A. Dressler
Reply to  Bellman
July 6, 2021 7:28 am

Bellman, I have no idea why you are so focused on Lord Monckton’s “prediction”, or not, as to the start date of the current pause in global warming.

Most reasoning people can just examine by themselves the graph presented at the start of the above article and easily see that the “pause” in GLAT warming started in 2013-2014, if not even earlier in 2001-2002.

“There are none so blind as those that have eyes yet refuse to see.” — old proverb

Bellman
Reply to  Gordon A. Dressler
July 6, 2021 9:45 am

Bellman, I have no idea why you are so focused on Lord Monckton’s “prediction”…

Possibly because you insisted I had to present evidence or you wouldn’t accept it.

Most reasoning people can just examine by themselves the graph presented at the start of the above article and easily see that the “pause” in GLAT warming started in 2013-2014, if not even earlier in 2001-2002.

I’m not interested in what most people think they can see, I’m interested in what the evidence is. I’d argue with anyone claiming a pause started in 2013-14, but it’s Monkton’s pause, starting in 2015 which is the one that gets posted here every month.

TheFinalNail
Reply to  Bellman
July 4, 2021 3:31 am

Bellman, Lord Monckton does indeed say that the new pause starts in March 2015. Carlo, Monte must be thinking you’re a ‘psychic remote viewing mind reader’.

Geoff Sherrington
Reply to  Bellman
July 3, 2021 3:55 am

Bellman,
Starting now and going back until a trend changes by some specified amount is a part of time series analysis that i find to be good for picking up inflexion points. If there is something wrong with the method. please describe the problem. It is nought to do with cherry picking. Geoff S

Bellman
Reply to  Geoff Sherrington
July 3, 2021 4:36 am

Starting now and going back until a trend changes by some specified amount is a part of time series analysis that i find to be good for picking up inflexion points.

And is it all right if I do that to find October 2010, where the trend “changed” to over 0.4°C / decade?

If there is something wrong with the method. please describe the problem.

A couple of problems spring to mind. First you want to make sure that the change in trend is statistically significant. Second you want to ensure that the inflection point doesn’t result in a discontinuity.

Here’s Monckton’s pause in context. Does this look like an inflection point?

20210703wuwt1.png
Bellman
Reply to  Bellman
July 3, 2021 4:39 am

And here’s my equally nonsensical inflection point.

Bellman
Reply to  Bellman
July 3, 2021 4:40 am

Sorry, here’s the graph.

20210703wuwt2.png
Loydo
Reply to  Bellman
July 3, 2021 5:47 am

Thats pretty stark Bellman. It’s about time some the actual skeptics here stopped giving his ‘pause’ disinformation a free pass. It’s ludicrous as you plainly show.

Jim Gorman
Reply to  Loydo
July 4, 2021 6:38 am

When you see an inflection point, it doesn’t mean that something is wrong with the analysis. It means something occurred that needs to be addressed.

Have you ever heard of differentiation? It means breaking a function into smaller and smaller (or shorter and shorter) pieces to determine slopes. Difficult to do with time varying periodic functions such as the oscillations contained in climate.

Bellman
Reply to  Jim Gorman
July 4, 2021 8:47 am

The problem is I’m not seeing any inflection point, and Monckton is not providing any statistically significant evidence that one exists. That doesn’t mean there hasn’t been a change around 2016, but at present there is not remotely enough evidence to suggest either a slow down or an acceleration.

Gordon A. Dressler
Reply to  Loydo
July 4, 2021 8:02 am

Guess what? It would even that much more stark in one were to plot the linked liked linear regression analysis performed, say, over every tow-year interval.

Statistics can be misused by those on either side of a argument when they seek to spread disinformation.

Clyde Spencer
Reply to  Bellman
July 3, 2021 10:42 am

What one does, and how they do it, is determined by the purpose of the exercise. That is, it is necessary to define what the analysis intends to accomplish.

When Monckton says that there has been a hiatus in warming for 6 years, it is a true statement. What should be argued is the meaning or importance of the statement. You are engaging in unproductive nit-picking! It strikes me as an emotion-based response because you understand the implication — there is a possibility that the recent upward trend is transient or cyclical and the CO2-driven paradigm would be invalidated.

Bellman
Reply to  Clyde Spencer
July 3, 2021 1:43 pm

When Monckton says that there has been a hiatus in warming for 6 years, it is a true statement.

It’s a true statement with no meaning. Statistics doesn’t make “true” statements. Statistics understands that everything is uncertain, and hence everything has to be treated with skepticism. The best you can do is have enough evidence that it is very likely you are correct.

If you define a hiatus or pause as any arbitrary length of time using the data set of your choice where the average trend is zero and ignore any uncertainty in the trend, then yes, it’s a true fact that there has been a hiatus of exactly 6 years and 10 months. It is without anymore meaning than that.

If on the other hand you have a hypothesis that warming actually stopped a few years ago, then statistically you have to device an argument that supports that, show that this is unlikely to have happened purely by chance. The standard way to do that is to define a null-hypothesis – that is what would the last 6 years have looked like if there had been no hiatus, and then see if the the actual data is significantly different to the null-hypothesis.

Monckton fails to even attempt that, and with good reason. The uncertainty of a trend line of 7 years is not significantly different than the previous 30+ years, the overall trend has actually increased during the hiatus.

Gordon A. Dressler
Reply to  Bellman
July 4, 2021 1:19 pm

Bellman posted “Statistics doesn’t make “true” statements” and “Statistics understands that . . .”

When one thinks that mathematical equations and/or methods behave like humans, it is pointless to respond.

Bellman
Reply to  Gordon A. Dressler
July 4, 2021 2:07 pm

Sorry if my language was a little poetic. Substitute statisticians for statistics if it will help you understand.

Gordon A. Dressler
Reply to  Bellman
July 4, 2021 4:09 pm

Your language was not “a little poetic”. It was dead wrong.

That I understand fully.

Bellman
Reply to  Gordon A. Dressler
July 5, 2021 4:46 am

Aside from the pathetic fallacy, what exactly did you understand to be wrong with what I
said?

Jim Gorman
Reply to  Bellman
July 3, 2021 5:57 am

You are missing the whole point. Why has CO2 ceased to be relevant to increased temperature? That is what your “discontinuity” shows! You need to be able to recognize these points in order to analyze natural variations.

Bellman
Reply to  Jim Gorman
July 3, 2021 8:11 am

What makes you think CO2 has become irrelevant? I purposely avoid talking about CO2 and only talk about time, i.e the trend against time, but as CO2 has been increasing with time it follows that there is a similar correlation with CO2. When I get back to my laptop I’ll show you what the correlation between UAH and CO2 looks like. I suspect as with time, the last 6 years will if anything have strengthened the correlation

Note before anyone jumps on me for saying that, correlation us not causation. I’m not saying the correlation proves that CO2 caused the warming, but it certainly doesn’t disprove it

Clyde Spencer
Reply to  Bellman
July 3, 2021 10:49 am

I suspect as with time, the last 6 years will if anything have strengthened the correlation

I look forward to you demonstrating your claim.

The importance of 6 and 20 year pauses is that the coupling between CO2 and temperature is not as tight as you and others would like to think. That then brings into question cause and effect, and whether the correlation could be spurious.

Bellman
Reply to  Clyde Spencer
July 3, 2021 4:36 pm

“I look forward to you demonstrating your claim.”

Happy to be of service.

Bellman
Reply to  Bellman
July 3, 2021 2:36 pm

Here’s the CO2 vs UAH correlation I promised.

This shows monthly UAH anomaly compared with monthly CO2 levels (I haven’t adjusted CO2 to remove the annual cycle here, though I probably should). The Pause months are highlighted in blue. The black line shows the linear best fit for the pre-pause months, the blue line the best fit for all months.

20210703wuwt3.png
Bellman
Reply to  Bellman
July 3, 2021 2:46 pm

In figures – before the pause, the slope is rising at the rate of 0.006°C per ppm of CO2, including the pause months the slope is 0.007°C per ppm of CO2. R^2 for pre-pause is 0.29, whilst the current R^2 is 0.47.

As I suggested, including the pause months makes the relationship stronger.

This becomes even stronger if I look at the log of CO2. Pre-pause the relationship is 1.6°C / doubling of CO2, after pause it’s 1.9°C / doubling.

Clyde Spencer
Reply to  Bellman
July 3, 2021 3:13 pm

You criticized Monckton for not showing uncertainties:

https://wattsupwiththat.com/2021/07/02/uah-global-temperature-update-for-june-2021-0-01-deg-c/#comment-3283742

Where are your uncertainties?

Need I remind you that an R^2 value of 0.47 means that less than half the variance in temperature can explained or predicted by CO2? Again, flipping a coin gives one a 50% chance of coming up heads. That’s better than 47%. And that is before taking into consideration the uncertainty associated with the coefficient of determination.

Bellman
Reply to  Clyde Spencer
July 3, 2021 4:13 pm

With uncertainties:

Up to March 2015
UAH ~ CO2: 0.0062 ± 0.0005
UAH ~ log2(CO2): 1.59 ± 0.12

Up to May 2021

UAH ~ CO2: 0.0073 ± 0.0003
UAH ~ log2(CO2): 1.90 ± 0.09

All uncertainties are one standard error.
p value for all cases is below 2.2e-16

They really don’t have much of an effect, though non of this is including any correction for auto correlation.

Need I remind you that an R^2 value of 0.47 means that less than half the variance in temperature can explained or predicted by CO2?

That’s right. There’s a lot of variation in the monthly values, yet almost half of that, despite all the “pauses” can be explained by the CO2 that month.

Again, flipping a coin gives one a 50% chance of coming up heads. That’s better than 47%.

Not this again. An r^2 value of 0.5 has nothing to do with flipping a coin.

Carlo, Monte
Reply to  Bellman
July 3, 2021 9:32 pm

There is a lot more to proper uncertainty calculations than just simple standard deviation of an average.

Bellman
Reply to  Carlo, Monte
July 4, 2021 2:39 pm

You’re free to do your own calculations.

Clyde Spencer
Reply to  Bellman
July 4, 2021 11:49 am

I should have been more specific, even though I was talking about R^2. What is the uncertainty of the correlation coefficient (R) and its derivative coefficient of determination (R^2)?

The point of the R^2 value is how well it can be expected to predict the value of an interpolated or extrapolated dependent variable, in this case, temperature. Low R^2 values mean low confidence in predictions. Coin tosses work similarly. With an unfair coin (two heads) one has 100% confidence that heads will come up on a toss; with a two-tailed coin, one has 0% confidence that a head will come up. With a fair coin, one has approximately 50% confidence in a head (or tail) coming up, allowing for ‘runs.’ While a coin toss has a bi-variate outcome, it illustrates that after a series of tosses that the prediction will not be perfect, but it will be around half heads.

In a similar manner, an R^2 value of 1.0 means one has perfect confidence in the prediction, and with an R^2 value of 0.0, one has no confidence. Thus, with 50% confidence, one can expect a set of numbers above and below the nominal y-value, showing the behavior of runs just as with a coin toss. It is not a perfect analogy (no analogy is, or else it wouldn’t be an analogy) but it illustrates that an R^2 value equal to or less than 50% has little practical utility for prediction, other than regressing around the mean or nominal y value. That is, almost all will be wrong, with about half being too high and half being too low.

I think that it would be reasonable to speculate that with R^2 being less than 50%, the variance will increase, meaning that the magnitude of the error (delta y) will increase. More formally, I would expect the average absolute value of ysubn would approach infinity as R^2 approaches zero.

Bellman
Reply to  Clyde Spencer
July 4, 2021 2:25 pm

The point of the R^2 value is how well it can be expected to predict the value of an interpolated or extrapolated dependent variable, in this case, temperature.

It tells you how much of the existing variation can be predicted. That doesn’t translate into how good the prediction is.

“In a similar manner, an R^2 value of 1.0 means one has perfect confidence in the prediction”

Correct.

“and with an R^2 value of 0.0, one has no confidence.

Not necessarily true. It means your independent variables tell you nothing about the dependent one. You might still be able to have confidence in the prediction, it’s just that it will be the same as looking at the mean value.

Thus, with 50% confidence, one can expect a set of numbers above and below the nominal y-value, showing the behavior of runs just as with a coin toss.

But that would be true no matter what the r^2 value was. You can always have approximately the same number of result above or below the prediction.

What an r^2 value roughly tells you is that however good or bad your prediction is, it might have been twice as bad if you had just used the mean as a prediction.

I think that it would be reasonable to speculate that with R^2 being less than 50%, the variance will increase, meaning that the magnitude of the error (delta y) will increase.

You’ve completely lost me here. What do you mean “will increase”? Are you predicting how the variance of temperature will change in the future, or what?

Clyde Spencer
Reply to  Bellman
July 5, 2021 11:38 am

It tells you how much of the existing variation can be predicted. That doesn’t translate into how good the prediction is.

Classically, R^2 is used to assess either the percentage of variance that is explained, or the percentage of the predicted variance that is applicable. I assert that an R^2 value of 1.000 means that any predictions will be perfect. Ergo, R^2 values approaching 1.000 will approach perfect prediction.

If you disagree, then explain how you would assess “how good the prediction is.”

Bellman
Reply to  Clyde Spencer
July 5, 2021 1:46 pm

I don’t know what you mean by “the percentage of the predicted variance that is applicable”, but I don;t disagree with your comments about R^2 values approaching 1. I just don’t know what you mean by an R^2 value below 0.5 having “little practical utility for prediction”, even less why you think “almost all will be wrong, with about half being too high and half being too low.” applies to R^2 of 0.5 or below, and not say an R^2 of 0.95.

ResourceGuy
Reply to  Bellman
July 4, 2021 5:33 am

So CO2 is correlated with AMO.

Bellman
Reply to  ResourceGuy
July 4, 2021 2:40 pm

If it is, there was’nt any AMO till the 20th century.

Richard M
Reply to  Bellman
July 3, 2021 9:08 am

Here’s a better view of where understanding “pauses” helps.

https://woodfortrees.org/plot/uah6/from:1997/to/plot/uah6/from:1997/to:2015/trend/plot/uah6/from:2015.25/to/trend

You can see two relevant pauses. That begs the question of what happened to lead to such a big step change. Obviously, CO2 cannot be the answer.

What one finds with a little investigation is the PDO was negative over most of the first pause and went positive in 2014. It remained positive until the recent La Nina occurred. Where it goes from here is still unknown.

Bellman
Reply to  Richard M
July 3, 2021 2:51 pm

“That begs the question of what happened to lead to such a big step change. Obviously, CO2 cannot be the answer.”

It also raises the question of whether the selected trend lines are spurious or not. You choose to end one and start another at a point that results in an instantaneous rise in global temperature. This seems unlikely.

CO2 couldn’t cause that, but what could? How does a cyclic PDO cause a two flat temperature trends joined by an instantaneous jump of around 0.25°C?

Jim Gorman
Reply to  Bellman
July 4, 2021 10:24 am

How can temperature when plotted over time result in “spurious trends”? The trends are determined by the temperature variable. The chosen length of time merely shows what temperature did over that time.

As to instantaneous “jumps”, you obviously have no experience with signals made up of multiple components whose frequency and phase all drift over time. Not only can amplitudes change rapidly but the sign can change instantaneously too.

Bellman
Reply to  Jim Gorman
July 4, 2021 3:02 pm

How can temperature when plotted over time result in “spurious trends”?”

Elsewhere you said

The worst assumption is that a linear regression will give an adequate representation of a trend to begin with.

I think that answers your question. I would say a trend is spurious if it doesn’t properly represent the correct trend. This might happen say because you are using the wrong formula for your trend, e.g. a linear formula when the actual trend is logarithmic, or because you are over fitting the available data, seeing trends that are really just noise.

Not only can amplitudes change rapidly but the sign can change instantaneously too.

Global climate is not a n electrical circuit. I’m sure there are situations where a long term change in temperature could happen on the scale of a month, e.g. a comet hitting the earth, but there doesn’t seem to be any thing happening in March 2015 that could have caused an actual global 6 year rise in temperature of around 0.25°C. And if it did, I wonder why it wasn’t obvious a few months ago, when it was being claimed this discontinuity happened in October 2015.

Jim Gorman
Reply to  Bellman
July 5, 2021 9:18 am

A trend is spurious if it doesn’t adequately forecast the future. Otherwise it is nothing more than an exercise in drawing a pretty curve.

It all depends on your perspective doesn’t it? In terms of tens of thousands of years, the cooling trend in the last several months is pretty instantaneous.

Bellman
Reply to  Jim Gorman
July 5, 2021 9:43 am

Nonsense, a trend can be to eamine what has happened, without it predicting the future. Monckton says the pause doesn’t predict the future. That’s not why it’s spurious.

Bryan A
Reply to  Bellman
July 2, 2021 7:56 pm

Warming rate of 350C / Year since 7am

Bruce Cobb
Reply to  Bellman
July 3, 2021 6:33 am

“Pause now starts March 2015.”
Nope. It starts from today, working back. Do try to keep up.

Bellman
July 2, 2021 7:00 pm

There have been 14 months with USA48 anomaly higher than 1.44, but none during summer months. August 1995 was closest at 1.41.

This is one case where it would be interesting to see absolute temperatures, rather than monthly anomalies.

Reply to  Bellman
July 2, 2021 8:18 pm

This is probably what happened in 1936: a year of seasonal extremes in the USA48.

Mike Dubrasich
Reply to  Bellman
July 2, 2021 9:58 pm

Why? It’s called Global warming for a reason. Or you mimic Moonbeam Brown, who called California “the epicenter of global warming,” and concentrate on one square meter in your driveway.

July 2, 2021 7:15 pm

Not long ago I suggested here that the ongoing La Niña conditions might soon send the anomalies below the zero line.
The next step is to see whether on average La Niña events have begun to dominate over El Ninos.
If they have and if that continues then at some point I would expect the first downward step at the end of a PDO phase that anyone has seen in living memory.
After that the rise in CO2 should slowly tail off and eventually go into reverse.
Unless the sun reactivates.

MarkW
Reply to  Stephen Wilde
July 2, 2021 7:27 pm

No matter how cool it gets, the usual suspects will still assure us that we are about to be cooked.

Mike
Reply to  MarkW
July 2, 2021 11:28 pm

Yes it’s either warming or weirding. That’s it and that’s all.

Pillage Idiot
Reply to  Stephen Wilde
July 2, 2021 7:45 pm

The IRI site has the La Nina probability overtaking the neutral probability by the OND quarter.

Their prediction is even greater for NDJ – with the La Nina probability over 50% for that time period.

Reply to  Stephen Wilde
July 2, 2021 8:19 pm

La Nina is set to return this coming NH winter.

Loydo
Reply to  Stephen Wilde
July 2, 2021 9:20 pm

“If they have and if that continues”..

Hope springs eternal.

This study shows more frequent strong El Niño events and predicts that trend will continue.

“Since the 1970s, El Niño has changed its origination from the eastern Pacific to the western Pacific, along with increased strong El Niño events due to a background warming in the western Pacific warm pool”

https://www.pnas.org/content/116/45/22512

Richard M
Reply to  Loydo
July 2, 2021 9:46 pm

The land areas near the PWP are highly populated. That means lots of pollution. This could be a significant factor in the warming mentioned in your study.

I’ve been focused on Atlantic warming from increased salinity and pollution but directly warming the PWP is another way that pollution can have an oversized impact.

Julian Flood
Reply to  Richard M
July 2, 2021 11:35 pm

Our civilisation dumps increasing amounts of light oil and surfactant onto the oceans which smooths them. Smoothed surfaces have reduced wave breaking – fewer salt aerosols so less low level stratocumulus. Smooths have lower albedo. Reduced wave stirring decreases nutrient uplift, plankton have to live deeper where light levels are lower – different and lower populations and carbon fixation methods which reduces carbon export to the deep ocean and that carbon has higher proportion of C13 – more CO2 left in the atmosphere with a high C12 signal.

Evans and Ruf use GPS signals to find microplastics which are associated with oil/surfactant smooths so their paper gives some idea of the size of the problem.

Part of AGW is caused by pollution. It’s imperative that science quantifies how much. Dr Evans to lead. (!)

JF

Reply to  Julian Flood
July 3, 2021 12:11 am

Explain 1940’s cooling. Huge numbers of WW2 shipping sunk with oil sheens across the oceans;

Julian Flood
Reply to  Joel O'Bryan
July 3, 2021 12:22 am

‘Why the cooling?’ said Prof Tom Wigley… Oh no, sorry, he said ‘why the blip?’ Look up how climate ‘science’ conspired to hide the warming during WW2 which was in the ’40s I believe.

JF

Clyde Spencer
Reply to  Julian Flood
July 3, 2021 11:03 am

Smooths have lower albedo.

But higher specular reflectance.

Julian Flood
Reply to  Clyde Spencer
July 3, 2021 2:53 pm

Interesting. When albedo is measured, does it not include all effects, including specular reflection? Otherwise the ‘warm the fish pond with oil trick’ wouldn’t work.

JF

Clyde Spencer
Reply to  Julian Flood
July 4, 2021 12:03 pm

You seem to be unacquainted with specular reflection. However, you are in good company, as most climatologists are similarly unacquainted. Check here:

https://wattsupwiththat.com/2016/09/12/why-albedo-is-the-wrong-measure-of-reflectivity-for-modeling-climate/

Albedo is a relative measure of retroreflectance, only capturing light returned in the direction of the source. It is typically only appropriate for diffuse reflectors such as granular materials and clouds. The important point is that when the angle of incidence is not normal to the surface, light is reflected away from the source, and cannot be observed or measured looking in the direction opposite to the source.

You may be confusing different phenomena. First of all, oil on a pond will probably reduce the rate of evaporation, thereby reducing cooling. Also, many common oils have a lower refractive index than water, meaning that less light will be reflected and more light will be absorbed.

Chris Hanley
Reply to  Loydo
July 2, 2021 10:36 pm

I don’t see any wish or desire expressed in Stephen Wilde’s comment, merely an expectation.
From your link: “… This suggests the controlling factors that may lead to increased extreme El Niño events in the future. If the observed background changes continue under future anthropogenic forcing, more frequent extreme El Niño events will induce profound socioeconomic consequences …”.
Is that what you hope for?

Loydo
Reply to  Chris Hanley
July 3, 2021 6:05 am

No, I wished SW was right, but I think his fervant hopes for multidecadal cooling are utterly forlorn. Instead, and I really am sorry to crush his hopes here, we’ve already baked in a few thousand years of the opposite. Such a pity.

Last edited 2 months ago by Loydo
Mike
Reply to  Loydo
July 2, 2021 11:33 pm

This study shows more frequent strong El Niño events and predicts that trend will continue.”

Predicts? Based on? What’s going on now? Or based on the anthropogenic warming hypothesis? Either way, I predict their prediction will decompose into a big heap of steaming compost – sooner or later.
Some of us keep one eye on history, others just bend over..

Loydo
Reply to  Mike
July 3, 2021 1:29 am

“Predicts? Based on?”

Observations. Read it.

Archer
Reply to  Loydo
July 3, 2021 3:51 am

Past performance is no guarantee of future results.

Reply to  Loydo
July 2, 2021 11:40 pm

Don’t trust any organization that calls itself PNAS.

clarence.t
Reply to  Loydo
July 3, 2021 12:56 am

Based on CMIP5 models.. so meaningless and totally irrelevant to reality.

Even the modellers have admitted as such, and gone to CMIP6..

which is even worse.

Loydo
Reply to  clarence.t
July 3, 2021 1:28 am

Based on obervations.

MethodsDefinition of El Niño Years (1901 through 2017).The SST anomaly averaged in the NINO 3.4 region (5°N–5°S, 120°–170°W), known as Oceanic Niño Index (ONI), is averaged for October-November-December-January-February (ONDJF) to identify El Niño years, because ONI has largest variances during ONDJF (SI Appendix, Fig. S3B). An El Niño year is defined as ONDJF ONI is greater than or equals to 0.6 °C. Using the linearly detrended data, 33 El Niño years are identified.

lee
Reply to  Loydo
July 3, 2021 2:13 am

Since 1901? You mean back when Phil Jones CRU said the normals between 40 and 60S were mostly made up. There were few real observations in the Southern Hemisphere until the 1950’s. That makes the early “data” more than suspect. The oceans aren’t really flash even now.

clarence.t
Reply to  Loydo
July 3, 2021 4:02 am

Seriously Loydo?

Observations of the future.. so funny !!

“The Coupled Model Intercomparison Project phase 5 (CMIP5) models’ projections demonstrate that both the frequency and intensity of the strong El Niño events will… blah..blah..”

Please at least try to read and comprehend links before you post them !

Does it really surprise anyone that El Ninos eventuate after the series of strong solar maxima in the latter half of last century.

The late 1970’s was the coldest period in the last 120 or so year, bottom of the AMO etc.. always the chosen starting point for alarmist trash-talk.

We are heading into solar minima period now.

Loydo
Reply to  clarence.t
July 3, 2021 4:29 am

What part of1901 through 2017′ did you not understand? Can’t be easy with your arms waving about so hard like that.

SAMURAI
July 2, 2021 9:34 pm

The CAGW zealots will soon have some serious explaining to do….

NOAA’s ENSO projections, Pacific equatorial SST and deep water ocean temperature data show a rare double-dip La Niña cycle will very likely occur at this end of this year, meaning sometime next year, the UAH6.0 could very likely hit -0.3~-0.4C.

The “new and improved!!” CMIP6 mean computer model projections show the global temp anomaly should now be +1.4C. with a global warming trend of around 0.28C/decade—- not so much..

The PDO and AMO Indexes show both ocean cycles will likely reenter their respective 30-year cool cycles in a few years, bringing falling global temp trends for 30+ years as occurred from 1880~1913 and 1945~1978.

CAGW is so busted.

TheFinalNail
Reply to  SAMURAI
July 3, 2021 1:09 am

Bear in mind, we heard all this over a decade ago too, before the last double-dip La Nina, around 2010.

Who can forget Don Easterbook’s forecast of imminent prolonged cooling caused by the PDO (2008); or David Archibald’s forecast of drastic cooling caused by reduced sunspots during solar cylce 24 (2009), etc.

The funny thing is, the PDO did hit a cooling phase, and solar cycle 24 did have much reduced sunspots; but the climate responded by producing the warmest decade on record, even using UAH. Just cautioning against hubris, that’s all.

Loydo
Reply to  TheFinalNail
July 3, 2021 1:40 am

Who can forget Macrae’s:

“I’VE PREDICTED MACRAE’S PREDICTED COOLING TO START VERY, VERY SOON. I EMAILED THAT. LOOK WHAT’S HAPPENED TO AUSTRALIA – SEVERAL VERY COOLISH MONTHS. I EMAILED THAT TOO.”

Reply to  Loydo
July 3, 2021 3:13 am

Loydo – I keep records of all my posts – I cannot find that quote. Please provide the url.

Loydo
Reply to  ALLAN MACRAE
July 3, 2021 4:52 am

Of course you do, and you email them to the newspapers too right? Btw, I think that’s a tactical mistake. I find hand-written is so much more gratifying.

Reply to  Loydo
July 3, 2021 10:17 pm

Given your non-response, I can only assume Loydo that you fabricated that alleged quotation.
Shameful, but typical warmist misconduct.

Dave Fair
Reply to  TheFinalNail
July 3, 2021 1:23 pm

The UN IPCC CliSciFi models predicted significant 21st Century warming. The Pause killed that. The effects of the Super El Nino are now disappearing. I don’t know what the future holds, but it is not CAGW.

angech
Reply to  TheFinalNail
July 3, 2021 5:40 pm

Samurai, if only it were so easy or true.
El Niño/ LaNina are extremely difficult to predict even when all the ducks line up in a row.
Lets hope it is true or even that we have a triplet.
Yet wanting La Ninas to restore the balance is sort of ducking the global warming effect of CO2.
I want cold, not because of the consequences but because the arguments and science have reduced to pointing at the weather and temperature only.

bdgwx
Reply to  SAMURAI
July 3, 2021 7:19 pm

Assuming you are the same poster that posts on Dr. Spencer’s blog you also said the UAH TLT anomaly would hit -0.2 to -0.3 by April 2021. And that was based on the old 1981-2010 baseline. On the new 1991-2020 baseline your prediction is -0.32 to -0.42. The lowest UAH TLT got was -0.05 on the 1991-2020 baseline or +0.07 on the 1981-2010 baseline.

TheFinalNail
July 2, 2021 9:38 pm

The linear warming trend since January, 1979 remains at +0.14 C/decade…

I make it +0.13 C/decade from Jan 1979. It’s 1.4 C/decade from the start of the data set, which is Dec 1978. However, Excel could be wrong.

Reply to  TheFinalNail
July 3, 2021 1:28 am

You can look it up here I make it 0.1354 °C/decade from Jan 1979 to May 2021. That gadget doesn’t have June yet, because Roy has not put the updated table online yet. June would have brought it down, so 0.13 is a plausible rounding. And it is also plausible that Dec 1978 (not in the gadget’s data) would put it back over the line to 0.14.

Last edited 2 months ago by Nick Stokes
clarence.t
Reply to  Nick Stokes
July 3, 2021 4:11 am

Yep, keep starting in 1979, coldest period in last 120 years.

Bottom of the AMO cycle etc etc.

Great for alarmist trash-talk.

TheFinalNail
Reply to  clarence.t
July 3, 2021 4:50 am

Erm, it was Roy Spencer who started quoting the trend from 1979. Roy the Alarmist?

Tom Abbott
Reply to  TheFinalNail
July 3, 2021 12:43 pm

Roy’s satellite readings didn’t start until 1979, and since Roy is discussing his satellite results, 1979 is as far back as he can go.

No conspiracy there.

TheFinalNail
Reply to  Tom Abbott
July 4, 2021 3:41 am

I understand that. Apparently clarence.t doesn’t.

Tom Abbott
Reply to  TheFinalNail
July 4, 2021 4:09 am

I was replying to what clarence said and I’m not sure how your name got up there. I thought I had clicked in the right place. Sorry about that.

Loydo
Reply to  clarence.t
July 3, 2021 4:55 am

“coldest period in last 120 years”

Really? Can you share a link to what you’re basing that on?

TheFinalNail
Reply to  Nick Stokes
July 3, 2021 5:07 am

Your figures for Jan 1979 – May 2021 are replicated in Excel, Nick. Adding the June update gives 0.1348 C/dec, which rounds to 0.13 from Jan 1979, not 0.14 as Roy says. Looks like he’s running his trend from Dec 1978, which even with the June update is still rounding to 0.14 C/dec.

Quick, somebody alert Roy! He’s missing a trick here.

Clyde Spencer
Reply to  Nick Stokes
July 3, 2021 11:14 am

Is the “4” in the first sentence the result of truncation or rounding? Are you sure you don’t want to add one or two more digits to make it look more precise?

Reply to  Clyde Spencer
July 3, 2021 4:32 pm

It is the result of a calculation. I always give those to whatever number of digits anyone is likely to want to know about.

Here the issue is, what might cause a change of rounding in Roy’s figures. The fact that it is 0.1354 is definitely relevant.

Clyde Spencer
Reply to  Nick Stokes
July 4, 2021 12:08 pm

I only want to know about the significant figures, not the digits returned by a calculator that is designed to assume the numbers entered are exact, with trailing zeros.

One has to be smarter than one’s tools. Otherwise someone is liable to accuse you of being as dumb as a hammer.

Last edited 2 months ago by Clyde Spencer
Bellman
Reply to  Clyde Spencer
July 4, 2021 1:43 pm

Of course the numbers aren’t exact and Dr Roy Spencer and Lord Monckton and all should really quote their confidence intervals. But, my problem with insisting on rounding to a couple of sf, is that it actually reduces the confidence as the rounding adds extra uncertainty.

Consider, the current rate of warming is around 1.35°C / century with a 2 sigma confidence interval of around 0.5°C / century. It might seem redundant to quote the 2nd decimal place, and to a a large extent I’d agree. But with 2dp, we know with 98% confidence that the actual rate of warming is between 0.85 and 1.85, with 1.35 being the mid point. If we round this to 1.4, what is the confidence interval? Is it still ±0.5?

It’s not a big issue, it’s just that I find the idea of prematurely removing data unnecessary as long as people realise the uncertainty. Next month we may well see the trend drop below 1.35, at which point, with no mention of uncertainty, Spencer will be saying the rate of warming has dropped from 1.4 to 1.3°C / century, and many here will see this as a sudden 7% drop in the rate of warming.

Clyde Spencer
Reply to  Bellman
July 5, 2021 11:43 am

as long as people realise the uncertainty.

If the uncertainty isn’t expressed explicitly [which it rarely is], then the only way that people can know is implicitly through the accepted use of the proper display of significant figures.

Bellman
Reply to  Clyde Spencer
July 5, 2021 1:52 pm

But even if they use the “correct” number of digits, you still have little idea of what the uncertainty is. How do I figure out how uncertain Spencer’s 0.14 is, or Monckton’s -0.23? Both have the same number of digits but extremely different levels of uncertainty.

Richard Barraclough
Reply to  Nick Stokes
July 4, 2021 6:33 am

Including June, it’s still 0.14, rounded to 2 decimal places. Another month with zero anomaly would bring it down to 0.13

Bellman
Reply to  TheFinalNail
July 3, 2021 4:48 am

I make the trend including December 1978 to be 0.1351°C / decade, but starting in January 1979 it’s 0.1348°C / decade.

This is the problem with only showing 2 sf. The trends are only differing by a few thousands of a degree per century, but it looks much bigger when rounded up or down.

Clyde Spencer
Reply to  Bellman
July 3, 2021 11:19 am

What is the uncertainty on your 0.1351 figure? Are you sure you don’t want to add a couple more digits?

Bellman
Reply to  Clyde Spencer
July 3, 2021 12:48 pm

Using the Skeptical Science Trend Calculator, around ±0.05°C / decade. As I meant to say, the precise figure isn’t significant, but it does provide additional context.

bdgwx
Reply to  Clyde Spencer
July 3, 2021 7:01 pm

LINEST says the standard uncertainty is 0.0065. FWIW In the past I report all trends with the same digits as the 2nd significant digit from the calculated standard uncertainty like how Bellman does it. I do this as reasonable balance between transparency and not being excess with digits. If someone else wants to do further rounding that is fine. I’ve had people accuse me of artificially increasing the trend when I’ve rounded it like UAH does. Maybe from now on I should display the exact output and let the WUWT community figure out what they want to do it from there.

Here are the exact digits from LINEST from 1979/01 to 2021/06 via the UAH 3-digit file up to 2021/05 and manually adding the 2-digit value from Dr. Spencer’s recent post. Note that every month is reprocessed with each update so each month will have a different value from it was compared to the previous update. Usually the changes are small with only a few dozen tripping a change at the 3rd digit and only a handful tripping a change at the 2nd digit so what I report below is preliminary.

0.134753757788607000 +/- 0.006509661838057940 C/decade

Last edited 2 months ago by bdgwx
TheFinalNail
Reply to  bdgwx
July 4, 2021 3:51 am

I get those figures too, but I think the uncertainty calculation used for temperature data is done slightly differently than in Excel, to account for autocorrelation. Excel treats each data point as if it’s uncorrelated to the previous one; whereas monthly global temperature values normally don’t vary by too much – they tend to be more correlated with one another rather than truely random. This has to be accounted for in the uncertainty value used in the SkS calculator. Grant Foster (aka Tamino) used to be great at explaining this, but he doesn’t seem to post these days, unfortunately.

Clyde Spencer
Reply to  TheFinalNail
July 4, 2021 12:24 pm

It is a complex subject that most seem to just ignore, which is why I was pulling Bellman’s and Stoke’s chains. Displaying more digits than there are significant figures gives the impression of more precision than is warranted.

The situation isn’t as bad as when I was teaching and hand calculators first came out. I’d give the students a problem with only 3 significant figures (for the benefit of those still using slide rules) and I’d get answers back with 8 or more digits, demonstrating that they had not comprehended the basics of significant figures.

It is just a hot button with me because climatologists as a group seem to pay little attention to significant figures and uncertainty, and routinely report 4 digits when they are not uncommonly dealing with factors with only one or two significant figures! Note that the ‘albedo’ for Earth is commonly expressed as being approximately 0.3, and yet ‘answers’ using it are provided with 3 and 4 digits.

Last edited 2 months ago by Clyde Spencer
Bellman
Reply to  Clyde Spencer
July 4, 2021 1:48 pm

Sorry if I’m a bit defensive here, but there are many here who think that the number of significant figures is some sort of religious commandment. I’m all in favour of presenting a sensible figure, but I’d sooner see too much precision than not enough. This isn’t helped by some who insist that global averages should only be given to the nearest degree.

Clyde Spencer
Reply to  Bellman
July 5, 2021 11:56 am

How do you define “sensible figure?” Some arbitrary number with no logical justification?

I’d sooner see too much precision than not enough.

I’m not suggesting than anyone should use an insufficient number of significant figures. But displaying digits that are not significant is displaying meaningless digits that imply an unjustified precision.

It appears that you have not been exposed to the concept of significant figures in your academic background.

You might want to read this:

https://wattsupwiththat.com/2017/04/12/are-claimed-global-record-temperatures-valid/

Some of the illustrations may not display automatically. However, if you click on the blank space, they should come up.

Bellman
Reply to  Clyde Spencer
July 5, 2021 2:07 pm

How do you define “sensible figure?” Some arbitrary number with no logical justification?

I don’t. Like most things you have to use some judgement.

It appears that you have not been exposed to the concept of significant figures in your academic background.

I’ve been aware of significant figures when I was still using slide rules.

You might want to read this:

Glanced at it briefly. Seems to be making the same mistakes I’ve been arguing about these last few months. For instance you state:

“Averaging tends to remove the random error in rounding when measuring a fixed value. However, the caveats here are that the measurements have to be taken with the same instrument, on the same fixed parameter, such as an angle turned with a transit.”

What reason do you have to claim that you cannot use different measurements with different instruments? Tim Gorman pointed me to a section in Taylor where he specifically uses averages to measure the width and breadth of a sheet of metal, and he says you should be using different instruments and measure at different places to get the most accurate measurement.

Last edited 2 months ago by Bellman
Jim Gorman
Reply to  Bellman
July 6, 2021 7:33 am

I don’t think Dr. Taylor’s book ever said that. You may have interpreted it that way. When measuring the width and breadth you are measuring two different things.

When removing random errors from each direction, you must use the same device for width, but you can use another for the breadth. When averaging each set of measurements, if the distributions are Gaussian, you will remove random measurement errors from each measurement. When combining the measurements, say for area, the uncertainty in each value will propagate through through calculations based upon orthogonality.

Bellman
Reply to  Jim Gorman
July 6, 2021 10:01 am

I’m using a PDF, I found on the internet, of Taylor’s An Introduction To Error Analysis, second edition. The example in question is from Chapter 4 “Statistical Uncertainties of Random Uncertainties”, section 4.5 “Examples”, page 104.

(my emphasis)

Example: Area of a Rectangle

As a first, simple application of the standard deviation of the mean, imagine that we have to measure very accurately the area A of a rectangular plate approximately 2.5 cm X 5 cm. We first find the best available measuring device, which might be a vernier caliper, and then make several measurements of the length l and breadth b of the plate. To allow for irregularities in the sides, we make our measurements at several different positions, and to allow for small defects in the instrument, we use several different calipers (if available). We might make 10 measurements each of / and b and obtain the results shown in Table 4.3.

Screenshot 2021-07-06 180012.png
Clyde Spencer
Reply to  Bellman
July 7, 2021 8:17 am

You said, “Like most things you have to use some judgement.” On what do you base that judgement? How do you defend that it is good judgement? Do you have sufficient status such as held by Lord Kelvin that no one should question your opinion?

You said that you “glanced at it briefly.” That is obvious because you then ask “What reason do you have to claim that you cannot use different measurements with different instruments?” That is answered in the text.

Bellman
Reply to  Clyde Spencer
July 7, 2021 8:57 am

Do you have sufficient status such as held by Lord Kelvin that no one should question your opinion?

No status at all, but then I don’t belief in argument by authority either.

That is obvious because you then ask “What reason do you have to claim that you cannot use different measurements with different instruments?” That is answered in the text.

Could you give me a quote, and do you have “sufficient status” to question Taylor?

Best I can find to justification is:

Different instruments will have different accuracies and may introduce greater imprecision in averaged values.

But that’s not an argument for only using the same instrument, quite the reverse. If all the instruments have different inaccuracies in them, you want to have a variety of them, so you have more chance of eliminating the inaccuracies. If you only have one inaccurate instrument, using it multiple times will still leave you with the same inaccuracy.

Having less precision in you average of inaccurate measurements is a feature not a bug. Use one inaccurate instrument for you sample will give you a precise but inaccurate mean.

Jim Gorman
Reply to  Bellman
July 6, 2021 7:25 am

You are trying to justify adding extra precision to a measurement through simple arithmetic. That is just not possible. It borders on fraud. You have participated for several months about this subject yet you have obviously not availed yourself of the opportunity to study any metrology at all.

I’ve posted the following before and it explains in detail why significant digits are important.

http://www.chemistry.wustl.edu/~coursedev/Online%20tutorials/SigFigs.htm
 

Defining the Terms Used to Discuss Significant Figures

 

Significant Figures: The number of digits used to express a measured or calculated quantity.

By using significant figures, we can show how precise a number is. If we express a number beyond the place to which we have actually measured (and are therefore certain of), we compromise the integrity of what this number is representing. It is important after learning and understanding significant figures to use them properly throughout your scientific career.

 

Precision: A measure of how closely individual measurements agree with one another.

Accuracy: Refers to how closely individual measurements agree with the correct or true value.

“compromise the integrity of what this number is representing.”

I don’t know what this means to you, but it is pretty plain to me. If you add significant digits through arithmetic calculation you are inappropriately telling folks that the measurements used to calculate the number were performed using devices capable of more precision than what was actually used.

If I give you a group of data where each entry is the number of balls in a bucket, are you going to report the mean of balls in a bucket out to five decimal points? Why do you think it is more appropriate to report integer temps any different.

In essence, you are saying the arithmetic calculation of the mean is more important than the measurements themselves. That is what the professors at Washington Univ. are trying to emphasize to the students, when they say, “compromise the integrity of what this number is representing.”

You are basically substituting precision of arithmetic calculations for precision of measurements. You are telling the world that you know better what the measurements truly are because you can do arithmetic on them. You should ask yourself why the world spends billions every year on more and more precise instruments and calibration of those instruments. Your logic leads one to the conclusion that if you just measure something enough times you can increase the precision of the measurements. Who needs micrometers or 10 digit voltmeters? Why can’t we use yardsticks and cheap voltmeters to get measurements precise out to 1/10000th?

Bellman
Reply to  Jim Gorman
July 6, 2021 10:25 am

You are trying to justify adding extra precision to a measurement through simple arithmetic. That is just not possible. It borders on fraud.

That’s a pretty serious accusation.

If you add significant digits through arithmetic calculation you are inappropriately telling folks that the measurements used to calculate the number were performed using devices capable of more precision than what was actually used.

What do you think the measurement is here, and what do you think is it’s precision?

We are talking about a linear regression of several hundred monthly global averages. The calculation can be as precise as you like, the issue is its accuracy. The precision of any specific satellite reading is pretty much irrelevant to the calculation of the trend. The precision of any one satellite reading is also not the same as it’s accuracy.

You are basically substituting precision of arithmetic calculations for precision of measurements.

What I’m really saying in the case of the rate of warming is you shouldn’t confuse the number of decimal places with accuracy. But yes, an arithmetic calculation can be more precise than the sum of it’s parts.

You are telling the world that you know better what the measurements truly are because you can do arithmetic on them.

As I said, my preference is to tell the world what the uncertainty of the measurement is, rather than to let them guess it from the number of decimal places.

Your logic leads one to the conclusion that if you just measure something enough times you can increase the precision of the measurements.

I can’t claim credit for the logic, it’s been around for at least 100 years. But to be clear, for the umpteenth time, averaging measurements does not make those measurements more precise, what it makes more precise and accurate is the mean of the measurements.

Who needs micrometers or 10 digit voltmeters? Why can’t we use yardsticks and cheap voltmeters to get measurements precise out to 1/10000th?

How accurate is a 10 digit voltmeter? How many measurements of a cheap one would you need to get an accuracy of 0.0001? If the cheap one is accurate to ±0.1 say, you would need around a million independent measurements to get that sort of accuracy.

And to get back to the beginning, none of this has anything to do with my point, which is about the uncertainty of a linear regression, where the uncertainty is all about the randomness of the variation and not about how many decimal places are in a monthly global mean.

Carlo, Monte
Reply to  Bellman
July 6, 2021 11:46 am

The usual Bellman-BS.

Bellman
Reply to  Carlo, Monte
July 7, 2021 4:38 am

BS is plant food.

bdgwx
Reply to  TheFinalNail
July 4, 2021 3:58 pm

I just read the Foster and Rahmstorf 2011 publication which documents the method you speak of. Thanks. I was not aware of that technique until now. I like that much better and I think it better represents the true uncertainty. It would be nice if there was easy way to implement that in Excel.

Bellman
Reply to  Bellman
July 7, 2021 5:43 am

Here’s a graph showing the evolving trend, rounded to 3 sf.

20210707wuwt2.png
Bellman
Reply to  Bellman
July 7, 2021 5:44 am

And here’s the same rounded to 2sf

Bellman
Reply to  Bellman
July 7, 2021 5:45 am

Sorry – why won’t the comments allow you to add a picture in the edit?

Here’s the graph to 2sf

20210707wuwt3.png
Clyde Spencer
Reply to  Bellman
July 7, 2021 8:28 am

You should consider reducing the resolution of the graphing scale when using low-precision data. There is no need for an expanded scale in this case, and it is actually inappropriate. You are not using “good judgement.” You are letting your graphing program think for you.

Bellman
Reply to  Clyde Spencer
July 7, 2021 9:27 am

More commandments. So what is the ISO approved scale for a graph of Trends? Also, why do you say the graph is using low precision data? There’s very little difference between each monthly trend.

Bellman
Reply to  Bellman
July 9, 2021 4:21 pm

Here’s the graph with my (very rough) estimate for the 2 sigma uncertainty in grey.

20210710wuwt1.png
Joel Patterson
July 3, 2021 12:42 am

Hard to believe global temp has dropped off a little bit the last six years. It hit 116 in Portland Or. where I live a few days ago!! Beat the old record of 107 back in 1981.

Greg
July 3, 2021 1:29 am

Oh dear, temperatures back to where they were in early 2000s, twenty years ago.

angech
Reply to  Greg
July 3, 2021 1:48 am

Where does this month sit in terms of coolest or warmest? 10th UAH?

Bellman
Reply to  angech
July 3, 2021 4:52 am

It was the 19th warmest or 25th coolest June. Very much in the middle.

angech
Reply to  Bellman
July 3, 2021 8:03 am

Ta

Peta of Newark
July 3, 2021 3:53 am

Makes perfect sense and sea-level rise tells you as much, on 2 counts..

The sea is rising because there is more water in it.

  1. Because there is less water in the land. Water here stores/traps heat and so moderates temperatures rises. As does evaporative cooling from plants and wet-lands
  2. Because there is less water in the atmosphere. Water has very high emissivity and does whatever ‘radiating to space’ that the Troposphere does. Because O2 and N2 have very low emissivities, less water means considerably less radiated IR, in turn meaning the temp has to rise to compensate. There is your trapped heat. ##

Now the crazy bit – that equals cooling of the land surface
Hello hello, what’s the jet stream doing in Seatle when it should be well North of Anchorage

## Lets all go visit the womenfolk of Venus, that’s where they come from innit.
But first, do a Stefan Calc for Earth with a thick-ish dry atmosphere of Oxygen Nitrogen. An Earth which will have weather (convection- heat transport)
You will get a figure of about 730 Celsius. Sound roughly familiar?
Boy, those Venus Girls are hot hot hot

Take a look around Earth and everywhere you look, you will see water doing an epic job of cooling everything/where it goes and touches.

Then visit the BBC here and read between the lines…
Headline:”Clouds of Venus ‘simply too dry’ to support life“Venus is hot because it has no water. period.

In conclusion: Earth atmosphere is warming up because we are drying it out

Loydo
Reply to  Peta of Newark
July 3, 2021 5:08 am

“we are drying it out”

Based on what? Not this obviously:

comment image

Trends in continental temperature and humidity directly linked to ocean warmingByrne et al (2018)

Anything but CO2.

Mark Fife
July 3, 2021 5:01 am

Don’t expect to see any dramatic warming any time soon either.

UAH June 2021.png
TheFinalNail
Reply to  Mark Fife
July 3, 2021 5:58 am

Maybe not dramatic, but it should start to warm up slightly soon. There’s a 5-6 month lag in UAH_TLT temperatures versus ENSO 3.4 that your chart doesn’t seem to account for. That uptick in ENSO at the end of your chart (light blue) hasn’t been reflected in UAH yet; but if the 5-6 month lag correlation holds, it should start to be reflected in warmer temperatures quite soon. The correlation isn’t exact (about 0.4, where 1.0 is perfect) and there are other influences on short term temperature fluctuations; but we should be close to the bottom of the lower troposphere’s cooling reaction to the recent La Nina.

Loydo
Reply to  Mark Fife
July 3, 2021 6:17 am

Until the SO reverses? Thats when the warming is most dramatic – every 2-7 years.

E. Martin
July 3, 2021 6:14 am

Pl show the USCRN data on these updates.

Bruce Cobb
July 3, 2021 7:11 am

While not exactly a cherry-pick, it is certainly highly convenient for Alarmists that the start of satellite based temperatures happens to be right around the end of a 30-year cooling period. Indeed, there was even concern then about a possible ice age.

July 3, 2021 10:17 am

The Northwest heat is indirect evidence of the strength and persistence of the La Niña, whose intensity is disguised by recent tampering with the Pacific SSTs (which also exaggerated the magnitude of the 2016 El Niño).

rah
July 3, 2021 11:36 am

Well while you guys argue about fractions of degrees, this truck driver is going to enjoy sitting around a campfire with friends tonight because of the nice and cool nights we’re having here in Central Indiana right now.

And here it is at 14:30 and I am going out to mow my acre. During most summers that would be a very hot task that would kick my butt but today it’s only 76 deg F right now. Still that lawn mowing beer will taste good!

Another Scott
July 3, 2021 7:18 pm

June included the apocalyptic high temps in the Northwest yet it was -.01 down from the average worldwide. Wonder if that will prompt any discussions about how climate is different from weather.

Bas
July 4, 2021 1:37 am

The claim “-0.01 deg. C” in the title is correct, but misleading. You cannot look at one month for assessing the effect of global warming. You have to take a longer period into account. Therefore, the sentence “The linear warming trend since January, 1979 remains at +0.14 C/decade (+0.12 C/decade over the global-averaged oceans, and +0.18 C/decade over global-averaged land)” is much more relevant.
It warms up. It warms up at a rate unprecedented in the past 2000 years. The reason for this is clear and based on lots of rigorous research. But still, there are people who want to ignore all that, pretend nothing is happening and only want look at the “-0.01 deg. C” (or give it unjustified attention by putting it in a title).

JCalvertN(UK)
July 4, 2021 7:22 am

Are we heading for a double-dip La Nina?
La Nina events seem to occur in pairs quite often.
If so, then this is bad news for California.
(Now that the dams have all been deliberately emptied of water. Desalination anyone?)

Last edited 2 months ago by JCalvertN(UK)