By P Gosselin
Alarmists have claimed lately that global warming has hit Antarctica, finally. But that hasn’t been the case at all at the Japanese Showa station.
Today we present a a temperature chart of the Japanese Antarctic Showa station, located on the East Ongul Island in Queen Maud Land, Antarctica.
The temperature trend at Syowa, also called Showa, has been modestly downward since 1973:
Chart by: Kirye. Data source:: JMA.
There’s been no warming. Right now it’s winter in Antarctica and temperatures there are way below freezing. Recent claims that the ice is melting there are absurd.
For more on Antarctic issues go to our EverythingClimate.com page
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
So basically July winter average temperature can vary by 13 degrees but don’t go down and hang out there without plenty of fossil fuels or you’re gunna die. Right got it.
“Recent claims that the ice is melting there are absurd.”
No-one has claimed that ice is melting at Syowa. People have noted that formation of sea ice has been slow in recent months. That is a matter of photographic observation, and is in no way contradicted by a trend at Syowa.
Temperatures have been a modest, steady decline for more than 40 years. That’s a matter of instrumentation, not photography.
Temperatures in July, only.
Proof?
In fact, the trends at Syowa of the 12 months have been, in C/century, since 1973
-1.03 0.73 0.10 -0.53 -0.37 2.39 -1.03 4.69 -1.59 3.19 2.14 -1.03
so yes, the trend for July was -1.03.
But the whole year trend was 0.64 C/century, warming.
How about also providing the uncertainty envelope for the whole year trend and telling us whether the trend of 0.0064 C/yr is statistically significant before extrapolating out to 100 years?
This article claims in headline that Syowa has been cooling for 40 years (in the fine print, July). Why don’t you ask them?
Your stuff about extrapolating is unscientific nonsense. But 0.064/decade (the other common unit) if you prefer.
So we shouldn’t extrapolate temperatures when hindcasting. Thanks for that.
Unbelievable. You claim something is unscientific nonsense and then turn around and speak unscientific nonsense. Extrapolating the future temps from the past temps *is* unscientific nonsense. It assumes nothing changes which is not how the natural world works.
As Clyde stated you need to state an uncertainty interval for your extrapolation in order for it to be realistic. The future *is* uncertain, you simply can’t ignore that fact – even though it is endemic in the CAGW crowd to do so.
At its base it is nothing more than the argumentative fallacy of Appeal to Tradition. “It’s been this way so it will remain this way”. It’s a failure mark for even middle school debaters let alone anyone claiming to be a physical scientist.
A graph was provided for the article, which by inspection, makes it clear that there was no warming. The July trend, being over -1 deg C, appears to probably be statistically significant.
I misunderstood what you were trying to say, thinking that you were only looking at one year. With 40 years, using units of “per decade” is justified; more than doubling it to “per century” requires that caveats about “continuing as it has been” be explicitly stated.
However, in the case of having, say, only one year of data which might not be statistically significant, multiplying by 100 to get a number two orders of magnitude larger is disingenuous, at best. When getting down into the weeds of two or three digits to the right of the decimal point to find a significant figure, it opens up the possibility of all kinds of mischief if one is talking about a period of time much larger than the measurement period, which requires the uncertainty to be provided.
“With 40 years, using units of “per decade” is justified; more than doubling it to “per century” requires that caveats about “continuing as it has been” be explicitly stated.”
This is the unscientific nonsense – that misunderstands the role of units in a fundamental way. .01 C/year=.1 C/decade=1 C/century. You can use any that you like.
If you’re charged with exceeding 30 mph, it’s no defense that they didn’t observe you for an hour.
If you say a tennis ball serve was 100 mph, that doesn’t mean that you observed it for an hour, or that it will travel 100 miles.
Nick Stokes August 3, 2023 2:03 pm
Nice try, Nick, but what is being measured in your examples is instantaneous velocity. And you’re right, you can put that in any units you want, parsecs per millennium if you wish.
But the same is NOT true about trends. For example, suppose this week averages 3°C warmer than last week.
It’s completely valid to say the trend over the period is 3°C/week … but claiming that means it’s showing a trend of 15,600°C per century is a scientific joke.
w.
No, it’s an instantaneous trend. A time rate of change of any quantity is a trend. Of course a longer duration would be more meaningful, but that has nothing to do with the units. It doesn’t make it more meaningful if you express it as 5e-6 C/sec.
You could equally say that the idea that a tennis ball would travel 100 miles in an hour is a joke.
So you’re claiming it’s meaningful to describe a one-week 3°C trend as a 15,600°C/year trend?
There’s a reason why your nickname is “Racehorse” Stokes.
I backed your true accusation that this was cherry-picking. Get smart, take the win, and move on …
w.
Nick,
You are trying to argue with people whose education and experience with varying functions far outweighs what you have.
There is no such thing as an instantaneous trend except for a linear equation of the form “y=mx+b”. For a constantly varying function, an ‘instantaneous trend” is an oxymoron.
Instantaneous in relation to weather or climate means a point in time where a derivative of the equation describing the changes will provide an instantaneous value at that point in time.
You can NOT trend that point in time because the next point, in the limit, will quite likely have a different value.
Be honest with this blog. What you are doing by extending the units of time of a measurement is assuming a linear function whose rate of change is constant. No one here believes the rate of change in the climate is constant.
The attached graph from UAH is an example of a function that reverts to a mean value over a period of time. It simply does not have a linear trend of constantly increasing temperatures. If, and notice the word “if”, temperatures follow past changes some time in the near future we will endure a move back toward the mean.
Too many CAGW advocates on here are statisticians with little if any familiarity with calculus.
The derivative is defined as
limit as x -> a of [ f(x) – f(a) ]/ (x-a)
As x -> a you are defining the slope of a point, not of a function between f(start) and f(end). The units of that derivative should be the same as that for f(x) and x. If x is being measured in units of time then the derivative only holds for one point in time. It is *not* a predictor of what the derivative will be at the next infintesimal increment on the x-axis.
They are numerically equivalent, but the implication is that to have meaning there must be an expectation that the rate of change will persist for approximately the longest period of time used. Only a student lacking the bigger picture will choose units that are inappropriate for the analysis or prediction being attempted. That is why we have invented units like nanoseconds or femtoseconds.
Scientific notation has long been accepted to be a number between 1 and 9, with associated decimal fraction, and an exponent multiplier, such as 3.14E00. Thus, an annual rate of change, which makes sense in the context of cycles that repeat as the Earth move around the sun, can be written as either 0.10°C/yr, or 1.0E-01°C/yr. I makes no sense to use millions of years unless one is perhaps talking about the average erosion rate of a mountain range.
To quote Einstein, “Make it as simple as possible, but not simpler.” An instantaneous rate of change should be shown in some natural small units such as centimeters per second, not kilometers per light year, when one is talking about the speed of a billiard ball, despite being able to convert any unit into another.
Another concern is that when presenting measurements there should be an associated 2-sigma uncertainty, which grows proportionately as the nominal value is multiplied. Thus, it becomes more awkward to mentally compare uncertainties when long time periods are used.
Yes, one can use any units they like, but it raises questions about their motives or understanding if inappropriate units are used.
As usual, clear and succinct!
“which grows proportionately as the nominal value is multiplied.”
And perhaps even *more* than proportionally.
In nature, a chaotic system, accurate projections require perfect knowledge. Accuracy typically degrades over time, just ask any meteorologist. Perhaps God can tell you what is going to happen in 100 years, Nick certainly can’t. He’s what karlomonte describes as a trendologist – trends continue forever. It’s the viewpoint of a statistician, not a physical scientist or engineer.
To paraphrase the Fiddler On The Roof: “Convention, convention, convention!”
I don’t think I have ever seen or felt a piece of ice that was “warm”.
No matter what month of the year it was.
Of course your ice cube will melt in your Bushmills if you’re a slow drinker, but then it’s just gone, not “warm”.
Syowa has warmed up since the coldest period in the last 100 years. It’s definitely time to start panicking.
Could I leave it a little longer to start panicking, please, because I’m a bit busy at the moment?
Mr Stokes, are your results from using NASA/GISS data or JMA which the author used?
Is there any difference in the values?
From below
I used unadjusted GHCN V4 data. It’s slightly different to the JMA mean. The difference seems to be that GHCN uses mean min/max (agrees with that calculated from JMA), but JMA’s monthly mean seems to be something else, perhaps average hourly values.
IOW, if you average JMA’s max and min, you get the GHCN mean, but not the JMA mean.
Are you unable to recognize the pattern here? Do you really think that a linear regression or a simple average would capture it correctly?
Has anyone checked that straight line fit on the graph ? Looks like a 1.5 degree rise since 1973 to me.
No, but it does show a discrepancy between SSTs and atmosphere temps down there. 40 years of cooling ,however slight, is noteworthy.
Cooling over July only (cherry pick?). Warming over the whole year.
Cooling over July, warming over the whole year in odd places and flat lining to slightly cooling over the whole continent = ?
So you’re claiming that the other 11 months it has warmed? C’mon, where’s your data?
I gave it here, using GHCN V4 data, where TAVG=(TMAX+TMIN)/2
JMA have same TMAX and TMIN, but calculate the mean somewhat differently.
Uh, oh. Where’s the JMA calculation?
July, being the middle of the southern hemisphere Winter, is not a ‘cherry pick.’ The theory behind radiative forcing of GHGs predicts that warming should be greatest at night and in the Winter. I’d personally prefer to see the whole year, but sometimes extracting an exemplar month, and de-seasonalizing the data, can provide some insights that might be difficult to pull out of a larger data set. It is just another analysis technique.
Speaking of cherry picking – I noticed that you completely failed to mention that, although sea ice had failed to fully form in some regions (due to unusual weather conditions) giving a below average result as a whole, in other regions they have actually exceeded the recent trend extent. Lying by omission, cherry picking, hiding the decline or anything that doesn’t fit your narrative? You’re worth everything they’re paying you.
On the other hand, if the temperature at Syowa had been going up, you would have been using it as an explanation as to why the ice was slow in forming.
As support for this, I point to your first response to the previous article about UAH global temperatures.
“As support for this”
And as refutation, the temperature in Syowa has been going up. And I didn’t.
This graph indicates a 40 year annual cooling trend.
I looked into that. I used unadjusted GHCN V4 data. It’s slightly different to the JMA mean. The difference seems to be that GHCN uses mean min/max (agrees with that calculated from JMA), but JMA’s monthly mean seems to be something else, perhaps average hourly values.
So Nick you “cherry picked” your dataset? You have one dataset showing very slight cooling and one showing very slight warming. My observation is nothing is really changing over the years. Y’all are arguing over nothing.
No you didn’t. You spend so much time gassing about things on WUWT that no one would have enough time to look at or study all the things you claim to have looked at or studied.
Does that not give you pause? Do you maybe think an hourly average may provide a better view of a “degree • day” value.
Actually, integrating temperatures from sunrise to sunset and sunset to sunrise would give a much better picture of temperature changes.
The data exists for 30 – 40 years that would allow this. Why has climate science not progressed to using a better measurement system?
It is like physicists refusing to use atomic time clocks for measurements because they had always used stop watches.
“It is like physicists refusing to use atomic time clocks for measurements because they had always used stop watches.”
Or pendulum clocks.
Nevermind, not worth my time responding.
“…temperatures there are well below freezing.” How cold was it? It was so cold the squirrels in Central Park were warming their nuts with a hairdryer. The only redeeming value of David Letterman were his squirrel comments.
The data in the graph closely matches NASA / GISS data for Syowa for July, also the GISS annual data eyeballing it shows no apparent warming for forty years. Warmest year since 1957 was 1989.
“no apparent warming for forty years”
For July. Other months were different, and the full year showed warming.
If you bothered to read what I said “the GISS annual data eyeballing it shows no apparent warming for forty years”. Or can you see warming where others can’t? I don’t believe you had time between my posting and your reply to calculate a trend.
I had already calculated and posted the trend.
I said forty years, not since 1973 or 1966.
Sorry typo 1980 not 1989 was the warmest year
If only the eunuchs claiming Antarctica is warming had the balls to tell the truth…..
I’m only surprised that Mr.Stokes doesn’t point out that the “miraculous recovery” of the Amazon Basin’s trees in the last six months since that naughty “far-right” Jair Bolsonaro was replaced by Jail Bird Lula da Silva is a demonstration that Antarctica (just like the Amazon) can only be saved by electing more Communists!
That’s what his photographs and data prove!
Nothing at all seems to be going on in Antarctica.
Trend is indistinguishable from zero.
Actually, my Mark IV calibrated eyeball suggests that the post-2000 temperatures are characterized by more low temperatures than the pre-2000 temps.
By a whisker. You’d have to be a statistician to make much of that.
There are 7 years where the low reaches or exceeds -1.25 deg C since 2000, while there are none before 2000.
It is a strange graph, with the years ending in zero being repeated.
Continuous, gap-free data from Syowa starts in February of 1966. Just looking at July is special pleading.
The trend since Feb ’66 looks like this:
The p-value is 0.233, not statistically significant.
However, that doesn’t account for autocorrelation. When adjusted for that, the trend looks like this:
There is no statistically significant trend in the Syowa temperature.
w.
In other words, the trend is indistinguishable from zero.
“Special pleading” depends on what one is trying to characterize. July does a better job of representing what is happening in Winter than does the whole year. Winter is when radiative forcing of GHGs is supposed to be most effective in raising temperatures.
“July does a better job of representing what is happening in Winter than does the whole year.”
The trend in July was -1.03 C/century. The trend in August was +4.69. In June, +2.14.
That is special pleading.
The question is whether models can exhibit this behaviour.
Nick, please stop being a special pleader. Anyone with more than 2 brain cells knows that obsessing over hundredth’s of a degree plus or minus in isolation is stupid. What have the weather patterns been like and what are they going to be like? Weather patterns can bring warmer or cooler air masses to a region that completely swamp the underlying temperature. If you continue with this obsession over hundredth’s of a degree in isolation without considering the weather patterns and meteorology of a region then you will be perpetually running in circles.
How much snow/ice is on the ground compared to 40 years ago – I see a lot of bare (low albedo) rock in that photo
Did the ice melt? Did it evaporate?. How has the annual total and seasonal pattern changed.
Did it simply never even fall out of the sky?
How has the size if the station changed in that time. I see quite some monstrosity there, it cannot always have been.
From what I’m learning of the giant freezer cold store, 3 miles away from my thermometer, that unless you are at least 5 miles away from a large man-made ‘temperaure controlled object’ – any thermometer inside that radius will be reading junk.
Is that the case here, at Ice Station Snowy?
We’re told that the sea-ice has changed.
Question arising: How has the colour of the water changed. Is it clearer or muddier than it was before?
Is it muddier or more turbid because of soil erosion elsewhere on the planet.
Such is why sea-level is rising after all.
i.e. Torrents of mud and dust storms coming off farmland
Plus ash, smoke and more dust from wildfires: whether deliberate, criminal, accidental, natural or preventative
Simple Archimedes Principle.
Is there more or less algae/slime/seaweed or other aquatic mush floating in there than previously?
IOW Has the albedo of the water changed since ‘the start of industrial times‘
What about the albedo of ice & snow ‘on the ground’ because of soot/smoke and ‘pollution’ coming from ‘Oop North’
That pollution will melt the ice but also feed the ocean – vast amounts of it were/are = Oxides of Nitrogen and Sulphur.
To include finely divided Iron – plus Calcium and Magnesium from cement manufacture plus the myriad trace elements released from the shale in that cement as that cement erodes and wears
All = truly epic fertilisers.
There are a Million Plus One Reasons why what is happening (not happening) there and ‘The Dancing Angels‘ & ‘Phlogiston‘ are none of them.
But no matter, Dancing Angels mean we can all pile into a cheap, common and drunken brawl
Today’s random wonderation:
What **is** the difference between ‘starvation‘ and ‘malnourishment‘
Reset button. Press your Reset button.
A few days ago on the UHI thread someone asked about truly rural site’s trends
The only UK site that the MO has data for and is pretty rural is Eskdalenuir in the Scottish Borders. I downloaded the data, some of which goes back to 2914.
I’ve done a bit of tinkering in Excel, not a Willis examination by any means, and so far only January and June.
As far as I can tell TMax in January shows no trend, nor did June until this year, although even that only just beat a year in the 1920s. Without looking at every month, which I might do if time permits, the only measure showing a warming trend is TMin.
The coldest decades appear to be the 1950s and 1960s. That seems right to me from my life experience. My mother always said it was warmer before the War than after Eskdalenuir seems to agree with her!
I could well be wrong, Stokesy can use the data to prove it if he likes. I’d be interested if he does.
Sorry typo, should be Eskdalemuir and the data is here
Second typo – should be 1914 not 2914. Not your day, today.
He should stick to “Bazball”…
Bazball been berry berry good to me.
So I’m totally confused. Back home in San Francisco we’re being told the earth is in crisis! We’re all gonna die and then this stubborn Antarctic thing comes along and rains on the parade. What is it? Is the Climate Crisis Planet Earth burning up or is it not?
Not. Not even close, not a sausage, nada, niet, nein.
Unless you happen to live in a city with plenty of UHI effect and are particularly unhinged in which case; yes we’re all burning here, it’s worse than we thought, it’s thermageddon.
Hmmm . . . from the graph in the above article, it seems like there was a tie for record-breaking warming in the years 1990 and 2006.
Funny thing is, I don’t remember such claims making MSM headlines back then. But perhaps the alarmist were looking the other way.
I think most of them were looking at the Middle East in 1990.
how to check a skeptic claim. red flags
here is what i have found to be universally true.
sceptics ALWAYS ALWAYS ALWAYS take a large dataset. Like worldwide historical records and select a small fraction to show.
they never show you this process of subsetting. or how the anser changes with every data cut.
cut number 1. from a record going back to 1957, they start at 1973. why.
cut number 2. from 12 months of the year they pick one. why. show all 12, show annual, show seasons.
cut number 3 from a globe they select antarctica
a. it has a negative greenhouse effect.
cut number 4. they select a coastal station rather than interior? why?
but we know the why, the sweetest cherries get picked.
proof? they dont show the process. they dont show the uncertainty in the selection methodology
full of tricks zone.
That’s because the global data set, taken as a whole, is meaningless. It starts off computing daily mid-range values from the stated min and max temperatures and then ignoring the measurement uncertainties associated with those stated values. All of a sudden the mid-range value is 100% accurate. no measurement uncertainty.
Then those mid-range values are combined into a monthly average with no propagation of uncertainty from the base temperatures. Again, the monthly average is assumed to be 100% accurate with no measurement uncertainty.
Then the same thing is done to compute annual averages and from there to compute baseline averages over a period of time -again with absolutely no propagation of measurement uncertainty from the base temperatures.
There isn’t even any upward propagation of the variance of the data from the base up through baseline average calculation. At each step where temperatures are combined the increase in the variance of the data is not calculated and propagated.
Then the climate science meme of “all measurement uncertainty is random and cancels” is applied and a shift to using the data variation as the uncertainty of the data is implemented. The data variation is at least as much from weather changes (daily, weekly, monthly, annually, decadily, etc) as from measurement uncertainty.
And then you have the problem of jamming NH temps with SH temps when the variances can be wildly different because of the tilt of the earth.
And YOU* want to complain about skeptics not worrying about sampling error when looking at the data from specific regions or locales? The sampling error is miniscule when compared to the errors in he global data base taken as a whole!
Anser? All lower case? Are you drunk typing?
Steven Mosher August 3, 2023 11:22 am
Mosh, here’s something which is not “universally true” but is almost always true:
Absolute all-inclusive statements like you made, about “everyone” and “ALWAYS” or “NEVER”, are almost always garbage. They are also very easy to falsify since all you need is one example to show they’re wrong.
Mosh, you REALLY need to dial it back. Your statements are becoming quite funny … but I doubt greatly that people laughing uproariously at your nonsense is the response you’re looking for.
w.
PS—A VERY large number of mainstream climate studies are about specific locations or classes of areas. Like say, the NINO3.4 zone … but according to you, that’s some kind of evil cherry picking, and they should be looking at the oceanwide data.
Do you realize how foolish that claim of yours is?
‘Recent claims that the ice is melting there are absurd.’
Of course the claim is absurd – actually, the ice is boiling!
OK, I’m calling triple cherry-picking on this post.
First, it’s cherry-picking to use just one month’s trends to claim that the station has been slightly cooling, when overall using all months it has been slightly warming.
Second, it’s cherry-picking to start the graph in 1973 when continuous data goes back to 1966 … especially when it’s been warming since ’66 and cooling since ’73
Third, it’s cherry-picking to claim cooling without pointing out that the “cooling” is far, far from significant. The p-value of the post-73 trend is a whopping 0.70.
Here’s the evidence:
The red/black line is post ’73, the orange/black line is the full dataset.
w.
Remember how July 2023 was the hottest month in the history of history? Well, someone forgot to tell my utility company which informed me that July 2023 was one degree COOLER than July 2022.
Hmmm.
Yeah, but that’s only where you live. There was a spot on the north coast of Svalbard that was a full degree and a tenth WARMER (-20.2 instead of -21.1) that cancels you out. My thermometer in Canada told me.