By Steven Goddard
No matter what happens with the summer Arctic ice minimum, NSIDC will report that the long-term trend is downwards.
Why? Because of mathematics. In order to reverse the 30 year downwards linear trend, this summer’s minimum would have to be nearly 20,000,000 km². More ice than has ever been directly measured.
In other words, we could have a “Day After Tomorrow” scenario, and the mathematical trend would still be downwards.
Conclusion: You can count on NSIDC to continue reporting a downwards trend, regardless of what happens over the next few years. For now, it will be fun seeing what happens over the next eight weeks.
You might find it interesting to know that from 1979-1998 the September trend is -0.032 +/- 0.017 million sq-km per year and from 1999-2009 it is -0.165 +/- 0.044 million sq-km per year. These trends are statistically different from each other at a 95% confidence level. A smoothed curved trend line computed using Locally Weighted Scatterplot Smoothing (LOWESS) for the entire 1979-2009 time-period further supports an acceleration of the September trend.
It is also interesting to note the increased variability after 1990. For the Arctic Ocean domain (i.e. excluding the E. Greenland Sea and the Canadian Archipelago), the standard deviation in September ice extent from 1979-1990 is 0.34 million sq-km, while from 1990-2009 it is 0.79 million sq-km, more than a factor of two larger. This increase in variability coincides with an early 1990s shift in the spring ice thickness distribution in the Arctic Ocean domain towards a larger fraction of thin, first-year ice (e.g. Maslanik et al. 2007), linked in turn to a period when the Northern Annular Mode (NAM) was in its strongly positive phase (Rigor and Wallace, 2004). An increase in variability of summer ice extent as the spring ice cover thins is a feature of coupled global climate model simulations (e.g. Holland et al., 2008). The reasoning is that as the ice cover thins, large regions become especially vulnerable to melting out during summer under favorable atmospheric conditions. Conversely, even in a warming climate, occasional summers are cool enough for much of the first-year ice to survive. The overall result is increased variability.
R. Gates says:
June 25, 2010 at 6:18 am
It is rather like a EEG …
Good lord! Do you preen in the mirror after each post?
Why I’ll bet that you think so highly of yourself that you verily hoist yourself up on a pedestal daily!
R. Gates at 6:18 am: good wrap-up of the situation.
I would like to add one more observation for that time series. In the last couple of years the graph has a different character, showing much wilder anomaly changes than before, making it get a much more ragged appearance. This is indicative of ‘catastrophe’ (mathematical concept). It happens when some phenomenon is subject to threshold behaviour and is near that threshold.
In the case of Arctic ice, the threshold to watch for is in thickness over large areas when it gets to about twee feet. At this point, the ice breaks up, letting the sun do its job on the bare sea surface and accelerating the melting process vastly. The two dramatic drops of 2007 are exemplaric for this. A comparable thing can happen any moment now in view of the already extremely thin ice over Canadian of the Arctic sea.
The combination of melt and currents this year actually suggests that absolutely all of the multiyear ice, at least the ice older than two years, will disappear by early autumn. A new first for the region.
Nice way to miss the point, Enonym. Anthony didn’t say he wanted to moderate anonymous bloggers, he just intended to ignore them as unimportant. And he’s right – you and carrot eater are nothing but typical net-trolls, unworthy of any serious consideration.
And carrot eater, you make a typically stupid analogy. Even in a double blind process, *someone* in charge of the process always knows who’s doing the reviews, even if the author and the reviewer don’t know who each other are.
Think about it – would you advocate leaving a scientific paper sitting out in a New York subway station, and then take seriously any anonymous comments left scribbled on it by random passersby? That’s what anonymous internet criticism is, and that’s why Anthony is correct to ignore any such scribblings from you. (and from me, if he chooses. But then I’m not pretending to be some mathematical genius like you are)
I think for everybody who is looking at trends, it is enlightening to look at figs 11 and 12 in
http://www.oism.org/pproject/s33p36.htm
We see that the sea levels have been increasing monotonically ever since the little ice age, which means that the ice has been diminishing by inference, because that is where the water is stored.
It shows that the trend has little to do with the hydrocarbon burning, since it starts a century before it takes off. The original references are appended in the review.
In these discussions we get carried away and lose the point. Is CO2 the prime mover? That is the point.
It is interesting to see what the ice is doing more because of the suspicion that we are close to the tipping point to the next little ice age, than trying to pin it on CO2. Data is clear that it is not CO2.
tallbloke says:
June 25, 2010 at 3:45 am
I wonder what the trend would look like if the data went back to 1000.
All of this fun aside, what would be the impact if the polar ice cap should just melt away?
No one has really explained to me why I should be concerned. Let alone what I am supposed to do about it.
I do know this, from history, it was very warm in the Arctic around 1000. This is based on my reading of the founder of CRU, Dr. Lamb.
http://en.wikipedia.org/wiki/Hubert_Lamb
See his book Climate, History and the Modern World (ISBN 0-415-12735-1).
At one time, Vikings raised cattle and sheep and grew wheat and barley, where today’s inhabitants must import fodder for their sheep. So it must have been warmer.
What was the status of the ice cap at that time? It must have been smaller, right? Or am I getting something wrong here.
Not only was it much warmer than today, the polar bears and seals somehow managed to get through it all. Otherwise, they wouldn’t be with us today.
Can someone, please, tell me why I should be concerned about any trends in the Arctic ice?
Seriously, anyone who thinks that a straight “trend” line is what applies here doesn’t really understand the physical world. Where’s the curve? Where’s the wave? Who failed to comprehend the concept of “cyclical”?
Tragically, it will require that the entire climate cycle be documented before some people will get it. Even then, maybe they’ll cling to the same artificial range as a “baseline”.
Arctic sea ice volume is of more interest than sea ice extent as a measure of climate change. While the history of sea ice extent before the satellite era (1979-present) is speculative, models such as PIOMASS have been used to hindcast the sea ice volume into the past.
http://www.vliz.be/imisdocs/publications/121856.pdf
The Polar Ice Center retro page shows a plot of the hindcast sea ice volume vs the NAO index
http://psc.apl.washington.edu/zhang/IDAO/retro.html#Satellite_ice
The models show that that the current sea ice volume low is not unusual.
A JGR study reported that there is no trend in arctic sea ice volume over the last century based on hindcasting.
http://www.agu.org/pubs/crossref/2007/2006JC003616.shtml
The NAO is currently slightly negative. If it goes strongly negative like the 1960’s
ice will build fast.
http://en.wikipedia.org/wiki/File:Winter-NAO-Index.svg
“Anthony: If you won’t allow anonymous commenters, you’ll have to moderate about 98% of all your commenters.”
I’m guessing Anthony differentiates somewhat between anonymous commenters and anonymous blog “owners” i.e. bloggers… e.g. “Tamino”.
vukcevic says:
June 25, 2010 at 3:31 am
What did that magnetic field change?. Magnetic fields are generated by electric currents. Is it the solar “wind” over there, as the auroras show?
Steven,
Why did you post a Mann graph and label it “September Artic Ice Extent Trend”?
😉
R Gates, I beg to differ. Climate has less to do with trends than it does with range. That includes precip range, temp range, clear sky range, etc. Almost the entire planet has successfully been divided up into climate zones. Climate zones have little to do with a straight trend, or even a smoothed running average. Were that the case, many seeders and planters would lose crops every year, basing their selection on such nonsense as “trend” or “average”. One does not plant around the trend or average if one wants to stay in business and feed you. One plants within the range. That is climate in my opinion. The fact that the trend slopes up, down, or the average wriggles around in the middle does not twist up my knickers in the least. Range I worry about, and follow record daytime and nighttime range closely.
To wit, not one climate zone has undergone designation change in the US since this global warming business has moved from intellectual discussion to front seat barn burner political correctness stage. Why? Range has not undergone any kind of drastic movement trend such that climate zone designation must change.
So much for death spirals, no matter which GPS address you are talking about.
Enonym says:
June 25, 2010 at 6:09 am
“David L:
Of course one can use a linear trend for an interval of 30-ish years if the variations are sufficiently slow. Being a fellow scientist, you know that. Further, let’s consider an imaginary situation where you have a planet where ice *is actually* melting linearly. Would you still not use a linear regression, since one can not have negative values as time approaches inf.?”
I don’t really like the use of lines because it gives the false sense that the trend will continue ad infinitum. There could be an asymptote lurking at very large non-zero levels, but the linearity focusses the discussion on the continued drop and not the very likely potential that there is a new, stable equilibrium of less ice
matt says:
June 25, 2010 at 7:30 am
Well doy! That’s because there IS a longterm downward trend…
Downward trend from where?
The “purpose of this post” is to point out that nothing which happens over the next few summers will change the 30 year trend.
We could have record high minimums for several years in a row, and the widely vaunted “long term trend” would still be downwards.
bob
You claim to know the current “melt rate.” Please tell us what it is, and how you differentiated melt from compaction.
Richard Holle says:
June 25, 2010 at 7:34 am
Dave Springer says:
June 25, 2010 at 4:18 am
[–snip–]
The answer my friend, is blowing in the solar wind,
If what you say is true, then a mere ‘rewinding’ of history would reveal that matter.
When were the last times such happened in, say, the Medieval —or other— times?
Dave Springer says:
June 25, 2010 at 4:18 am
It appears to me from the satellite record there was no downward trend in arctic sea ice from 1979 to 1996 then in 1997 something happened that started a downward trend that lasted for about 10 years then leveled off.
CO2 emissions didn’t skyrocket in 1997 did they?
[–snip–]
It sure isn’t the global average atmospheric temperature inching up or down a tenth of a degree or two per decade as the 1997 El Nino made the ocean surface temperature in a huge swath of the Pacific shoot up 5 whole degrees above normal in a matter of months – more extra energy than “a million Hiroshima bombs”!
VILLABOLO RESPONDS:
First of all Dave, there WAS a downward trend in the Arctic sea ice from 1979 to 1996. It was, of most importance, a thinning trend followed by a shrinking trend. This is a clear indication that the ocean is getting warmer. The link immediately below shows the downward trend in ice thickness.
http://nsidc.org/images/arcticseaicenews/20091005_Figure5.png
No, Carbon Dioxide levels do not have to shoot up immediately. Why should Nature respond instantaneously? CO2 levels have been helping the oceans to absorb more heat throughout time. This would not only affect the Arctic ice cap in its thinning but would also BUILD UP HEAT FOR THE NEXT EL NINO.
For reference to the temperatures throughout 1979-present, the following UAH chart should help put things in context. http://www.drroyspencer.com/wp-content/uploads/UAH_LT_1979_thru_May_10.gif
What started happening in 1997 was not, however, a single year event whose influence managed to affect a whole decade. What happened in 1997 onward was an approximately .15 degree Centigrade jump in temperatures from the previous two decades that was caused by El Ninos in 1997 AND 2001-2007 as well as that unmentionable factor, GW.
If El Ninos are getting warmer, as the chart above indicates, then what is increasing the heat of the ocean from which they draw their heat from? You cannot assume that the El Ninos, from 1997, and even La Ninas from 1996 onwards are consistently warmer than those from 1976 to 1996/7 out of mere coincidence.
As for1997 temperatures shooting “up 5 whole degrees above normal.” You did not specify whether the temperatures you used were in C or F. You also did not define what “above normal” was. “Normal” in relation to what time period or baseline?
Furthermore, you were referring to the temperature rise in the area immediately above the Pacific Ocean, where El Nino was coming from in 1997, and then comparing it to the “global average atmospheric temperature” for the past decade. That is comparing one apple to 10 oranges. To be consistent you have to compare Global to Global.
The UAH Globally Averaged Satellite-Based Temperature Chart, referenced above, shows that from 1997 onwards, there is an almost a .2 C rise above the average top peaks of 1979-1996. Or, if by “normal” we were to refer to the UAH chart’s baseline then it would be .5 C. That amounts to 3/4 to barely 1 degree Fahrenheit.
RR Kampen says at 7:53 am:
“A new first for the region.”
Aren’t you forgetting 967 A.D., and 1016, an 1124, 1125 and 1126 A.D., when the Arctic was completely ice free?
bob says:
June 25, 2010 at 7:35 am
At what point are you skeptics going to admit that the AGW proponents made a prediction that came to pass?
A ‘prediction,’ you say?
Well, a broken clock is ‘right’ once a day.
So if you’re basing your predictions on broken mechanisms, then I’d have to say that your declaration is —given the current down-turn in global temps— morally broke and mentally bankrupt.
I see the word “melt” being used repeatedly in the comments, and yet it is safe to assume that none of the authors actually know how much ice is melting, and how much is simply compacting.
Extent is a 2-D number. If you shovel your driveway, the mass of snow remains constant- yet the extent gets smaller. A reduction in extent is not necessarily evidence of melt.
Why can Steve Goddard not include a similar graph covering Antarctica for comparison. If warm water is thrust north by El Nino from a fixed ocean heat positon (short term) it is reasonable to expect that Antarctica sea ice is getting a bit of a cool bath and increases sea ice such as is shown here http://nsidc.org/data/seaice_index/images/daily_images/S_timeseries.png with a current positive anomaly http://arctic.atmos.uiuc.edu/cryosphere/IMAGES/seaice.recent.antarctic.png
That might perhaps also put a bit of a “damper” on R Gates.
Just off the top of my head…
The gorons predicted that the Himalayan glaciers would disappear within a few years, that the sea level would start to rise fast, and that the ocean is acidifying.
They said toads are going extinct because of global warming, that Tuvalu is being swamped by a fast rising sea level, and that the West Antarctic peninsula will soon be ice free.
They said that GCMs can predict the climate, that the climategate emails are just misunderstood, and that CO2 will trigger runaway global warming.
They said that the Sun causes global warming, that the Sun doesn’t cause global warming, and that CAGW is a Theory.
They said that DDT thins bird egg shells, that global warming is causing hurricanes to increase, and that it causes more tornadoes, too.
They said the MWP didn’t happen, or if it did it was just a local event, and that the LIA didn’t happen, and that the Minoan Optimum didn’t happen.
They said that James Hansen’s predictions were accurate, and that Cap & Trade will reduce CO2, and that corals are bleaching and dying out from CO2.
They said there is a large consensus of scientists who believe in CAGW, and that global warming causes more forest fires, and that it causes malaria to spread, too.
They said that shellfish are losing their calcium due to CO2, and that NOAA’s temperature “adjustments” are A-OK, and that the climate peer review process is legit.
Now they are desperately trying to convince everyone that the Arctic is acting outside of its natural long term parameters because of CO2.
Based on their past record of getting everything else wrong, what are the chances that they’re right about the Arctic?
stevegoddard,
I got the current “melt” rate from
http://www.ijis.iarc.uaf.edu/seaice/extent/plot.csv
which is 64,637 km2/day for the last month.
Which is on a pace to near the low of 2007, is it not?
Which would increase the trend, no?
Do you have any evidence for more compaction this year than any other year?
No trend until about 1990. After that a falling trend, with increasing steepness.
and
No trend until about 1990. After that a falling trend, with increasing steepness.
Running a trend analysis of September sea ice data with data, instead of just eyeballing the graph, I get:
A downward trend of ~155 000 sq/km per decade from 1979 – end 1988.
Adding one year’s data….
A downward trend of ~280 000/km per decade from 1979 – end 1989
Which should indicate the vagaries of using short-term trends, but to continue…
A steeper downward trend of ~407 000 sq/km per decade 1979 – 1998
An even steeper downward trend of ~770 000 sq/km per decade 1979 – 2008.
Conclusion: Arctic September sea ice has been declining every decade since since 1979, and the decline has been accelerating.
I’ll wager this – if this year’s minimum falls above the trend line, NSIDC will mention it, either directly or by saying that the downward trend has decreased. I’ll take on anyone for $100 even bet. Of course, this is nullified if minimum extent falls below the trend line.