Water Behaving Badly

Guest Post by Willis Eschenbach

Those who know me are aware that I’m a tropical boy, a hopeless addict of warm blue seas and coconut-laden islands with white sand beaches. Here’s where I used to live and work, Liapari Island in the Solomon Islands.

That is how I like my water to behave, soft, warm, and inviting. But when the ice jumps out of my tropical-type adult beverage and starts running around the countryside covering everything in white and floating in giant chunks all over the ocean, well, I call that “water behaving badly”. 

And as you might imagine, other than brief visits I tend to avoid places where water behaves badly.

However, thinking about such icy climes when I’m someplace warm, that’s a more pleasant matter. So I got to considering ice, in particular, sea ice. And as is my wont, when I consider something I go get the longest dataset that I can find. In this case, that was the HadCRUT ice and sea surface temperature dataset. It claims to go back to 1870 … but that doesn’t mean that it’s good back to 1870. Figure 1 shows why. 

(As an aside, Figure 1 also shows the importance of starting by running the old Mark I Eyeball over your data before subjecting it to any mathematical gymnastics … but I digress.)

Figure 1. Hadley Centre HadISST Ice data, 1870-2019.  Top panel shows global data, middle panel shows southern hemisphere data, and the bottom panel shows northern hemisphere data. Note that the three panels are all at different scales. Units in this and all graphs are percentages of the total surface area of the globe.

Notice the very regular signal in the early days of both the northern and southern hemisphere data, and as a result in the global data. This is obviously just a perfectly regular repeating signal added to the later actual observations. We can get a better idea of when the real observations start (and stop) by removing the regular repeating seasonal signal from our dataset. Figure 2 below is the same as Figure 1, but with the regular repeating average seasonal variation removed.

Figure 2. Global and hemispheric ice areas, HadISST Ice data. Note that the three panels are all at different scales. Units in this and all graphs are percentages of the total surface area of the globe.

The regular signal in the earlier parts of the record is an artifact. It is an interference pattern resulting from the removal of the seasonal signal. Only the latter part of the datasets contain valid observations.

In Figure 2 above we can see that Arctic measurements (northern hemisphere, blue above) are only good since about 1960. Note the odd lack of data (with missing data replaced by a regular signal) from about 1940 to 1952.

Antarctic ice area (southern hemisphere, red above) actual measurements are more recent. The Antarctic record is only good since 1973. As a result, we can only look at global data since 1973. However, that’s approaching a half-century, so it is still of interest. Here’s the global ice area since 1973, the period where we have actual observations. It’s worth noting that since 1979 we have full satellite observations of ice areas.

Figure 3. Global ice areas, HadISST Ice data, Jan 1973 to July 2019. Units in this and all graphs are percentages of the total surface area of the globe.

Now, I was surprised by Figure 3. Surprise is the very best part of science to me. I love the first sight of the graphics, turning what before was just a bunch of numbers into a record of the past.

There were a couple of surprises in Figure 3. First, from 1980 through 2004, a quarter-century during which there was general global warming, there was no trend at all in global ice area. None. Well, to be accurate, the trend 1980 through 2004 is -0.0000000000000001% per decade … and as you imagine, not statistically significant.

After 2005 the global ice area went down, but by 2010 it had recovered. From there to 2015, it was above average. And since 2015 global ice area has dropped precipitously but then recovered back to average. Finally, there is no statistically significant trend in the full 1973 – 2019 dataset.

So … lots of things of interest in Figure 3. However, I gotta say, I’m not seeing the evil hand of steadily increasing atmospheric CO2 in that record. Nor am I seeing any “anthropogenic fingerprint”. Perhaps most importantly, I am unable to detect any sign of any “climate emergency” in that record.

The final surprise was the recent several-year deep drop and then recovery of the ice area. I figured it must be from what alarmists have termed the “Arctic death spiral”, the widely trumpeted decrease in Arctic sea ice. So I added the separate Arctic and the Antarctic records to Figure 3 above. Figure 4 below shows those records.

Curiously, the amount of ice at the two poles is just about the same, at ~2% of the globe. But that makes it hard to compare the Arctic and Antarctic ice. So in Figure 4 below, I’ve offset the northern hemisphere (blue line) by 1% for clarity. You’ll need to add 1% to the northern hemisphere ice areas to get the actual values. Figure 4 shows the globe as well as the two halves of the planet separately. Note that in this graphic they are all to the same scale.

Figure 4. Global, northern hemisphere (Arctic) and southern hemisphere (Antarctic) ice areas. Northern hemisphere values have been offset downwards by 1% for clarity. Add 1% to the values shown to get the actual values. Units in this and all graphs are percentages of the total surface area of the globe

And for my final surprise, it turns out that the recent variations in global ice area are largely the result of variations in the Antarctic ice area, and not in the Arctic ice area that we spend so many electrons discussing …

So what I found out regarding the global ice areas was that I didn’t know all that much about global ice areas … and speaking of which, just what the heck did cause the drop and subsequent recovery in Antarctic sea ice area from 2015 to the present?

Here on the north coast of California, it’s the leading edge of autumn. We had our first rain this week, which left the forest full of the damp dark green smell of life, decay, and rebirth. And when I just looked outside, the rain had come again. What a joy it is to investigate the mysteries of this endless universe, even the vagaries of water behaving badly!

Best to everyone,


0 0 vote
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
Ian Magness
September 30, 2019 10:18 am

Thanks as ever for another very interesting article Willis.
Re the Southern Hemisphere figures, it seems to me that:
– the whole 2015 and onward period measurements seems to have an annual variability too far beyond previous figures to be consistent. Did the method of measurement change around 2015 or have the numbers been subject to the Adjustocene?
– even given the anomalous annual variability noted above, 2016 just looks like a spurious data point, unless, of course, it was subject to the aforementioned Adjustocene fun. I suppose volcanic activity might, theoretically present an answer but I am not aware of such on the required significant scale.
Any thoughts Willis?

Daniel Bryce
Reply to  Ian Magness
October 1, 2019 6:33 am

Larson C and Pine Island Ice Shelves broke off 2015-2017.

Nice to see the total ice recovering. For funsies, google doomsday articles from 2017.

Interesting to see the slow trend toward less ice in the NH, and a mirror increasing trend in the SH.

September 30, 2019 10:21 am
Reply to  Rob
September 30, 2019 10:58 am

Rob why is every weather event a “record” now? A storm is hitting england tomorrow, and the news papers are already scaring people with “record breaking winds” 😐

Bill Powers
Reply to  Sunny
September 30, 2019 1:06 pm

Fear and guilt Sunny. You control the populous with Fear and Guilt. They are taking their lead from the Catholic church which has controlled their congregates with fear and guilt.

On thing the Government/Education/Media troika need to understand is that that for the past 50 years the Catholic Church has been hemorrhaging congregates.

If they keep this Stormpocalypse nonsense up for a couple more decades and these children they have indoctrinated are going to wake up and realize what they are experiencing is weather and the changing seasons. The world is not coming to an end and they are going to revolt against this propaganda if they don’t lose their ever loving minds like Greta first.

Reply to  Sunny
September 30, 2019 1:50 pm


I trust you will be hiding under the table tomorrow? . Of course we here in the Uk have never known a storm before and heavy rain is something we have only experienced once or twice in our lifetime.


Reply to  Tonyb
September 30, 2019 1:54 pm

Tonyb Lol if we went with out rain for more then a few days then I would worry, the wind has been constantly blowing all year though. I remember my first storm at brighton pier, the wind was epic, I loved it…

Smart Rock
Reply to  Sunny
September 30, 2019 5:51 pm

Sunny: You are going to be blessed with Hurricane Lorenzo, which started out (as the nearly always do) headed towards the Caribbean, then did a 90-degree right hand turn and is now heading almost straight for Ireland and Scotland. As of today, it looks as though England will miss most of the fun. You can watch its progress at https://www.nhc.noaa.gov

Of course it’s going to be a record storm. The media make that decision well in advance of the event itself, making sure that the message is deeply embedded in the popular consciousness. Will they tell us afterwards “that wasn’t as bad as we expected”? Hmmm, I wonder /s

Reply to  Smart Rock
October 1, 2019 2:19 am

It looks like the western Azores are going to be hit. The birdwatchers who go there this time of year in order to get American vagrants on their Palearctic tick list will have a field day as soon as it has passed. A lot of confused migrants will probably be blown in.

Yes, there actually are people spending long periods in fall sitting around on the westernmost Azores for this purpose.

September 30, 2019 10:22 am

Do you know the Historical Ice Chart Archive ?

Reply to  Krishna Gans
September 30, 2019 11:12 am


Clyde Spencer
September 30, 2019 10:24 am

Figures 3 and 4 are strong anecdotal evidence that something other than CO2 is affecting sea ice area. That is, one would expect ice area to have a negative correlation with CO2 concentration, were it responsible for melting ice. Indeed, SOMETHING, which overpowers the assumed effects of CO2, caused a significant decline in Antarctic sea ice around 2016. Submarine volcanism?

Reply to  Willis Eschenbach
September 30, 2019 11:30 am

Excellent data view!

I wonder why there is such an effort to consider polar sea ice as distinct from total polar ice, including land based ice?

Given how recent Arctic history demonstrated how much storms detrimentally affect Arctic sea ice, I’m more surprised that both poles show such regular sea ice extents.

There always is a possibility that satellites and their algorithms have issues. “Satellite sensor errors cause data outage”

Bryan A
Reply to  ATheoK
September 30, 2019 12:18 pm

Eastern Ca. (Sierras) just had a “Dusting” ( 4-5″ at the summit) of White Water.

Geo Rubik
Reply to  Willis Eschenbach
September 30, 2019 12:02 pm

“Finally, over most of the record, the decrease of ice area in the Arctic is almost exactly matched by the increase in ice area in the Antarctic … coincidence?”

I remember being taught this back in grade school (1960’s). When one increases the other decreases. Or I’m getting old and don’t remember well.

Reply to  Willis Eschenbach
September 30, 2019 2:04 pm

However, I would expect it to respond to changes in temperature including SST. But globally, we’re not seeing that either.

Is this the giveaway that temperature records have been hopelessly, shamelessly adjusted?

Reply to  Willis Eschenbach
September 30, 2019 3:22 pm

Willis, isn’t sea level rise (adjusted for landform subsidence) linearly coupled to global ice melt (sea and land)? And if sea level rise is very slow and steady (evident during human observation) doesn’t that indicate global ice melt (in aggregate) has been slow and steady as well? The wild swings and disparities between poles suggest that sea ice alone is a poor proxy for global temperature resulting sea level.

Reply to  Willis Eschenbach
October 1, 2019 8:22 am

thank you Willis, for stating the obvious about sea ice.

Reply to  Willis Eschenbach
October 2, 2019 6:13 am

Hi Willis,

mind to honestly object. What is result of ice sheet pressure on Earth crust? If you suddenly get rid of all ice from Antarctica, Antarctic crust would immediately surge up. Uncovering beautiful Antarctic beaches…
Water from around those would go somewhere else increasing sea level.
Earth crust is floating on mantle, so any weight above has impact on surface level e.g. seafloor.

Greetings from my boat making repairs after 30th September storm around 120km/h, salvaging my outboard which was partially torn from boat and remained jumping on mooring line kept in tension by wind just touching water and changing broken chain link from backstay, which would for sure give up next stronger wind.
All this imprisoned in landlocked country waiting for first bad water come and force us to dry dock all boats.

Reply to  Willis Eschenbach
October 3, 2019 12:13 am

It is Slovakia now.
But I lived 4 years between 2012 and 2016 in Bay Area, Mountain View. I did my boating course there in Berkeley and spent many hours on Bay.

Reply to  Willis Eschenbach
September 30, 2019 3:41 pm

Willis Eschenbach
September 30, 2019 at 10:55 am

Forwarding a wild guess here.

The signal patterns that you mostly describe here, have to do and have to rely in consideration of precipitation.

Such signals, according to the wild guess offered here, will be overwhelmed by the main long term climate signal, unless that climate signal happens to be a plain one… like in the case of climate stagnation, kinda of no warming or cooling in the consideration of the long term climate.

Unless the Mann’ s hockey stick considered as valid, then if carefully looking at the data, for the last ~2000 years, including LIA, there is no any clear up or down on that 2000 year period trend there.

So in the consideration of such a wild claim, if such small signals can be still observed clearly enough in the data of the last 150 years, then no difficult to consider that despite the warming during this time period still the climatic long term main signal remains a plain one, so the main climate signal and it’s long trend still in the condition of stagnation… no in a warming or cooling trend either, yet.

Considering a clear observation of the effect of thermal ingredient in regard of the surface being water one in one case and land in the other, and in a quasi equal quantitative variable values and with an opposite sign, may push for a consideration that it is plausible enough to contemplate that happens to be possible as a result of a the lack of climatic long trend direction… as in a climatic stagnation.

Oh well, just a thought, first that came to mind.


Reply to  Willis Eschenbach
October 2, 2019 8:09 am

A possible reason the Northern and Southern hemispheres have nearly identical areas is that the earth is a pretty symmetrical globe. The climate transfers energy all around the globe all the time. Any change in the ice caps would likely be symmetrical also. Given the relatively small range(in physical terms) of global temperatures any major is quickly muted.

Reply to  Clyde Spencer
September 30, 2019 11:02 am

Or possibly something happening with ocean currents, such as a slowdown or overturning of the Gulf Stream or “Atlantic Conveyor”?

Whatever it was, it was very short term.

Bryan A
Reply to  Duane
September 30, 2019 12:19 pm

It was Trenberth’s missing heat finally coming to light.
There one minute and gone the next

Reply to  Clyde Spencer
September 30, 2019 12:24 pm

There was also the strong 2016 El Niño event.

Richard M
Reply to  Renee
September 30, 2019 12:51 pm

I agree, the 2016 super El Nino was the culprit. A little of that warm water found it’s way south and since SH ice is located closer to the tropics it wasn’t that far.

Once that ended the ice recovered back towards its average.

Richard M
Reply to  Willis Eschenbach
September 30, 2019 1:47 pm

Willis, different types of El Nino events (regular vs. semi-Modoki). Also no Bjerknes feedback in 2016.

Of course, there are other teleconnections as well in winds and ocean currents. It could be the El Nino blocked some currents which led to upwelling of warmer water in Antarctica.

Not sure we will ever know for sure.

Reply to  Willis Eschenbach
September 30, 2019 2:34 pm

Maybe the Big Blob forced the El Nino waters southwards?

Reply to  Willis Eschenbach
September 30, 2019 2:48 pm

Multiple oceanic and atmospheric processes were probably involved in 2016. This paper discusses several events that occurred such as strong northerly atmospheric flow causing rapid ice retreat in the Weddell Sea, 2) an unusually negative southern annular mode in 2016 causing rapid ice retreat in the Ross Sea, 3) the extreme El Niño that peaked in 2016, contributing to unusually warm ocean waters in the Bellingshausen, Amundsen, and eastern Ross Seas, 4) a persistent zonal wave 3 atmospheric circulation around Antarctica contributing to reduced sea ice extents in the Indian Ocean, Ross Sea, Bellingshausen Sea, and western Weddell Sea, and 5) a weakened polar stratospheric vortex. Not sure if those same conditions existed in 1998. https://www.pnas.org/content/116/29/14414

Reply to  Willis Eschenbach
September 30, 2019 10:48 pm

I agree that 2016 was a climate extreme event for Antarctic melt, perhaps the IPCC can use as an example case. The NSIDC data does not show full recovery. However, it does look like it has stabilized. Interestingly, sea ice stratigraphy is like geologic layers. Stable correlative phases interrupted by brief catastrophic events.

Reply to  Willis Eschenbach
October 1, 2019 12:23 am

Willis, the negative feedback you are looking for could be the “thinner ice grows faster than thicker effect”, see https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2018GL079223 .
It impacts the growing season. perhaps you can show with the data (HadISST1 or NSIDC dataset) that after a sharp decline of sea ice coverage in the melting season the sea ice grows much faster in the following freeze season. In the paper they mention this negative feedback working against the ice albedo feedback which must be overestimated, this show your figures. No “death spiral” when ice begins to decline as we could see it in 2016 in the SH and 2012 in the NH.

September 30, 2019 10:27 am


As a fellow researcher, I have a great admiration for your work and information you provide. Needless to say, I keep copies.


Mike Smith
September 30, 2019 10:30 am

Wow, another humdinger of an observation. Thanks Willis.

Certainly interested to hear any theories regarding that 2015 “glitch”.

Reply to  Mike Smith
September 30, 2019 10:58 am

I was wondering that myself, even when the Antarctic sea ice was rising faster than usual, and then dropping precipitously. Has anyone owned up to an equipment malfunction, operator error or use of some dodgy new algorithms?

Curious George
September 30, 2019 10:34 am

Interesting as always, thank you Willis. Just curious: you talk about the Arctic death spiral, but Figure 4 assigns it to the Antarctic – or perhaps something is mislabeled?

Robert W Turner
September 30, 2019 10:37 am

Changes in the >15% ice cover algorithm and changes in the satellites themselves seem to account for when the sea ice “declines”.

Reply to  Willis Eschenbach
September 30, 2019 11:25 am

I was under the impression that the 15% rule applies to ice extent, not ice area. Area is the actual percent of each grid cell that is covered with ice. For example, 50% ice on a 625 sq km grid cell would be ~310 sq km in area, but 625 sq km in extent.

Reply to  Robert W Turner
September 30, 2019 11:16 am

The chart above is labeled Area, which does not follow the 15 % criteria for Extent. I’m curious how things were done pre-1980: passive microwave or satellite photos?

Reply to  Willis Eschenbach
September 30, 2019 1:43 pm

By saying “ice cover percentage” do you mean concentration, i.e., 0-100%?

September 30, 2019 10:48 am

comment image 10 March 2016: A detailed analysis of HadISST has identified a number of problems, some of which might limit the utility of the dataset for certain scientific applications, particularly where high resolution SST data or spatial gradients of SST are required. Thanks to Dudley Chelton and Craig Risien for bringing this to our attention and their detailed analysis.

Your Figures 2 and 4 clearly show global warming. We know the CWP is due to human -influence i.e. AGW. The Arctic and Antarctic cannot be compared with each other as the former is surrounded by land and the latter is open sea/ocean. There is virtually no multiseasonal sea ice in Antarctica while the Arctic used to have up to 8 years old and now down to a few years which cracked and broke up for the first time in summer 2019.

I suggest you read the following description of why Antarctic sea ice has been below trend since 2016:


Reply to  Rob
September 30, 2019 11:44 am

Prior to the 1980’s there was an ice buildup during the “ice age cometh” decade according to eye witness historical accounts. Glaciers all over the globe reached century+ peaks then…and Arctic Ice expanded then as well from low ice levels in the 1920-1940 era. There are historical records of low Arctic Ice area and unprecedented passages by northern Mariners all over the place before 1940…as well as abundant historical records of far less open water during the cold decade.

So it is not known how OLD the ice was back 100 years ago…or even 60 years ago…so nobody knows if current ice content (by age) is unprecedented.

If the data was worth a damn prior to the satellite era, Willis would have presented it…BUT PRIOR TO 1950, IT WAS CRAP. And Willis kindly took the time to show us how totally worthless most of the sea ice record is.

Reply to  DocSiders
October 4, 2019 12:29 pm

Hi Doc,

what gave you the (wrong) idea that galicers allover the world reached in the 1980ies century+ peaks?
As a matter of fact they did gain and lenth gains are reported from teh European alps in the low meters per anum. The losses in the 90ies (contiuning until today) were often above 10 m range per year.


John Dilks
Reply to  Rob
September 30, 2019 5:04 pm

” We know the CWP is due to human -influence i.e. AGW.”

“We” don’t know any such thing. You assume it. I disagree.

Jean Parisot
Reply to  Rob
September 30, 2019 5:21 pm

” We know the CWP is due to human -influence i.e. AGW.”

Spoken ex-cathedra?

Martin Cropp
Reply to  Rob
September 30, 2019 6:13 pm

The report you cite is mainly correct. Also refer to the article identified of the bottom of your reference.
Most importantly this chart here

Antarctic SIE progressively increased to a peak area in 2014, the peak being achieved as winds dropped at the critical time, allowing cold air to flow down off the continent, cool and freeze the calmed surface. Winds are the number one enemy of of sea ice. Warmer SST will soften, but winds move and break up.

The decrease for 2016 started in September 2015 and continued through into 2016, as the image in the second link above identifies. at December 31st 2014 SIE (NSIDC) was at 9.50 msqkm, whereas 2015 was 7.22 msqkm.
The culprit was El Nino 2015. The true thermal impact on the 2 meter temperature level was negated by surface to high troposphere transport by the second largest East Pacific ACE since 1982. Coupled to that increased volumes of atmosphere traveled southward at various altitudes, surface through Brewer Dobson Cycle. Hence we had one of the largest and long lasting ozone depletion periods over Antarctica, via sustained SH Zonal winds, colder temperatures at altitude and therefore ozone dilution (a longer story), see links below.

In summary the 2016 reduction started in 2015 and did not recover, ending the year at 31st December 2016 at 5.70 msqkm. Most things are connected.
With regards


Martin Cropp
Reply to  Martin Cropp
September 30, 2019 7:10 pm

The two top links are identical. Open the article and expand the first image to show the whole chart.

September 30, 2019 10:49 am

“just what the heck did cause the drop and subsequent recovery in Antarctic sea ice area from 2015 to the present?”

Under sea/ice volcanic activity.?

September 30, 2019 10:51 am

The graph above appears to show a different pre-1980 trend in Arctic ice than this graph does.

comment image

Reply to  icisil
September 30, 2019 11:25 am

Is Tony Heller aka Steve Goddard still banned from WUWT for his infamous CO₂ condenses out over Antarctica? His work is not creditable and you should refrain from using it as support or a citation as that opens you to ridicule for being inept.

Reply to  Rob
September 30, 2019 12:09 pm

I don’t pay any attention to people who talk to me like you do.

Reply to  icisil
September 30, 2019 12:19 pm

It is an interesting chart, but the Vinnikov paper doesn’t appear to have direct data, they must be proxy based, since any year before 1973 will be spotty observational (regional, local) data, thus not comparable to each other.

Reply to  icisil
September 30, 2019 12:53 pm

That is your loss and why you never grow or be a scientist but remain a rookie a drift at sea.

Reply to  Rob
September 30, 2019 1:44 pm

Hubris is your name

Reply to  icisil
September 30, 2019 4:23 pm

You sound like Tony! 🙂

Reply to  Ric Werme
September 30, 2019 5:22 pm

LOL. Yeah that does, doesn’t it.

Reply to  Rob
September 30, 2019 1:24 pm

I am setting a mission for myself to find one wrong thing you said once, and use it to destroy your credibility, absolutely and utterly, now and forever. Should not take long.
Cancel Culture — Your Rules.

As a side note:
In the political sphere, intolerance has mutated into the Politics of Personal Destruction, which now seems to be acceptable.
Unfortunately, personal attacks are still not allowed here at WUWT, lucky you.

Reply to  TonyL
September 30, 2019 1:37 pm

Your ire and umbrage are misplaced. Can you answer the question? I rarely visit this site as it doesn’t change. However, my professional ethics will not let me pass up on schooling commentators or calling them out. I spent a long time being an activist against fascism and eventually prevailed. I sad to see it now rising in the US under the current administration.

FYI: I can refer you to climate scientists across the spectrum who pan Heller’s misleading and purposely deceptive work from Mann to Curry. Dr. Curry, a skeptic climate scientist writes on her blog that Heller/Goddard’s analysis “bogus” and “highly problematic”.

Reply to  Rob
September 30, 2019 2:13 pm

If you think the current administration is fascist, you’ve been fighting against the wrong thing.

Reply to  Rob
September 30, 2019 2:20 pm

Here, I’m calling you out. Address the graph immediately below.

And btw, if you think the current administration is fascist, you don’t even know what fascism is.

Reply to  Rob
September 30, 2019 4:38 pm

Lol you wrote “fascism.” Are those the talking points now? What happened to racist, bigot, homophobes, misogynist….?

Reply to  Rob
September 30, 2019 4:52 pm

However, my professional ethics will not let me pass up on schooling commentators or calling them out.

Do your professional ethics extend to the incompetence or corruption within the professional Climate Science community? Just asking for a Friend.

I spent a long time being an activist against fascism and eventually prevailed.

My undying gratitude!
You must have served with one of:
Gen. Omar Bradley: First Army.
Gen George Patton: Third Army.
Gen William Simpson: Ninth Army
Or maybe you served with the British under Field Marshal Montgomery, or maybe even with the Russians, the Free French, or the Polish Resistance.

I grew up at a time and a place where I was surrounded by adults who were combat veterans of WWII, many in Europe, some from the Pacific. I saw up close and personal, the price payed. It was a small town, the WW II memorial listed names from families I knew. The loved ones of those who did not come back.

If you did not fight in that war, then you did not fight fascism, what you are engaging in is called “Stolen Valor”.
Despicable. Arrogant. Self Important. Reprehensible.
We see who and what you are. Go Away Now. And take your self-righteous “professional ethics” with you.

Reply to  Rob
September 30, 2019 1:32 pm

That is one way of dealing with inconvenient facts.

How about the graph he has shown, with data from NOAA itself?

comment image


Also too inconvenient?

Reply to  Rob
September 30, 2019 1:44 pm

Try debunking this one. All the graphs, from sea ice, to see level, to whatever, are prepared to cause alarm. He shows you exactly how. Too inconvenient for you?

Reply to  Rob
September 30, 2019 1:58 pm

Well Rob if you are so sure of yourself why not go over to his site and offer to debate him. I’m sure with your sharp wit and rapier tongue you would be able to out do a non-credible person such as him.

But since you can’t answer 3 simple questions probably not.

Reply to  mkelly
September 30, 2019 2:08 pm

mkelly, he can’t since Tony Heller has the following:

BS Geology, Arizona State University
Masters Electrical Engineering, Rice University
Boston University Geology
Northern Arizona University Computer Science
Colorado State University Computer Science
University of New Mexico Geochemistry


that is why the snob will avoid that place.

Reply to  Sunsettommy
September 30, 2019 6:42 pm

Heller has been to college that is more than most at WUWT but he still manages to get everything wrong. His list of written apologies might fill a book by now. Not one piece of his work has been published in the peer-review climate science journals of standing. He is busy debating potholer 54 on YouTube and getting his clock thoroughly cleaned in the process. Even Mockton ran from his debate with Potholer54. Why?

Reply to  Rob
September 30, 2019 2:43 pm

The fact that the Antarctic reaches temperatures up to 10ºC BELOW the sublimation point of CO2 doesn’t seem to worry your lack of science.

Reply to  fred250
September 30, 2019 4:35 pm

Perhaps you could critique the discussion back then and learn learn from it instead of making unsubstantiated claims. If you can’t find it, I’ll post URLs for you. Should be an easy search. I’d rather I not scratch the scars.

I take you don’t live where it gets cold in the winter and have a car parked outside.

There have been many days when I’ve gone outside in the morning with a temperature of 20F or so. Some days I had to scrape frost off my windshield. Some days there wasn’t frost. Current CO2 levels are similar to days without frost.

Reply to  Rob
September 30, 2019 4:22 pm

I’m not really part of banishment discussions, but I was one of the principals in the CO2 frost “discussions.” Still a bit of a sore point with me, and apparently an even more sore point with Heller.

I think Tony and Anthony pretty much avoid each other, it works out better that way. I posted a couple times on Tony’s blog, and won a new post from Tony to point out what an idiot I am.

His historical newspaper and magazine article finds are useful, but the repetition and some of his analyses help keep me away.

Reply to  Ric Werme
September 30, 2019 6:51 pm

[Snip – see this from yesterday

[ This is what you wrote: “As a white Afrikaner/Hollander, you already carry the bias of being a vile christian supremacist and racist being a supporter of apartheid.”

You used the word racist in a derogatory way. There’s no dispute.

You are back on a time out, 96 hours this time. There will not be any further times out after this if you continue to call people names, and insult them in other ways. Clean up you act or you will be permanently banned – mod]

[Since then we have discovered two other aliases going back in time that you have used here from the same IP address. This is sockpuppeting on your part, and a clear violation of policy. You have multiple policy violations. You’ve been warned multiple times. Comparing notes with all moderators, it has become clear that you have no interest in adhering to policy. Therefore the decision has been made to permanently ban you for multiple policy violations. -mod]

Reply to  Rob
October 1, 2019 3:47 pm

So, why are you posting this here to me? Steven Goddard never said that.

Ah, https://wattsupwiththat.com/2019/09/28/global-climate-intelligence-group-founded/#comment-2810176

What aliases have you been using, and why are you an anonymous coward?

BTW, by the time Anthony bans someone, very few people left are sad to see the twit go.

Rick C PE
September 30, 2019 10:53 am

Willis: Great work as usual.

Might the drop off in Antarctic sea ice after 2015 be related to the much ballyhooed chunks breaking off the Ross and Larson ice shelves in 2016/17?

Oh, and why don’t the HADCrut folks just admit they have no real observation data before about 1950?

Reply to  Rick C PE
September 30, 2019 11:49 am


Scientists seem to find it hard to say they don’t know. Interpolation can artificially replace lack of data, but lack of data can’t replace lack of data.


Len Werner
September 30, 2019 10:54 am

A nice demonstration of what used to be a requirement of statistical analysis–that Step Number One was a ‘Critical Review of the Data’. There is plenty of evidence here that the accumulated data holds as much or more artifact of the system of measurement than of the actual parameter being measured. As a geologist I have no choice but to consider a data set existing in reliable (?) form from 1973 to present as relatively insignificant to a climate system over 4 billion years old, I studied that which was under and before Pleistocene ice, despite spending a lot of time walking across that ice to get to the rocks; they record climate too.

A very revealing treatment of dubious data; the exposure of the blatant uselessness of pre-1973 data is brilliant. And what are you doing still sitting in northern California; it’s snowing like stink just a few hundred miles away.

Reply to  Gunga Din
September 30, 2019 12:30 pm

That was my guess, that volcanic activity in Antarctica might be the cause. I also notice that when southern hemisphere ice levels trend up that southern hemisphere ice trends down and vice versa.

Reply to  Ryan Welch
September 30, 2019 1:13 pm

Heat capacity of ice is 2.09 kJ⋅(kg⋅K)⁻¹, the temperature of the interior of Antarctica is ~-60°C, so the energy to raise the temperature to 0°C is ~125 kJ⋅kg⁻¹, versus the heat of fusion which is 333.55 kJ⋅kg⁻¹. So yes, the energy to bring this temperature of ice to 0°C is a fraction of the heat of fusion. “The minimum average heat flow beneath Thwaites Glacier is 114⋅mW⋅m⁻² with some areas giving off 200 mW⋅m⁻² or more, the researchers report today (June 9) in the journal Proceedings of the National Academy of Sciences.” Now, compare that with the 2,000 – 3,000 mW⋅m⁻² that the human CO₂ increase adds. And don’t forget that CO₂ is adding that much everywhere, not just a few small areas in West Antarctica. Which matters more? Any guesses?

Reply to  Rob
September 30, 2019 2:08 pm

If your reasoning were true, we would be seeing similar melting in E Antarctica. We aren’t. I suspect that a small magma hotspot underneath the ice would add much, much more than 200, or even 3000, mW/m^2. The figures you gave for beneath Thwaites were gathered during a 50-year survey (who knows how long ago); there’s no telling what the fluxes are now; we simply have no way of knowing. Even so, can you guarantee that the magnetic flux measuring technology used for that survey had the resolution to detect very small hotspots beneath kilometers of ice?

Clyde Spencer
Reply to  icisil
September 30, 2019 9:01 pm

The rates may not be uniform. They probably are related to volcanoes we can actually observe — eruptions.

Aerial or spatial resolution should be governed by the inverse square law. That is, the thicker the ice, the lower the resolution for detection and characterization.

Richard Moore
September 30, 2019 11:08 am

The first figure seen on the WUWT Sea Ice information site is the plot by Professor Ole Humlum of total, Arctic and Antarctic ice area going back to 1979. It depicts the seasonal variation in smooth sine wave like curves and shows the recent decline in Antarctic ice area.
I do not see a recovery of Antarctic ice area in his figure.
comment image?ssl=1

Phil Rae
September 30, 2019 11:19 am

As always, impressed by your natural curiosity & your ability to tease sensible explanations out of apparently mundane observations. Always learning from you, sir! Thanks.

September 30, 2019 11:36 am

So do I understand, Will Eschenbach, that you are saying, essentially, that the ice in both polar regions is stable, with seasonal variations taken into account?

And basically, we hoomans really don’t have much of an effect on it?

Just askin’. I’m wading through Wilson’s 14 Points thingy and getting brain fry from it.

Gunga Din
Reply to  Sara
September 30, 2019 1:32 pm

Sara, don’t you “brain freeze”? 😎

Reply to  Gunga Din
September 30, 2019 3:20 pm

Well, considering that Woodrow Wilson was a blatant socialist and wanted a one-world government, hence his proposal for the League of Nations (guess what its descendant is!), it’s mostly brain fry rather than freeze.
Thank you for asking!

Gunga Din
Reply to  Sara
September 30, 2019 4:23 pm

One of my coworkers bought a copy of an old Time magazine a few years ago . He brought it into work.
He was more interested in the price of new cars in the ads.
I was interested in the International Section. No mention of Germany. But under the League of Nations, they did mention that radio reports from a Dr. said that the the Japanese had landed a force was moving toward Nanking. The Japanese rep said they had only landed supplies (or something like that.)
The “Rape of Nanking” began about a week later.
(It was … weird … reading the “Pre-UN” “diplomatic” comments before an atrrocity that happened a week or so later.)

Latus Dextro
Reply to  Willis Eschenbach
September 30, 2019 2:17 pm

GLOBAL TEMPERATURES Philip J. Energy & Environment · Vol. 26, No. 3, 2015
(truncated abstract)
“Holocene records up to 8000 years before present, from several ice cores were examined. The differences in temperatures between all records which are approximately a century apart were determined, after any trends in the data had been removed. The differences were close to normally distributed. The average standard deviation of temperature was 0.98 ± 0.27C.”

Reply to  Willis Eschenbach
September 30, 2019 3:22 pm

Thank you, Willis! A simple, direct answer to a simple question is ALWAYS appreciated.

September 30, 2019 11:45 am


‘Note the odd lack of data (with missing data replaced by a regular signal) from about 1940 to 1952.’

As you know, that little local difficulty of the war explains that.

My brave next door neighbour sailed on the northern convoys to Murmansk. The Northern sea route opened up in the 1920’s and the Russians set up an organisation in the 30’s to manage the shipping that was able to more easily traverse the route.

the arctic started to melt around 1908 (again) and as you know the bergs from the jakobshaven glacier is what did for the Titanic in 1912. That warming continued until the war but then became more intermittent

My neighbour assures me that oddly, there was no enthusiasm for the SST’ records they were asked to take and they should be disregarded, as the warmth of the cabin, the temperature outside and the risk of lights being seen by the enemy, meant the measurements were mostly imaginary.

My neighbour says the Russians in Murmansk were very hostile as they found it humiliating to take war aid. This naval force never received a specific medal from the allies for their very tough work, but were awarded one by the Russians only 4 or 5 years ago and the Russian ambassador in London presented them


Reply to  Willis Eschenbach
September 30, 2019 2:42 pm


When writing my article on the melting in the artic from 1908 it was clear there was a lack of international cooperation between scientists following the war, as the allies fell out and in particular the era of McCarthyism started.


There were more imprtant things to worry about than recording arctic ice in detail


Michael Ozanne
Reply to  Willis Eschenbach
September 30, 2019 3:52 pm

I’m surprised the anomaly doesn’t run longer.. My father was a NATO staff officer involved in Northern Wedding/Ocean Safari and their predecessors and successors… Operation of research stations and voyages of research vessels were extremely constrained up until the mid eighties….

Reply to  Willis Eschenbach
September 30, 2019 4:44 pm

The International Geophysical year was in 1957-1958, see http://www.nas.edu/history/igy/ . The prep started in 1953, so maybe that inspired some data collection in the run up to the main festivities.

Following a suggestion by NAS member Lloyd Berkner, the International Council of Scientific Unions in 1952 proposed a comprehensive series of global geophysical activities to span the period July 1957-December 1958. The International Geophysical Year (IGY), as it was called, was modeled on the International Polar Years of 1882-1883 and 1932-1933 and was intended to allow scientists from around the world to take part in a series of coordinated observations of various geophysical phenomena. Although representatives of 46 countries originally agreed to participate in the IGY, by the close of the activity, 67 countries had become involved.

I don’t know anything about the International Polar Years.

Mike Ozanne
September 30, 2019 11:50 am

“Note the odd lack of data (with missing data replaced by a regular signal) from about 1940 to 1952.”

WW2 and the Korean “Police Action” and frosty USSR-Rest of the World relations until after the end of the Berlin Airlift… Death of Stalin might have helped to.

September 30, 2019 11:54 am

I didn’t spend a lot of time looking onto this, but after 3 minutes of research I failed to find the ultra-erratic CO2 emissions levels that caused the ~2016 sea ice area instability (fall and recovery).

I have to assume the Mona Loa folks are hiding that erratic behavior in the atmospheric CO2 concentration levels for nefarious reason.

September 30, 2019 12:30 pm

Good article that discusses atmospheric and oceanic conditions that could have led to the extremely rapid 2014–2017 decline of the Antarctic sea ice cover.

John Tillman
Reply to  Renee
September 30, 2019 5:10 pm

As can plainly be seen in the NSDIC graphs, the decline was due to two weather events in 2016.

No surprise, after setting a record low minimum in early 2017, Antarctic sea ice recovered nicely in 2018 and this year, whose minimum and maximum are higher than last.


September 30, 2019 12:31 pm

What’s more arrogant? man causes global warming or man can stop global warming. Life is too short to worry and stop scaring the children.

Reply to  Willis Eschenbach
September 30, 2019 2:56 pm

Oh yeah, I like this article. It supports my 2016 El Niño theory. 2016 El Niño was different from 1998 El Niño in that it was followed by a weak La Niña.

Reply to  Renee
October 1, 2019 1:31 pm

Yes, different neighbors….

comment image

September 30, 2019 1:12 pm

The Antarctic is the “grand central station” of the deep ocean circulation (thermohaline circulation or THC).

Possibly the biggest “influencer” of the world’s deep ocean circulation.

Thus it’s noteworthy that flow of Antarctic deep water (dense cold bottom water) into the Atlantic, is increasing, reversing a previous declining trend:


This suggests that Antarctica, that as Willis points out generally leads climate trends, could be leading a cooling period with strengthening cold deep water formation at Antarctica.

This 2017 post by Wim Rost explains nicely how cold downwelling and deep water formation at the poles, refrigerates the planet.


John Tillman
September 30, 2019 1:37 pm

In satellite observations since 1979, according to NSIDC, Arctic sea ice summer minimum extent fell until 2012, while Antarctic winter maximum grew until 2014. Naturally, the media only cover the Arctic.

Two weather events in 2016 caused a drop in Antarctic sea ice during its melt season, leading to a record low summer extent in 2017, but minima and maxima rebounded in 2018 and 2019. Antarctic sea ice was still gaining as of yesterday.

The lows in Arctic summer minima–2012, 2019, 2007 and 2016–all have been associated with cyclones or late season wind shifts. Record low extent was seven years ago, and this year essentially tied with 2007 and 2016. The trend for the past 13 years is flat, after falling from 1979 to 2007 or 2012.

Waxing Antarctic ice and waning Arctic ice falsifies the CO2 hypothesis.

Reply to  John Tillman
September 30, 2019 1:51 pm

Waxing Antarctic ice and waning Arctic ice falsifies the CO2 hypothesis.

Instead it looks like the operation of the “bipolar seesaw”.

John Tillman
Reply to  Phil Salmon
September 30, 2019 4:55 pm

Indeed it does. Which means that globally, ice is not leaving the Earth, as advertised. Or at least not a rate significantly different from since the end of the LIA. Oceans are just moving the heat around. It’s not hiding in them.

September 30, 2019 2:08 pm

@ Willis
Nice post as usual.
I would like to call your attention to Figure 2, particularly the bottom data set, Northern Hemisphere.
What you said about this:
“The regular signal in the earlier parts of the record is an artifact. It is an interference pattern resulting from the removal of the seasonal signal.”

***** Stopped Me In My Tracks *****

Here is the story:
I like to keep as eye on the Boston Tide Gauge, just out of curiosity. I grabbed the data set back in 2017, and made some nice plots. Last summer, I thought it time to update my favorite plots, and so got the “Latest and Greatest”. After graphing, it was obvious the entire record had been altered. Of note is that the peak-to-peak level of the general data was increased dramatically, giving the appearance of a noisier data set. And this is clear through, from the start in 1921 to the present.
So I did what anybody would do. I graphed the difference. Simply, I subtracted the old(2017) from the new(2019) and got the difference.

The difference plot looks *exactly* like your Figure 2, bottom plot from start to ~1900.
Why should this be?
Why should an interference pattern suddenly appear, out of nowhere, in the Boston tide gauge data? And why should that pattern match a pattern you found in some old British Arctic sea ice data?
Here is the Difference Plot:
comment image

Anybody with any ideas, feel free to jump in.

Curious George
Reply to  TonyL
September 30, 2019 2:29 pm

While I don’t understand what a “BossDiff” means in terms of “xdata”, the idea of changing the past goes back to Orwell’s “1984”. Yes, we are witnessing all kinds of bowdlerization of all kinds of data. Yesterday’s data is gone, trust me, today’s data is much better.

These “scientists” could not be moved to use any kind of a version control system – a system that allows you to retrieve data as it was yesterday, a day before, and so on. Very dangerous in a post-1984 world.

Reply to  Curious George
September 30, 2019 3:36 pm

My apologies, I stare at my own plots long enough so I no longer see the obvious.
BossDiff = Boston Difference.
Y axis – Difference between 2017 data set and 2019 data set in millimeters. Total span from top to bottom in a typical annual pattern is 73 mm. The pattern is exactly 12 months long, repeating.
X axis – Data series in months, the data start is Jan. 1921

Here is a look at the first 5 years of numerical data, month by month, Jan. – Dec.
1921: 45 44 26 2 -11 -28 -14 -14 -14 -23 -11 25
1922: 45 44 26 2 -11 -28 -14 -14 -13 -23 -11 25
1923: 45 44 26 2 -11 -28 -14 -14 -14 -23 -11 25
1924: 45 44 26 2 -11 -28 -14 -14 -13 -23 -11 25
1925: 45 44 26 2 -11 -28 -14 -14 – 13 -23 -11 25

“These “scientists” could not be moved to use any kind of a version control system”
Back in the day, the data never changed. You would add to an ongoing data set, but you would never change existing data. In that world, a VCS made as much sense as a screen door on a submarine. I guess those days are long gone.

Steven Fraser
September 30, 2019 2:11 pm

Willis, a request: Please put units on the vertical axes of the first 3 graphs, or in the accompanying title or text.

September 30, 2019 2:21 pm

Antarctica’s ocean dynamics led the planet out of the last deep glaciation:



Antarctica’s ocean dynamics will likely lead us back into the next one.

Michael Palmer
September 30, 2019 2:25 pm

Congratulations Mr. Eschenbach for pushing one of the simplest ways of constructing a misleading graph. Your choice of vertical axis units stinks.


I suggest you build graphics using ACTUAL SEA ICE AREA instead of “% of global area. That way, the trend(s) in the data will be more obvious.

Michael Palmer
Reply to  Willis Eschenbach
September 30, 2019 5:13 pm

You are correct, the two graphs would be identical, BUT the vertical axis would encompass a much wider range. Your graph has a compressed vertical axis which hides the trend.comment image

See what an expanded y-axis does?
Secondly, using absolute values instead of anomalies hides the trend:
comment image

Your graphs are misleading, and you are using common techniques to do it.

Reply to  Michael Palmer
October 1, 2019 1:16 pm

So, what happens to the hockey stick when you plot to a vertical axis bottoming at last X years average minimum annual temp for the low end of the scale, and topping at last X years average maximum annual temp for the high end of the scale? ‘Bout as big as a pimple on skeeters’ tush, if I may say so.

Pot and Kettle come to mind…

Wim Röst
Reply to  Michael Palmer
October 3, 2019 10:50 am

Michael Palmer: “See what an expanded y-axis does?”

WR: Your graph is about sea ice extent, as Willis’ graph is about sea ice area. See comment https://wattsupwiththat.com/2019/09/30/water-behaving-badly/#comment-2810181

R Moore
Reply to  Willis Eschenbach
September 30, 2019 5:17 pm

I like to use Texas-area as a unit T which is a little over 250, 000 square miles and a little under 700,000 square kilometers. Then I hum a song by Asleep at the Wheel.

Reply to  R Moore
September 30, 2019 7:13 pm

Miles and Miles of Texas
Shout Wa Hey!

thanks for the memories

Reply to  Michael Palmer
October 1, 2019 6:50 am

Michael writes

Congratulations Mr. Eschenbach for pushing one of the simplest ways of constructing a misleading graph. Your choice of vertical axis units stinks.

One might suggest that graphing temperature anomalies with a vertical axis in tenths of a degree is misleading in conversations of rapid warming, with extinctions inevitable and existential threat to all life on earth. Put the changes on a full Kelvin scale and see what people think. Add the perspective of daily and seasonal variation and the crowds of protesters will all go home.

Michael Palmer
Reply to  TimTheToolMan
October 1, 2019 11:58 am

Putting the changes on a full Kelvin scale is misleading in the same manner that Willis is doing.

Reply to  Michael Palmer
October 2, 2019 3:56 am

Misleading in what way? If the goal is to show the effect of warming in the context of daily and seasonal variation then its the ONLY way.

Duke C.
September 30, 2019 2:34 pm

After watching this time-lapse video representing 500 ma to present, I was taken by how rare and uncommon glaciers and polar ice caps really are on this planet. 50 years of data? Pffft…


September 30, 2019 2:40 pm

Clearly the ice trend in the arctic is going down. Not so in the SH.
How come?

Michael Ozanne
Reply to  henryp
September 30, 2019 3:37 pm

The SH ice is a million miles from nowhere, the NH ice is within chimney belch of every source of soot and blacking and ice breaking you can think of… The SH is a n actual land mass with hills mountains and volcanoes with attached Ice the NH is floating ice alone constrained and compressed by the current flows between the fixed land masses. The two could not be less comparable…

September 30, 2019 2:43 pm


“just what the heck did cause the drop and subsequent recovery in Antarctic sea ice area from 2015 to the present?”

As I indicated in a post above, undersea/ice volcanic activity.

John in Oz
September 30, 2019 2:58 pm

Can anyone postulate on the effect so much changing ice distribution, hence weight, has on the Earth’s motion?

As it is spinning at approx 1,000Km/hr, I would expect the change of weight distribution from North to South as the seasons change to at least give us some ‘wobble’ due to precession effects. If so, how does this affect ocean circulations, ENSO’s, etc?

(Asked by an interested non-mathematician/scientist)

Reply to  John in Oz
September 30, 2019 4:54 pm

Yeah, I could, but don’t have time. Besides, XKCD has done a better job. The ice/water winter/summer effect is measurable, but it’s not the most important, at least over the long term.

Leap seconds are a great way to measure most of the effects, just not very good for the seasonal effect which is too small and cycles too quickly.


Leonard Weinstein
Reply to  John in Oz
October 1, 2019 12:29 pm

floating ice in liquid water make no significant difference. The displaced water from floating ice is the same mass as the floating ice in a given area. Only land ice make a mass distribution change effect.

September 30, 2019 3:02 pm

“are largely the result of variations in the Antarctic ice area, and not in the Arctic ice area that we spend so many electrons discussing”
Where the electrons go is in discussing summer Arctic. The big difference in the hemispheres is that except for summer, the Arctic ice is largely land bounded, which reduces full-year changes. Antarctic is the opposite; unbounded in winter, but the edge comes closer to land in summer. But mainly much freer to move.

September 30, 2019 4:09 pm

Willis, it seems perhaps you don’t respond too gar below root-level comments (forgive my file system orientation.) So reposting here:

Isn’t sea level rise (adjusted for landform subsidence) linearly coupled to global ice melt (sea and land)? And if sea level rise is very slow and steady (evident during human observation) doesn’t that indicate global ice melt (in aggregate) has been slow and steady as well? The wild swings and disparities between poles suggest that sea ice alone is a poor proxy for global temperature and resulting sea level. And smoothed sea level may be the only true proxy for global temperature.

Reply to  brians356
September 30, 2019 5:39 pm


“The wild swings and disparities between poles suggest that sea ice alone is a poor proxy for global temperature and resulting sea level.”

What is in your opinion the sense of comparing the poles?

The North pole is a piece of frozen ocean surrounded by Earth’s most land masses; the South pole is a piece of land surrounded by Earth’s most water surfaces.

Better proxies for global temperature would imho rather be the surface mass balances for Greenland’s and Antarctica’s ice sheets.

September 30, 2019 4:18 pm


You might have obtained the same results by using a more direct info, namely the sum of sea ice extent and area (aka 100% pack ice) you can easily obtain from colorado.edu:


Here is an evaluation of their monthly data using a spreadsheet calc. (It takes some little time to collect the stuff out of the monthly files, and to sort it.)

1. Absolute data

2. Anomalies wrt mean of 1981-2010

The monthly plots are interesting: both Arctic and Antarctic experienced a rather big drop in their anomaly time series nearly at the same time (Oct vs. Dec 2016).

What isn’t quite apparent in your graph is that the Arctic extent+area drop (-4.4 Mkm² wrt mean) is even lower than that in the Antarctic (-3.5).

Strrange things happen!

J.-P. D.

September 30, 2019 4:52 pm

Calling Nick and Griff…calling Nick and Griff….?
Come out come out climate changers wherever you are and show us where this fits in your computer models.

September 30, 2019 5:59 pm

It’s called paleoclimatology in case you’ve forgotten dudes. LOL

September 30, 2019 6:44 pm

Very interesting as all Willis posts tend to be. My humble contribution to the intense interest in sea ice in the context of AGW is that the evidence does not support the assumed causation.

Pls see


As suggested for example in alarmism such as this


Scott Bennett
September 30, 2019 6:49 pm

Talk about behaving badly!

As a completely unqualified observation, that dramatic dip in the Antarctic data doesn’t look right. I would entertain strong doubts about that isn’t a glitch, without very good evidence that it was real.

Are we sure that there wasn’t a change in the methodology at that time.
It looks like a processing artefact, (In the supplied data and not by Willis!) as it is the largest dip – by far – in the entire record and it just happens to begin when it was at its greatest extent (2014) after growing for the previous 40 years.

I’m sure somebody will school me on this but I’m not likely to be convinced without a a very very good justification and physical explanation to account for – what is IMHO – is a very very odd looking graph! 😉

Reply to  Scott Bennett
October 1, 2019 2:46 am

There actually was a big dip in the Antarctic data. But it isn’t unique, the NIMBUS data indicates a even greater abrupt dip in 1964-66:


Scott W Bennett
Reply to  tty
October 1, 2019 5:15 am

I know about Nimbus but you have linked to nothing that substances your claims.

Reply to  Scott W Bennett
October 1, 2019 8:09 am

Read the link:

“In fact, 1964 was the largest sea ice extent until 2014. Then in 1966 we saw the lowest ice extent that was ever seen. “

Reply to  tty
October 1, 2019 9:13 am

And then there’s the fact that Antarctic sea-ice area goes from huge to practically nothing in 6 months. So what does it matter all that much how it varies in that short a time-frame.

Oh yeah, it would mean Emperior penguins wouldn’t have to walk so far to get to their nesting sites.

Reply to  Scott Bennett
October 1, 2019 11:36 am

Scott Bennett

When I read such comments, I’m always wondering that certain persons considering the graph below:

(1) https://drive.google.com/file/d/1uMFrTs2tAILeE4CVCNbU2-5Ad7OuM7Fi/view

so often feel the need to consider sea ice drops to be ‘odd looking’ or simply artefacts, while sea ice peaks are always ‘good’. The same holds of course for ‘odd’ sea level and temperature peaks vs. ‘good’ drops.

It is really amazing that the commenter didn’t bother about the Antarctic peaks in 2008 and 2014/15, but did very well look at the ‘odd’ drop in 2016!

Let us have a look at the year 2016, by using in a first step the absolute daily data for sea ice extent:

(2) https://drive.google.com/file/d/1KADx20o9trSqzbXevHvvZWleLLwieIQ9/view?usp=sharing

We see in blue the Arctic (with 2016 in dark, and with the mean of 1981-2010 in light), and similarly in red for the Antarctic.

Oh! Impossible! No odd drop in the Antarctic data! There is no more to observe than what looks like a small deviation from the mean, beginning in the September of the year.

But nobody looks at absolute data time series, especially when having to compare many of them. Anomalies wrt some mean are used instead:

(3) https://drive.google.com/file/d/1cGEBcRMkNr1Nk0Ee_aM8tCMV2DuHXqQl/view

And here we see that in December 2016, the daily anomalies for Antarctic’s sea ice extent reach a maximum.

I don’t have daily data for the area (aka pack ice) but though it had its maximum in November, the sum of extent and area, averaged over the whole December, gave a deficit of 3.5 Mkm² for the Antarctic, as can be seen in the graph (1) showing the monthly anomalies.

Yes, Scott Bennett: it was a ‘completely unqualified observation’ 🙂

J.-P. D.

Reply to  Bindidon
October 1, 2019 11:58 am


We see in blue the Arctic (with 2016 in dark, and with the mean of 1981-2010 in light), and similarly in red for the Antarctic.

Oh! Impossible! No odd drop in the Antarctic data! There is no more to observe than what looks like a small deviation from the mean, beginning in the September of the year.

But nobody looks at absolute data time series, especially when having to compare many of them. Anomalies wrt some mean are used instead:

(3) https://drive.google.com/file/d/1cGEBcRMkNr1Nk0Ee_aM8tCMV2DuHXqQl/view

And here we see that in December 2016, the daily anomalies for Antarctic’s sea ice extent reach a maximum.

In June 2014 (from the Cryosphere data set), the Antarctica sea ice anomaly exceeded 2.16 million sq kilometers. More “excess” sea ice around ANtarctica than the entire area of Greenland.
The NSIDC couldn’t have that of course, and so the Cryosphere sea ice lab was shutdown, and the Antarctic sea ice responsility was shited to the NSIDC, Boulder Colorado. The ANtarctic sea ice promptly shurunk from +2.1 Mkm2 to -1.5 Mkm2 in only a few months. ANd no one cared, no “icebergs the size of Greenland were lost” headlines and screaming press releases. But, regardless of the NSIDC best efforts, the sea ice around ANtractica continued to expand, just like it has for most of the years since 1992.

Well, today (Oct 2019) the Antarctic sea ice is back above its daily 1979-1990 average for 1 October, and they are still growing. And there is no “Antarctic sea ice expanded by 1/2 the area of Greenland in only 2 years” screaming headlines either.

Oh. By the way, the maximum Antarctic sea ice occurs in late September, a few weeks after the mid-Sept Arctic sea ice minimum. Difference is, the Antarctic sea ice at maximum is up past 58-60 degrees latitude in the high angle sunshine. The Arctic sea ice minimum is in the dark at 79-80 north latitude (edge of sea ice) and 82-83 (middle of the sea ice.) Over the course of the entire year, ANtarctic sea ice reflects some 1.7 times the differential solar energy that the Arctic sea ice does.

And, above all, remember: “The less Arctic sea ice from today’s extents, the greater the heat loss from the newly uncovered Arctic Ocean to the infinite black cold of deep space.”

Reply to  RACookPE1978
October 1, 2019 5:06 pm


“In June 2014 (from the Cryosphere data set), the Antarctica sea ice anomaly exceeded 2.16 million sq kilometers. More “excess” sea ice around ANtarctica than the entire area of Greenland.
The NSIDC couldn’t have that of course, and so the Cryosphere sea ice lab was shutdown, and the Antarctic sea ice responsility was shited to the NSIDC, Boulder Colorado. The ANtarctic sea ice promptly shurunk from +2.1 Mkm2 to -1.5 Mkm2 in only a few months.”

Well, Sir, without presenting valuable sources (data, original documents), what you write here has zero value.

J.-P. D.

Smart Rock
September 30, 2019 7:15 pm


I got the impression that there was a monotonic downward trend in the Arctic ice area on your graph. So I took a screenshot of it, adjusted the 1% offset of the Arctic ice extent, and eyeballed a straight line through the squiggles. It shows Arctic ice area dropping from 2.24% of the globe to 1.89% between 1978 and 2019, a relative decrease of 16 percent (numbers are approximate).

The 13-month average of Arctic ice area on the NSIDC plot (which can be seen at the WUWT Sea Ice Page) drops from 12.2 million km² to 10.8 million km² over the same time interval, a relative decrease of 15 percent. They are both using the same satellite data sets, so this is not surprising.

It looks as though the Arctic Death Spiral is alive and well – it just doesn’t look that serious when you plot it with zero at the bottom of the Y-axis.

And of course the absence of satellite data pre-1979 means that you can write your own narrative about how this all fits into the long-term picture. There is, as other commenters have pointed out, a lot of anecdotal evidence that there was much less sea ice in the Arctic during the 1930s and 1940s, but it’s easy for alarmists to dismiss that stuff as unreliable. Or just ignore it. After all, the message is more important than the facts.

Matthew R Marler
October 1, 2019 12:05 am

Willis, thank you once again.

October 1, 2019 2:37 am

One thing should be clear to everybody:


But there are satellite data long before 1979. The early IPCC reports included data back to 1973, and more recent reanalysis of NIMBUS imagery has extended the record back to 1964. The data are available, but have, remarkably, never been properly published, one wonders why?



The coverage and data quality could be improved by using declassified CORONA imagery (which goes back to August 1960), particularly from the KH5 mapping camera missions, but apparently nobody is interested.

There is reasonably good data for the North Atlantic sector going far back, and for the Siberian coast (at least in summer) back to 1933, but Northern Canada is almost completely blank before the sixties.

Reply to  tty
October 2, 2019 9:32 am


Not to mention the whole continent of Oz only had a reasonable Stevenson Screen rollout by 1910 so they won’t accept the stinkers of the 1880s but still the climate changers live in the moment and we’re all doomed-
Snake oil and catastrophism.

October 1, 2019 11:39 am

“there was no trend at all in global ice area. None. Well, to be accurate, the trend 1980 through 2004 is -0.0000000000000001% per decade … ”

This makes me think that there is very little change in ocean temperature. Any increase in Ocean temperature over 25 years should cause a statistically significant change in ice volume.

Years ago, when I was calibrating temperature elements at power plants to within 0.1 percent accuracy, we used insulated ice/water filled containers with a magnetic stirrer. We had procedures requiring that the quantity of ice was at least 30% of the container. We were comparing one “Laboratory” grade NTSB instrument for the reference with the instrument being calibrated. Occasionally, towards the end of the day you would see a change in the last digit of the reference instrument.

Reply to  Usurbrain
October 1, 2019 1:28 pm

Interesting. How much of an effect does water purity/cleanliness have on freezing temperature? What effect at the end of the day caused the change? How many digits on the instrument? C or F?

Reply to  Vince
October 2, 2019 9:46 pm

The freezing point of water decreases 1.86ºC/mole of solute, if the water is in contact with air during the day you’d end up with about a millimole of dissolved gas so a change of about 0.001ºC

Ulric Lyons
October 1, 2019 12:11 pm

Definitely the polar see-saw effect. The increasing trend in the Antarctic since 1993 is directly following the weakening of the solar wind, and all that AMO and Arctic warming is an amplified negative feedback.

james feltus
October 1, 2019 1:53 pm

“Perhaps most importantly, am I unable to detect any sign of any “climate emergency” in that record.”
As written, the sentence is a question.
More seriously, that’s another fine article. Thanks, W.E.!

October 2, 2019 7:25 am

As usual I am impressed by Willis’s data presentation. I hope he won’t feel offended that I have taken his analysis a step further.

I took account of both sea ice and snow cover. The basic data come from NSIDC and Rutgers University. It shows a slight downward trend from 1978 to present but equivalent to only 0.1% per year.

You can see my posting at:

October 2, 2019 11:03 am

Tony Heller has posted this video of September Arctic ice showing that in the last decade September as a whole has changed from a falling to a marginally rising trend of ice extent:

comment image

Reply to  Phil Salmon
October 2, 2019 11:11 am

Oops – misread the graph. From the start to the end of September the trend specific to that date has changed from ice loss to ice gain.

As Javier once commented this (unsustainable) trend of forward movement of autumn may signify an impending climate transition of some kind.

Reply to  Phil Salmon
October 2, 2019 12:35 pm

Phil Salmon

That is simple cherry-picking.

Trends for absolute Arctic sea ice since 2007 (as chosen by Javier):
– September months: +0.2 Mkm² / decade
– March months: -1.2 Mkm² / decade.

When you look at ice on sea or on land, you look at both the highest and lowest levels.
The gain in March has a much lower trend than the loss in September.

Nothing dramatic, but it is as it is.

October 2, 2019 11:41 am


Thanks a posteriori for this guest post which motivated to do a similar job, through which I indirectly obtained a 1° land mask 🙂

I also wondered about the data before 1970.

What I don’t understand in your text is this:

“Finally, there is no statistically significant trend in the full 1973 – 2019 dataset.”

This can be correct only when you calculate the linear estimates on the basis of the absolute data time series. The HadISST Ice Globe trend then is

0.10 ± 0.09 Mkm² / decade

and that is indeed not very significant.

But it is due to the fact that absolute data often has higher deviations from the mean, which lead to a higher standard error. If you compute the linear estimate on the basis of anomalies e.g. wrt the mean of 1981-2010, you will obtain

0.11 ± 0.03 Mkm² / decade

due to the fact that you removed those seasonal dependencies Roy Spencer calls the ‘annual cycle’.

J.-P. D.

Reply to  Willis Eschenbach
October 2, 2019 4:12 pm


Thanks, sounds good.

October 2, 2019 4:34 pm

It might be interesting to look at a comparison of (hopefully correct) anomaly-based plots, processed out of HadISST Ice and Colorado SIDADS data for the Arctic and Antarctic regions.

1. Arctic sea ice

2. Antarctic sea ice

Hadley’s data looks smoother than the 100% satellite data, especially in the Arctic.

October 13, 2019 4:15 am

test post

%d bloggers like this: