By Steve Goddard and Anthony Watts
If you zoom in far enough, most anything looks scary, like this picture of a human head louse.
But when you look at it in the scale of our normal experience, not so much.
Be it lice or ice, the scale of presentation matters.
There is often criticism of cherry picking when it comes to time scales of climate data. In the case of satellite sea ice data presentation, both time scale and vertical scale are magnified. There’s only about 30 years of satellite ice data, whereas Arctic sea ice has been around for millions of years. Vertical scale is magnified to show the smallest fluctuations. Willis Eschenbach made and excellent point about scale when he comparatively demonstrated the scale of ice melt in Greenland in his essay: On Being the Wrong Size. When compared to the bulk volume of ice, the current Greenland melt is statistically insignificant.
There has been a lot of talk about commercial shipping opportunities through the “soon to be ice free” Arctic. These are normally based on highly magnified graphs published by organisations like NSIDC, similar to the one below.
A different view emerges when you take the raw data from NSIDC’s web site and plot it on graphs with a more appropriate vertical scale. Done that way, the downwards trend for April ice is 0.039 million km²/year.
The surprise of scale?
When you calculate the slope, it suggests that April sea ice extent won’t reach zero until the year 2385.
Oh, that can’t be right. How about May? May will be ice free in the year 2404, only 394 years from now. (The US is 234 years old. Copernicus was placed on the “Catholic Forbidden index” 394 years ago.)
June will be ice free in the year 2296.
July will be ice free by the year 2151.
August will be ice free by the year 2103
September will be ice free by the year 2065. (Note that September 2009 was right on the trend line.)
All of the data and plots are available here in this Google online spreadsheet.
September is the minimum and ice starts to freeze up again. No chance of an ice free Arctic in October. But something must be wrong. The experts said that the Arctic would be ice free by 2008, and that it would be ice free by 2013.
“Our projection of 2013 for the removal of ice in summer is not accounting for the last two minima, in 2005 and 2007,” the researcher from the Naval Postgraduate School, Monterey, California, explained to the BBC. “So given that fact, you can argue that may be our projection of 2013 is already too conservative.” “In the end, it will just melt away quite suddenly”
NSIDC director, Dr. Mark Serreze also says this in this 5/20/10 Globe and Mail article:
“We are going to lose the summer sea-ice cover. We can’t go back.”
Dr. Serreze is still on the ‘death spiral’. He hasn’t changed his tune.
While skeptics see cycles, by saying “we can’t go back” Dr. Serreze apparently assumes the linear trend will continue to zero.
You can see from the graphs above how ridiculous those claims are. Even if the current trends continue, there is no reason to expect an ice free Arctic anytime in the next 50 years. And even more interesting to me is the fact that September, 2007 was really not that interesting. It was only 1.5 standard deviations off the trend line, i.e. almost following the 30 year trend.
All of the the main Arctic ice experts underpredicted the 2009 minimum, except for WUWT – which predicted it correctly and early.
—————————————————————-
Science is the belief in the ignorance of the experts
-Richard Feynman
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Steve, I’m a little puzzled by one of your earlier statements:- “”” There’s only about 30 years of satellite ice data, whereas Arctic sea ice has been around for millions of years. “””
Assuming that those thirty years are the most recent 30 years (maybe that’s not true), then that would take us back to the time around 1979-80 as the beginning of satellite icea data.
Now I recently got jumped all over for making that same allegation; well I put is slightly differently, in saying that the first polar orbit satellites went up around 1979. So Phil and barefootgirl jumped all over that and said it was 1961 (Tiros-1 or somesuch).
OK; maybe pedantically correct; but when I made my (erroneous) statement; the discussion was about satellite sea ice observations in the arctic. So to me it was obvious that I was referring to the first satellites that started making these satellite sea ice observations. So the Russians showed in 1957 that you could toss an inert lump of metal up there, to fly around the earth; but it couldn’t take arctic sea ice pictures in any frequency range; so who cares if they did it in 1957.
So what is the real skinny; could Tiros-1 take arctic sea ice pictures or not; and if so why doesn’t the satellite data go back to 1961 ?
Seems like one has to spell everything out here in words of one syllable, or peop-le just don’t seem to catch on. No wonder it takes so many thousands of “scientists” to ferret out a few simple physical facts.
And I specifically asked why the period from 1979-2000 was taken as the baseline for what is normal arctic sea ice; and I drew responses that some people were using data back to 1972 (as I recall).
So does the peer reviewed widely accepted satellite arctic sea ice data start in 1979 or doesn’t it ?
Steve writes: A different view emerges when you take the raw data from NSIDC’s web site and plot it on graphs with a more appropriate vertical scale. Done that way, the downwards trend for April ice is 0.039 million km²/year.
———————
Can you please tell me how the trend changes when you plot the extent from 0 to 20 million sq-km? It shouldn’t so your sentence above is very misleading. And why is this a more appropriate vertical scale? If you want to highlight interannual differences, plotting it on an appropriate scale is the right thing to do.
You seem to avoid mentioning the fact that the slopes are statistically significant in all calendar months, that not much change is expected in the winter ice cover (and thus the April trend being small is what is expected), or that no one expects the trends to remain linear, and in fact the data shows the changes are not linear.
George E. Smith says:
May 24, 2010 at 9:51 am
The real skinny is that indeed, there are earlier satellite observations and have been used in data sets such as the Had1SST data set. Basically observations from 1953 onwards are the most reliable. NSIDC only shows the modern satellite data record (i.e. based on multichannel passive microwave sensors with similar frequencies). But one can blend that record with earlier data. Of course, what would likely happen is that skeptics would complain about how the data were blended together to make a longer time-series.
stevengoddard says:
May 24, 2010 at 9:20 am
The reason that Serreze forecast an ice free pole in 2008 was because it started out with first year ice that summer. The lesson to be learned is perhaps that even first year ice won’t disappear that far north.
—————————–
Correct. And Serreze didn’t forecast an ice free summer, but simply thought there was possibility of open water at the pole in summer of 2008 because it was covered by first-year ice. So there is a lesson to be learned there…but I would expect as temperatures in the atmosphere and ocean rise further, that even first-year ice at the pole may not survive…
skye
If the April slope is statistically insignificant, why did NSIDC include it in their sea ice news?
http://nsidc.org/images/arcticseaicenews/20100504_Figure3_thumb.png
You should take that up with them – because if it weren’t for that graph, I wouldn’t have had any reason to write this article.
wildred,
http://wattsupwiththat.com/2008/08/15/arctic-ice-extent-discrepancy-nsidc-versus-cryosphere-today/#comment-32714
stevengoddard says:
May 24, 2010 at 9:11 am
R. Gates
Great link thanks. It shows exactly why scientists shouldn’t jump to conclusions about non-linearity based on a one-year blip. I’ll bet the authors regret having written it now.
———————————————
I’m not sure I follow you Steve. That presentation doesn’t disprove non-linearity in the sea ice system. Perhaps you are thinking that 2009 proved the presentation was in error. But looking at the presentation the authors clearly state that natural variability remains important. Which I believe is what the scientists have been saying all along. But what I think is interesting is that you can have a weather pattern that favors ice retention, and while it may help keep the ice from setting a new record low, it certainly does not bring the ice back up to conditions seen 20-30 years ago. Last summer is a perfect example of that, and this year is showing that to be true again since your much-touted negative AO this winter didn’t help the ice pack out like you thought it would.
I read that powerpoint and didn’t walk away with an assumption being made about a single year but rather that processes seem to be working together to allow for large ice losses to occur in summer. The key behind it appears to be thin ice, and continuance of anomalously warm air and ocean temperatures.
I also didn’t see any mention of CO2 being responsible for the observed ice loss. Why ignore real science analysis of the declining sea ice?
http://www.dailymail.co.uk/news/article-1050990/The-North-Pole-island-time-history-ice-melts.html
Anthony, it’s not a concern, it’s a prediction. If your and Steven’s gamble turns out wrong, this thing will haunt you more than any other screw-up so far.
REPLY: No pain, no gain, and unlike you and your cowardly shoot from behind the rocks friends, we have the courage to put our name on it.
I’m sure you’ll do everything in your cowardly power to make sure everyone knows if we miss the mark, and ignore it if we do. Like you ignored our correct prediction last year.
Such a negative person you are, safe in your cocoon of anonymity, shooting from behind the rocks. I see what you write elsewhere, and your MO is that of a coward unable to stand up for his own words. – Anthony
stevengoddard says:
May 24, 2010 at 10:02 am
skye
If the April slope is statistically insignificant, why did NSIDC include it in their sea ice news?
http://nsidc.org/images/arcticseaicenews/20100504_Figure3_thumb.png
You should take that up with them – because if it weren’t for that graph, I wouldn’t have had any reason to write this article.
——————————————–
It is statistically significant. When did I say it wasn’t? Just because it’s small, doesn’t mean it’s not statistically significant. Antarctic sea ice trends are small, but have recently become statistically significant and you point to them on a regular basis as evidence for no global warming.
BTW..NSIDC shows the monthly trends for every end-of-the-month post, regardless of how large they are. If they were not statistically significant, they would have stated so.
wildred
What do you think Hansen meant by the statement below? Was he suggesting that ice melt was driven by CO2?
“Conceivably it will be necessary to return carbon dioxide even closer to its pre-industrial value of 280 ppm, especially for matters such as ice sheet stability”
http://solveclimate.com/blog/20091123/tipping-points-melting-ice-rising-oceans
rogerkni
May 24, 2010 at 9:10 am
If we only have 30 years data, how can you possibly assert there are 60 or 200 year cycles?
She meant to say, “satellite data” on arctic ice extent. We have other data on the PDO.
There is data on Arctic sea ice before the satellite era (as we know, the further back in time we go, the more porous the instrumental record). But if that’s what Gail meant, it applies to other data.
It’s difficult, if not impossible, to reconcile various contributions at this site. The data is deeply flawed, we are told, but then people use the data to say all sorts of things with great confidence. Solar data is good in the satellite era, but before satellites we had to use our eyes (with some telescopic assist). Yet Gail asserts with confidence that there are 200-year solar cycles – something I’ve never heard of, and assume comes from, at best, some obscure paper. The PDO is a recently discovered climate phenomenon (1997), so we rely on proxy measurements to assess it’s variability through time. How confident should we be that we have that down, let alone correlation to the instrumental temperature record?
It seems to me that people pick and choose what data supports their position. Whatever uncertainties are inherent are completely ignored when the message suits the predilection. This is so for casual conversation in blogs. In the vast majority of science papers, uncertainties feature prominently. Acknowledging uncertainties lends credence to analysis. I wish we’d see more of that in discussions of the science in the comments section. We’re too quick to judge, too slow to reflect. Not always, but too often, IMO.
stevengoddard says:
May 24, 2010 at 9:17 am
Wren,
I didn’t realize that plotting NSIDC data was arrogant. Perhaps it would be more reasonable to forecast an ice-free Arctic by 2013?
=====
I didn’t say it is arrogant, and I’m still scratching my head trying to figure out what I said that may have implied arrogance.
My money would be on the Arctic being ice-free closer to August 2013 than to your projection of August 2103.
barry says:
May 24, 2010 at 7:30 am
“I thought the sea ice recovered last month. Back to normal was the message from WUWT. I see that all indicators show the sea ice extent has dropped below average and is now near 2007 levels for this time of year.
Can we get an update on whether it has recovered or not?….”
______________________________________________________________________
Barry it is a complex subject with a lot of factors. Air temperature (minor) ocean temperature (major) and most of all WIND. The wind blows the ice out of the arctic circle into the warmer seas where it melts.
Juraj V. presented some very nice pictures showing the ice extent in May 2007 (the year of the greatest ice melt) and May 2010. The color on the pictures shows the percentage open water vs ice in each area. Notice that the 2010 ice is all deep purple (100% ice) while the 2007 ice had lighter purple areas throughout. This indicates 80% ice/20% open water and the fact the ice was already breaking up. This means the wind could push it out of the arctic circle more easily.
Here is Juraj V. ‘s pictures: http://weather.unisys.com/surface/sst_anom.html
stevengoddard says:
May 24, 2010 at 9:11 am
R. Gates
Great link thanks. It shows exactly why scientists shouldn’t jump to conclusions about non-linearity based on a one-year blip. I’ll bet the authors regret having written it now.
_______
Steve, the very real potential of a non-linearity of Arctic Sea ice loss didn’t change because of the so-called “recovery” of 2008-09 summer sea ice minimum. Arctic Sea Ice volume has shown no such recovery. I’m wondering how you will characterize the summer sea ice minimum if it fails to meet your expectations of a continued recovery above the 2008-09 levels, and in fact, falls back toward the very non-linear level of 2007? You have put a lot of expectations for recovery based simply on more multi-year ice, when your perception of that multi-year ice might be different if you would look at its volume and realize that it is not the “solid core” of ice that it might have been in decades past.
Finally, I don’t think that Dr. Mark Serreze nor Julienne Stroeve or Don Perovich regret any of their predictions for a non-linear arctic sea ice melt. They may regret not having a precise enough model to predict the exact ups and downs (i.e. 2008-2009 summer melt) within that non-linear and accelerating melt, but I would bet they are still quite confident in the overall downward spiral of arctic sea ice.
@ur momisugly Bob Tisdale says:
May 24, 2010 at 7:10 am
As this is the Arctic, I think it would be safer to go by the AO;
http://jisao.washington.edu/ao/
Its positive phase till 2035 will frequently bring low ice extent, except where there are colder winters. I have 2014, 2016, 2017, 2018 and 2020 mapped as hard winters, 2025 till 2038 will be overall very warm. So my long term outlook is completely the opposite to Joe Bastardi`s.
Mark Serreze has apparently placed his bet for 2010
May 24, 2010 Mark Serreze of the center forecast the ice decline this year would even break 2007’s record.
http://www.allheadlinenews.com/articles/7018784377
barry says:
May 24, 2010 at 7:30 am
“I thought the sea ice recovered last month. Back to normal was the message from WUWT. I see that all indicators show the sea ice extent has dropped below average and is now near 2007 levels for this time of year.
____________
Barry, Arctic sea ice EXTENT almost (but not quite) returned to the 30-year normal line in April, but that gain was almost entirely driven by a persistent low pressure system over the Bering Sea that created a great deal of new and very thin ice. Almost all of that March-April “bump up” in Arctic Sea ice was from this thin ice that formed, and now is just as quickly gone, and we are back once more to the very negative anomaly situation we’ve seen in the past few years (over 1,000,000 sq. km as of today). Also, those grand statements of a “recovery” were only looking at extent, and not volume, which is far more indicative and important to the long term condition of the Arctic sea ice. Bottom line…any insinuation that the Arctic sea ice had returned to some normal conditions would be misleading at best. (with normal being based of the last 30+ plus years of reliable data, and not pictures of submarines coming up) somewhere in the Arctic in the 1950’s)
http://climaterealists.com/index.php?id=5636
April 30th 2010 Mark Serreze, director of the National Snow and Ice Data Centre, has admitted that it had been an overstatement to conclude in 2007 that global warming had pushed the Arctic to a tipping point from which it might not recover.
R. Gates says:
May 24, 2010 at 8:06 am
“……An ice free Arctic in September of 2065 would be unprecedented in human recorded history, and would have lots of other implications. ….”
As a disgruntled Icelander mentioned a few weeks ago. Scientist keep ignoring Icelandic history in their claims of “unprecedented warmth.” You just did it again. The Norse history is recorded in their sagas and proof that the sagas were historic has been found.
Here is the Norse history:
The Norse in the North Atlantic
online features
The Fate of Greenland’s Vikings
Here are the Greenland temperatures:
Greenland Ice Core Data
Your are twisting the facts again.
Gail Combs says:
May 24, 2010 at 11:21 am
R. Gates says:
May 24, 2010 at 8:06 am
“……An ice free Arctic in September of 2065 would be unprecedented in human recorded history, and would have lots of other implications. ….”
As a disgruntled Icelander mentioned a few weeks ago. Scientist keep ignoring Icelandic history in their claims of “unprecedented warmth.” You just did it again. The Norse history is recorded in their sagas and proof that the sagas were historic has been found.
Here is the Norse history:
The Norse in the North Atlantic
online features
The Fate of Greenland’s Vikings
Here are the Greenland temperatures:
http://www.archaeology.org/online/features/greenland/“>Greenland Ice Core Data
Your are twisting the facts again.
____________
Gail, I really resent being accused of “twisting facts”, but I do appreciate you providing links to fables and stories of the past. There is nothing in any of what you gave me that proves or refutes the scientific claim that an Ice Free Arctic would be a unique event in recorded human history.
Wren,
Sorry – I was responding to Neven and accidentally put your name at the top.
jeff brown says:
May 24, 2010 at 9:04 am
Probably you are just being willfully obtuse but I think most readers understand that Steven Goddard is just pointing out the absurdity of magnifying the y axis for dramatic purposes and extrapolating trend lines to reach ridiculous or unfounded conclusions. Nowhere in the article do I read that Steven actually thinks these extrapolations are any kind of forecast. Indeed, the article simply reinforces the fact that 30 years of data are insufficient to forecast Arctic ice coverage in the future.
barry says:
May 24, 2010 at 8:15 am
Gail,
There are natural cycles of about 60 yrs (ocean) and 200 yrs (sun), but we only have data for about 30 yrs so there is no way in heck we have a good handle on the natural variability or what the “true average is”.
If we only have 30 years data, how can you possibly assert there are 60 or 200 year cycles?
_________________________________________________________________________
Super easy. The GRAPHS are for the scientific measurement of sea Ice in the Arctic they only go back to the mid 20th century, so THAT data is only available for about 30 years. Other information from proxy, not direct measurement goes back a lot further.
If you want to pick nits (or lice) how about this graph from Greenland that shows we are in a gradual cooling phase and the last thirty years is just a minor uptick? Greenland GISP2 Ice Core – Interglacial temperature
Here’s an excellent link on how loss of Arctic Sea ice can cause an amplification in warming on land:
http://www.the-cryosphere.net/3/11/2009/tc-3-11-2009.pdf