NASA scientists seek to improve sea ice predictions


Many melt ponds on sea ice north of Greenland, as seen during an Operation IceBridge flight on July 24, 2017. CREDIT Credits: NASA/Nathan Kurtz

Sea ice in the Arctic Ocean is in a downward spiral, with summer minimum extents about 40 percent smaller than in the 1980s. But predicting how the sea ice is going to behave in a particular year is tricky: There are still many unknowns about the conditions of the sea ice cover, to say nothing of the difficulties of forecasting weather and ocean behavior over seasonal timescales.

NASA researchers are working to improve their forecasts of the size of the Arctic sea ice cover at the end of the summer melt season — but the goal is not just to have a better prediction of sea ice coverage. The challenge of making summer sea ice forecasts allows scientists to test their understanding of the processes that control seasonal sea ice growth and retreat, and to fine-tune computer models that represent connections among the ice, atmosphere and ocean.

A grassroots initiative launched in 2008 compares the efforts of different research teams each year to predict end-of-summer Arctic sea ice. This year, three NASA teams are among over three dozen groups that submitted at least one prediction. The participating groups in the Sea Ice Prediction Network (SIPN) use different methods, ranging from statistical analysis to dynamic models, gut feelings and combinations of techniques. Teams can submit forecasts in June, July and August based on the current sea ice conditions at those times. Each fall, the project releases an analysis of the year’s submissions, reviewing lessons learned and providing guidance to future research efforts.

The comparison of different methods allows scientists to go beyond simply collecting observations of sea ice characteristics to test hypotheses about the drivers of sea behavior in a particular year. As an example, such methods will allow researchers to more closely examine the impact of summer ice thickness or the presence of melt water ponds at the beginning of the melt season. It will also let them determine whether a singular characteristic of the ice is more useful than analyzing a combination of factors.

Among other data, the participating teams use sea ice concentration — how much of a given area is covered in ice — and extent, which takes into consideration all areas of the Arctic Ocean and surrounding seas where ice covers at least 15 percent of the ocean surface. Higher sea ice extents and, more importantly, higher concentrations make the ice more resistant to melt, because less of the ocean is exposed and thus less able to absorb solar warmth. The concentration and extent datasets are based on methods developed back in the 1970s and 1980s by scientists at NASA’s Goddard Space Flight Center in Greenbelt, Maryland, and currently curated by the NASA data archive at the National Snow and Ice Data Center (NSIDC) in Boulder, Colorado.

In addition, a long-running NASA aerial survey of polar ice called Operation IceBridge provides measurements to SIPN teams on spring sea ice thickness, melt ponds and snow conditions, which are three other parameters that influence the summer melt season.

Alek Petty, a sea ice researcher at Goddard, has submitted forecasts to SIPN for the last two years that are based on a statistical model that analyzes satellite measurements of sea ice concentration and melt onset. One aspect of the summer melt season Petty tests with his model is the question of how melt ponds — the bright blue puddles of melt water that appear on the sea ice in the spring and summer — can also help predict of the extent of sea ice at the end of the summer. Petty can run his model with data on the spring melt pond coverage obtained by a climate model run by researchers in the United Kingdom.

“One of the difficulties with the melt pond model is that we don’t have many good observations of melt ponds in the Arctic for validation,” Petty said. “That’s what Operation IceBridge is now trying to provide with their recent summer sea ice campaigns.”

Richard Cullather, a climate modeler at Goddard, uses a dynamic model of the climate system that takes into account the interactions among Arctic sea ice, the ocean and the atmosphere. This year, his fourth participating in SIPN, he applied sea ice thickness measurements from satellite data to define the conditions of the sea ice pack before the beginning of the melt season. Cullather expects that this addition will improve the forecast.

“It seems that the ice that remains in the late summer is a function of how thick the ice was to begin with,” Cullather said. “We ran our forecast in June without the sea ice thickness and it came with a much higher prediction than when we actually introduced the thickness measurements. That suggests that adding in the thickness observations is changing our values.”

NASA IceBridge data, along with measurements from ESA’s (European Space Agency) CryoSat-2 satellite, provide key thickness information. With NASA’s Ice, Cloud, and land Elevation Satellite-2 (ICESat-2), a mission scheduled to launch in 2018 that will measure the height of the sea ice above the ocean surface, improved thickness information will become available to inform melt predictions.

Finally, Walt Meier, a sea ice scientist who worked at Goddard until July and is currently with NSIDC, uses a simple statistical method that looks at the total extent of sea ice from 2005 to 2016 and averages the daily changes in extent day by day through the end of September.

“I only use data from the past 12 years because before 2006, the edge of the sea ice pack at the beginning of the melt season was located much farther South, so it would melt more quickly in July and August than it does now,” said Meier, who has participated in SIPN since its beginnings and is a member of the leadership team.

Despite the sea ice pack starting the melt season from a record low maximum yearly extent in March, none of the predictions from Petty, Cullather and Meier call for this year’s September extent to be less than that of 2012, which is the lowest ever observed during the satellite record of Arctic sea ice measurements that began in 1979. Still, their forecasts are suggesting an extent in line with the Arctic sea ice’s long-term downward trend.

A common challenge to existing forecasts is that as the sea ice has dramatically thinned over the past decades, the impact of weather phenomena such as large summer cyclones on the ice cover has increased. Also, weather cannot typically be predicted more than 10 days in advance.

“Summer storms can have different effects,” Cullather said. “If the ice is reasonably thick, the storm can distribute the ice pack so it then covers a larger extent. If the ice is very thin and you spread it, it’ll allow it to melt faster because there’s more open water being warmed by the sun.”

“There’s more and more interest in seasonal predictions of the Arctic sea ice, but at the same time things are challenging because of the changing conditions of the ice pack and how it responds to weather,” Meier said. But improvements in the models and statistics, together with a growing record of sea ice measurements, make forecasters optimistic about the future. “The hope is that the models can outrun the changes in the Arctic, that we’ll be able to catch up and produce more reliable forecasts.”


0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
September 1, 2017 12:35 pm

Pen Hadow looks for a new assignment to evaluate downward spiral of the sea ice in the Arctic Ocean.

george e. smith
Reply to  vukcevic
September 1, 2017 2:47 pm

As I recall in the late 70s and early 80s, they were so concerned about global cooling and ice ages, that they were hot to scatter soot on all of that sea ice to stop the ice age in its tracks.

Reply to  george e. smith
September 1, 2017 5:06 pm

a few nuts thought that. not the majority.

Reply to  george e. smith
September 1, 2017 7:27 pm

Steve, PIOMAS is a model?

Reply to  george e. smith
September 1, 2017 8:08 pm

“a few nuts thought that. ”
Just like a few nuts, many of them the same people, think that ENHANCED ATMOSPHERIC CO2 is a problem…
…..which it isn’t !!
Crushed nuts and dodgy used car salesmen.

Reply to  vukcevic
September 1, 2017 3:18 pm

In terms of Arctic sea ice volume, it seems to be a peculiar downward spiral that doesn´t seem to go down:comment image

Reply to  Science or Fiction
September 1, 2017 5:08 pm

if so its a model.

Reply to  Science or Fiction
September 1, 2017 9:18 pm

Steve, Like RSS ESTIMATIONS of sea ice thickness? 😉

Reply to  Science or Fiction
September 3, 2017 3:00 am

Yeah I noticed that when Griff was going on about it. The data from 2008-2017 is basically flat with 2012 the obvious low standout. So I was wondering where all the decline came from and then when you look at the min max one you can see it is all historic 1980-2008. It definitely isn’t accelerating, I don’t have a problem showing it as a linear decline which would mean ice free in 30 or 40 years. There obviously must also be a chance it isn’t going to go any lower for a number of years given it hasn’t really moved in decade.

Reply to  vukcevic
September 1, 2017 3:21 pm

“Sea ice in the Arctic Ocean is in a downward spiral …”
Not a good start to improving ice predictions. First you need to honestly and accurately describe what has already happened. The Death Spiral is Dead
With current Arctic sea ice extent at the same level as it was a decade ago, the constant assertions melting at “unprecedented rates” or death spiral etc are becoming farcical .
Like the iconic polar bear, this particular canary just refuses to fall off its perch.

Reply to  Greg
September 1, 2017 4:07 pm

At the moment, Arctic sea ice is well above a decade ago, and most years since then.

Reply to  vukcevic
September 1, 2017 4:04 pm

The day 241 Arctic sea ice extent doesn´t seem to be in a downward spiral either:comment image

Reply to  Science or Fiction
September 1, 2017 4:07 pm

Yup. Five-year uptrend confirmed.

Reply to  Science or Fiction
September 1, 2017 5:10 pm

if so it cannot be used for year to year comparisons.
ur on roll

Reply to  Science or Fiction
September 1, 2017 7:29 pm

Steve, you going to say no to this too?

Reply to  Science or Fiction
September 1, 2017 9:33 pm

How does the same chart look using volume / concentration. That is, area plus thickness calculation.
It would make an interesting comparison.
Area is one dimensional and not a complete value.

Reply to  Science or Fiction
September 2, 2017 1:24 am

No , area is TWO dimensions, not one. volume is three.
Both are relevant in different ways, neither is “better”.
Area/extent is relevant to feedbacks since most depend upon the exposed surface: evaporation, insolation, outwards IR emissions. . Volume is relevant as an indication of shortage of energy.

Reply to  Science or Fiction
September 2, 2017 1:26 am

storage of energy !

Reply to  Science or Fiction
September 2, 2017 2:24 pm

Your comments are not correct.
A top down image of sea ice area is one dimensional. It provides no consideration of the thickness. It is area.
From one year to the next the area may be the same but the thickness may be on average 30% less. There is a considerable difference in both the energy required to reduce the volume, and the residual energy.
The key influence’s controlling area, also control thickness.
Therefore by only using area without consideration of thickness limits the understanding of volume loss per year and the energy required to affect the change.

Reply to  Science or Fiction
September 2, 2017 2:40 pm

if so it cannot be used for year to year comparisons.”
From masie faq:
“Use the Sea Ice Index when comparing trends in sea ice over time or when consistency is important. Even then, the monthly, not the daily, Sea Ice Index views should be used to look at trends in sea ice. The Sea Ice Index documentation explains how linear regression is used to say something about trends in ice extent, and what the limitations of that method are. Use MASIE when you want the most accurate view possible of Arctic-wide ice on a given day or through the week. ”
So MASIE gives the best information on the amount of ice at any given point in time but still the results aren’t comparable. And since it shouldn’t be done NSIDC presumably presents results for several years together from pure perversity:

Reply to  Science or Fiction
September 3, 2017 5:36 am

ozonebust, you obviously do not have enough scientific or maths training to understand what “one dimension” means.
One dimension is a scalar or linear measurement, eg this road is a mile long; 2 dimenstions gives an area for example: this field is 600 sqr yards. Area or extent is a two dimensional measurement, The fact that this is measured from above has no relevance. The volume estimations are also made from above and that is 3D.
The key influence’s controlling area, also control thickness.
That’s your assertion, I explained wyy both area and volume are relevant and why both are useful in different ways. You reply with unargumented assertion. I guess you must be a warmist and/or “climate expert”.

Reply to  Science or Fiction
September 3, 2017 5:40 am

The key influence’s controlling area, also control thickness.

” influence’s ” is a possessive, not a plural.
Not only are you scientifically illiterate you are literally illiterate.

Reply to  vukcevic
September 1, 2017 4:40 pm

Is downward spiral a new type of correlation? like linear, logarithmic, sigmoid and such?

Ian Magness
September 1, 2017 12:37 pm

SHOCK! If sea ice is either thinner or significantly further south, it melts more quickly.
Who knew?
More grant money please.

george e. smith
Reply to  Ian Magness
September 1, 2017 2:52 pm

No ! The more there is, the more surface area (most of it is under water) the faster it melts. The melt rate depends on how much surface area is in contact with that surface water that went up there from the tropics. We are talking of kg/s, not percent.

September 1, 2017 12:38 pm

Yes, by all means fine tune those computer models. Maybe one year we can see somebody actually sail a boat to the North Pole as has been predicted for how long now? While you’re at it, work on those models that predicted 5 degree global warming by the end of the century.

Reply to  daveandrews723
September 1, 2017 1:38 pm

Why not just have a computer model sail a predicted boat to the modeled ice free pole.

Reply to  rocketscientist
September 1, 2017 2:21 pm

Would you adjust the model based on the assigned number of captains on the modeled boat?

Reply to  rocketscientist
September 1, 2017 3:58 pm

Shhhh! You’ll just give them more silly ideas, and the press will be too stupid to see the silliness of it.

September 1, 2017 12:39 pm

The description “Downward spiral” is loose-talk over-the-top and alarmist. For sure there is a distinct downward linear trend – as can be seen herecomment image

Reply to  JCalvertN(UK)
September 1, 2017 1:45 pm

Except for the fact that the ice level in 1979 was an anomaly with exceptionally high ice levels.comment image

Reply to  RAH
September 1, 2017 1:55 pm

RAH, I have tried to source this Vinnikov et al. 1980 graph. I find it hard to believe sea ice “amount” could have been graphed this accurately pre-satellite. I would love to see the source and justification. Must have been created by temperature proxy estimation.

george e. smith
Reply to  RAH
September 1, 2017 2:54 pm

Yes the age of the coming ice age; 1976-79.

Reply to  RAH
September 1, 2017 3:41 pm

There were satellite observations in the 1960s and ’70s, just not by a sea ice-dedicated satellite.
And before that, ice edge observation maps were made by various services earlier in the 20th century.

Reply to  RAH
September 1, 2017 5:12 pm

another fake chart with no real data behind it.

[I call bullshit on Mosher, these are two charts from two separate scientific reports. One is Vinnikov et al. 1980, and the other is from NOAA data of the last 30 years, superimposed. There’s real data there even if presented in a way you don’t like. -Anthony]

Bill Illis
Reply to  RAH
September 1, 2017 5:15 pm

I note that Walt Meier used recovered Nimbus satellite pictures from 1964 and his number is bang on for the graph in September 1964.
There are also satellite-based estimates from 1972-1979 which are a somewhat higher than the above graph.
So it is more like the sea ice continually expanded through the cold 1970s which peaked at about 1979. I can link to the data if someone wants.

Reply to  RAH
September 1, 2017 5:24 pm

IMO ice expansion began in the late ’40s and proceeded through the ’70s, with of course large annual differences.

Reply to  RAH
September 1, 2017 7:32 pm

Steve Mosher, the Vinnikov paper contains the chart, oh you didn’t know that came from a “peer reviewed” and published paper?
Why don’t you stop being an insufferable snob?

Reply to  RAH
September 2, 2017 5:18 pm

I like it when Anthony speaks his mind 🙂

Reply to  RAH
September 2, 2017 8:49 pm

Here’s the source of the Vinnikov graph. DOE report (1985), page 152, Figure 5.2
Arctic sea ice extent dropped 15% from 1925 to 1945. Similar percentage drop from 1979 to 2015. It’s cyclical

Reply to  JCalvertN(UK)
September 1, 2017 1:55 pm

Nice graph. Note that the decline in maximums is not large, it is the fluctuations in annual minimum that drive the average. And note that up to 1996 the minimums were flat, then a sharp drop to 2007, and flat since. And we know there was an unusual flux of old ice out through the Fram Strait starting about 1998.

Reply to  Ron Clutz
September 1, 2017 1:56 pm

My comment pertains to JCalvert’s graph.

Reply to  JCalvertN(UK)
September 1, 2017 2:05 pm

BS, her is Vinikov;
Your graf is a spleis made by The Great “Sea Ice Communicator”, AKA Tony Heller.

Reply to  Emeritus
September 2, 2017 2:56 pm

You cite the wrong paper (from 1999, not 1980) and can’t even spell “Vinnikov”, very convincing. The correct citation is:
Vinnikov, K. Ya., G. V. Gruza, V. F. Zakharov, A. A. Kirillov, N. P. Kovyneva, and E. Ya. Ran’kova, Recent climatic changes in the Northern Hemisphere, Soviet Meteorology and Hydrology, 6, 1 – 10, 1980.

September 1, 2017 12:51 pm

Agree that if anyone should be doing this, it ought to be NOAA, not NASA. But whoever wants to forecast summer Arctic sea ice should definitely not hire Griff. As his betters told him would be the case, he was, as always, as wrong as wrong could be.
Not only will there be no new record low this year, but 2012 will probably the low for the next 70 years or so, until the next low point in the 35-year sea ice cycle.
Past highs and lows of Arctic sea ice:
Low: 2012
High: 1977 (year of PDO flip)
Low: c. 1942 (free navigation of Siberian sea route, aka NE Passage)
High: c. 1907 (year btw Peary’s two NP attempts)
Low: c. 1872 (year after Hall’s Polaris expedition, first to attempt the NP)
High: c. 1837 (Franklin died ten years later).

Reply to  Gloateus
September 1, 2017 12:53 pm

Note difference in approach of Hall, who hoped to reach the Pole by sea, and Peary, who went over the ice.

Reply to  Gloateus
September 1, 2017 3:15 pm

Let’s not forget Clarkson 2007 who did it by car.
[That attempt was successful, but remember, they drove to the Magnetic North Pole, not the geometric north pole. .mod]

Reply to  Gloateus
September 1, 2017 12:54 pm

Guess I should have said next 65 years, since we’re already five years into the waxing Arctic sea ice cycle.

Reply to  Gloateus
September 1, 2017 11:01 pm

Yes, it’s early days but I suspect that OMG. OMG low in 2012 was the turning point. That is bracketed by the OMG lows of 2007 and 2016.comment image
I think the next five year span will show a positive trend.

Pillage Idiot
Reply to  Gloateus
September 1, 2017 2:50 pm

Sorry Gloateus, I disagree with your NOAA hiring recommendations.
Griff has a nearly perfect record of predictions. The fact that he is essentially always wrong has great predictive value, so good that NOAA should definitely hire him. We can easily take his Arctic sea ice predictions, and then plug the opposite results into the models for improved accuracy.
\sarc off

Reply to  Pillage Idiot
September 1, 2017 3:01 pm

Of course you’re right. On the Street, the 100% loser would be worth his weight in blue chip securities, as a perfect counterindicator.
But to be of use to NOAA, he’d have to make sea ice predictions more precise than “a new record low is a sure thing”.

September 1, 2017 12:54 pm

Maybe I missed it but I didn’t see any mention of measuring current inflow from the N. Pacific or Atlantic. Back when the “catastrophic” melting began there were a couple of “peer” reviewed journal articles that fairly well documented that the ice began this period of melting due to warmer water inflows from the North Pacific. (associated with PDO??) As I understood it that set up a loop. More warmer water meant more melt and exposed sea surface. Then even after the warmer water began to retreat the ice would continue to melt do to the amount of exposed sea surface. The authors speculated that there was some mechanism that would ultimate reverse the ice melt or it could be just a given amount of time after warmer water stopped flowing into the Arctic Ocean.

Reply to  Edwin
September 1, 2017 12:57 pm

It’s the PDO and AMO. Arctic sea ice is cyclical, although general global climate plays a role, too.
But since the mid-19th century global climate has been in a warming cycle, so the current 35-year waxing and waning is liable to continue at least until the next LIA cooling.

Reply to  Edwin
September 2, 2017 2:02 pm

I think that surface wind flows are of importance to understanding sea ice changes, along with temp changes on both sides of the Arctic/oceans connection. Greenland temps, for example, starting dropping at the end of the second week of August. That was partly due to shifting wind patterns south of Greenland. As a result of those wind changes the Greenland SMB has once again soared as the melt ended on August 15th, early as I had forecast back in late June of this year that there was a strong probability that the melt would end early from other indicators which I view daily. In this case I noted a cold footprint which set in early around late June in upper central Siberia. …,57.51,581/loc=94.643,62.883
This new Siberian cold spot sits around 30 degrees of longitude closer to Europe than in the past 4 years of my observing. Which leads me to wonder if that means that this upcoming winter’s cold front will extend an equal distance or greater to the ever increasing deep cold spells which have struck that region of the world in the last 3 winters. Last winter the edge of -10F or lower reached eastern Poland. A shift of 30 degrees further west would bring minus temps into France. The changes in the Greenland SMB over the last 3 winters are another clue to the ongoing climate shift, imo, which I see see as a probable herald to a moderate cold spell setting in around the NH. …

Reply to  Edwin
September 3, 2017 9:10 am

I am pretty sure from some solidly based past papers that the change in any one year is dictated primarily by winds and currents. Sorry I don’t have references. Might even have come from NASA/NOAA. That means that trying to predict using past years’ numbers is a useless way of predicting this year’s number.

Ed Zuiderwijk
September 1, 2017 12:56 pm

And just by the time they figured it out the Arctic ice will be returning. By 2030 its extent will be like back in the 1970s. Which makes prediction of its extent extremely easy and reliable.

September 1, 2017 1:02 pm

Of what we can be sure however is that having now four instead of three, as a century ago, molecules of CO2 per 10,000 dry air molecules has no detectable effect on Arctic sea ice extent.
CO2 was rising during the high ice decades of the ’50s, ’60s and ’70s, and has continued rising since 2012, as the ice has recovered from its record low that year.
By the same token, Antarctic sea ice grew alarmingly from 1979 to 2014 during steadily rising CO2. Then came the 2015-16 Super El Nino, which, along with other weather events, led to lowered SH sea ice.
Ocean oscillations control sea ice. CO2, not so much. As in, not at all.

September 1, 2017 1:02 pm

Mixed emotions about the science because the duplication of effort suggests overall too expensive and wasteful. NSIDC is in Colorado under NOAA, not NASA. It is the lead repository for all NASA satellite cryosphere data. And responsible for NOAA sea ice research. Why are NASA scientists working on duplicative Arctic sea ice models that are the primary responsibility of NSIDC? In my corporate world this would never be allowed.

Reply to  ristvan
September 1, 2017 1:30 pm

Hi Rud
” based on a statistical model”
Could it be that Gavin Schmidt a climatologist, climate modeler (applied mathematics) is the Director of the NASA Goddard Institute for Space Studies – GISS (New York division only?)

Reply to  vukcevic
September 1, 2017 1:46 pm

you mean the person hand-picked by DR Doom to carry-on his ‘good work ‘

Alan Robertson
Reply to  vukcevic
September 1, 2017 1:52 pm

Don’t you see? The whole effort has required someone (Schmidt) who was willing to make and repeat such pronouncements as “Hottest year/month ever recorded…” and to run a blog supporting CAGW that disallows any dissenting viewpoint, whatsoever.
Gavin Schmidt is just such a man.

Reply to  vukcevic
September 1, 2017 4:12 pm

He’s a computer programmer, not a climatologist.

Wayne Delbeke
Reply to  vukcevic
September 1, 2017 9:12 pm

Gloateus: And apparently not too good at either.

Reply to  ristvan
September 1, 2017 4:11 pm

The public has more respect for NASA than NOAA. That NASA scientists believe in “climate change” is right up there with the Big Lie of 97% as a substitute for thought and research.
Except of course for the fact that 49 NASA retirees signed a letter in 2012 objecting to the politicization of the agency in the cause of CACA.

September 1, 2017 1:05 pm

It should be NOAA’s job. Pretty soon NOAA will be launching noaanaughts to Mars.

September 1, 2017 1:09 pm

Arctic sea ice conditions HAVE been improving since the extreme levels of the late 1970’s.
Unfortunately, the recovery back downwards to more normal Holocene levels seems to have stalled.
There would be huge benefits for fishing, commerce, travel etc etc from having less sea ice

Reply to  AndyG55
September 1, 2017 1:14 pm

Biodata shows that current levels are much higher than for most of the Holocene.
And that is the main problem for the people living up there..
I mean , look at the massive effort that the Russians have put in to try to make the region navigable for even a few month. These new ice-breaking oil tankers. Nuclear powered icebreakers that can actually melt the ice around them.

John Bell
September 1, 2017 1:09 pm

What difference does it make, anyway?

Reply to  John Bell
September 1, 2017 1:14 pm


Reply to  John Bell
September 1, 2017 1:16 pm

Lower sea ice levels would be a massive benefit… that’s what difference it would make.

Reply to  AndyG55
September 1, 2017 2:30 pm

More specific question:
What difference does it make if an annualized prediction accuracy is improved from -/+ 15% to +/- 3% ?

Reply to  John Bell
September 1, 2017 9:45 pm

John Bell
It means that you start to understand the mechanisms that control sea ice area and thickness. Once established it will be another nail in the practice of using CO2 as a cause for all things that change from how they used to be.

September 1, 2017 1:17 pm

Again, what we should be asking is
“What SHOULD the level of Arctic sea ice be” ? give reasons for your answer.

Reply to  AndyG55
September 1, 2017 3:53 pm

Depends on whether or not some yutz is trying to sail to the Pole. If so, there should be just enuf ice to trap him (and his 3 co-captains) 500 miles away from his stated goal and hold him ’till rescue.
Otherwise it should be whatever Griff seys it be.

Dr. Deanster
September 1, 2017 1:20 pm

it’s all about the Money Money Money …. don’t worry bout the price tag …..

September 1, 2017 1:30 pm

The arctic is screaming!. Oops. Mark Serreze, head of NSIDC, already said that.

Reply to  SkepticGoneWild
September 2, 2017 3:07 am

Quick, grant money before it melts. Seth Borenstein @AP told us in 2007 that ‘climate prophet’ James Hansen says it’s gonna melt before 2019, so we have only one season to go before there is no ice to research. 🙂

September 1, 2017 1:40 pm

“The challenge of making summer sea ice forecasts allows scientists to test their understanding of the processes that control seasonal sea ice growth and retreat , and to fine-tune computer models that represent connections among the ice, atmosphere and ocean.”
Based on so many forecasts that the ice would be gone by now I would have to say that their understanding of the processes is near nil and their computer models need more than a fine tuning. How many Wadhams are there up there now?

September 1, 2017 1:44 pm

Lets us have a guess the ‘models ‘ will forecast things getting worse , which by ‘lucky chance’ is a good match for those using them . Now where is my PHd and fat research grant ?

September 1, 2017 1:51 pm

Arctic ice melts from the bottom up and freezes from the top down.
Indicate Gloateus’ highs & lows of Arctic sea ice on approximation of net effect on avg global temperature of SST oscillations, Fig 4 in . When SST is up, Arctic sea ice is down. Time between low and high on Fig 4 is 32 yr vs 35 yr for measured ice. It will be interesting to see what NASA predicts.

September 1, 2017 1:53 pm

Really? Actual scientists asking for “gut feelings” for summer sea ice predictions?
Why not just ask a gypsy as she peers into a crystal ball?

Gary Pearse
Reply to  AZ1971
September 1, 2017 4:51 pm

WUWT used to put in a blog commenters estimates and they were among the best.

Bruce Cobb
September 1, 2017 1:59 pm

Clearly their agenda isn’t science, though they pretend that it is. Arctic sea ice is the last bastion of Alarmist pseudoscience.

Stewart Pid
Reply to  Bruce Cobb
September 1, 2017 2:27 pm

Re alarmist pseudoscience cue the griffster 😉

Mark - Helsinki
September 1, 2017 2:24 pm

“It seems that the ice that remains in the late summer is a function of how thick the ice was to begin with,”
Oh sweet frikkin jebus!!
he said this, I cant believe he actually said remaining ice depends on how much you started with.
rant seeking losers.
USGS admit openly they know frikkin next to nothing about arctic tides and currents.

Mark - Helsinki
Reply to  Mark - Helsinki
September 1, 2017 2:24 pm


Reply to  Mark - Helsinki
September 2, 2017 2:27 pm

Your original also works

Reply to  Mark - Helsinki
September 1, 2017 3:46 pm

AND, he didn’t even start with the initial condition!!
He said he started with a bunch of other variables and got an output that was higher than reality. So then, AFTER introducing the initial condition as a part of the model, they gots a gooder answer.
(I need to know how much fuel I need to drive to Vancouver BC (850 miles) this weekend. I am going to base it on the mileage that I got last time I traveled eastery to Great Falls, Montana. My odometer read 153,002 when I got to Great Falls, and I used 52.537 gallons of gas to get there; how do I model & project my Vancouver fuel needs?)

J Mac
September 1, 2017 2:25 pm

“The hope is that the models can outrun the changes in the Arctic, that we’ll be able to catch up and produce more reliable forecasts.”
What the heck does that gobble-de-goop mean? Buehler? Anyone?

Mark - Helsinki
September 1, 2017 2:25 pm

they have no idea what they are doing, the fact they cant agree on methods or predictions makes that clear

Reply to  Mark - Helsinki
September 1, 2017 2:37 pm

Good observation. They are being paid to play.

J Mac
September 1, 2017 2:35 pm

NASA-Goddard Space Flight Center should focus on ‘Space Flight’ and leave earthly sea ice measuring and modeling to NOAA-NSIDC. This is yet another example of Waste, Fraud, and Abuse within the the US government agencies. The funding for this activity at NASA-Goddard should be stripped and the associated employees reassigned or laid off. The manager(s) responsible should be fired.

September 1, 2017 2:50 pm

I’ll give them a slight clue:

September 1, 2017 2:53 pm

Here is the arctic cap as it appears today from cryosphere U of Iliinios
Here is the image from the same time in 1980
Yet the graphs show that extent today is Significantly less that the average. The image of today clearly shows the extent and thickness to be much higher……what am I missing

Reply to  Jamie
September 1, 2017 3:01 pm

The first chart is doesn’t match the September 1 date.

Reply to  Sunsettommy
September 1, 2017 3:11 pm

One day. Of no consequence. Thanks for the comparison, Jamie. Had not known of that additional resource.

Reply to  Sunsettommy
September 1, 2017 3:31 pm

ristvan, the August 31,2017 chart is wrong. Cryosphere has not been updating the current sea ice chart for a long time now.
There is currently LESS sea ice now than in August 31, 1980.

Bob boder
Reply to  Sunsettommy
September 1, 2017 5:29 pm

Unfortunately their charts are messed up and your 2017 is not correct.

Reply to  Sunsettommy
September 2, 2017 2:20 pm

What is 9/1/1980 or 9/1/2017 September 1 (US) 9th January (EU).
I think there is a mix up with dates.

Reply to  Jamie
September 1, 2017 3:05 pm

“Special Sensor Microwave Imager and Sounder (SSMIS) on the Defense Meteorological Satellite Program (DMSP) F-17 satellite that provides passive microwave brightness temperatures (and derived Arctic and Antarctic sea ice products) has been providing spurious data since beginning of April. Working on resolving problem or replacing this data source.”
The first chart is probably for month of March.

Reply to  Sunsettommy
September 1, 2017 3:31 pm

Yeah…there must be something wrong with image
Here’s Aug 2013…..much different…..although it not march…..must be the latest image…but the satellite is giving poor data

Reply to  Sunsettommy
September 1, 2017 3:44 pm

The first chart is probably for month of March.

The Cryosphere reports from University of Illinois Urbana were providing spurious (unreliable jumps of 1-3 Million sq Km’s per day) in March and April 2016, then went down completely in mid-April 2016. They have been providing bad anomalies (false area reports) automatically since then. No person at Cryosphere is now studying ice mass, ice area, ice anomalies this scholastic year.

Reply to  Sunsettommy
September 1, 2017 7:26 pm

Scratch that source.

Reply to  Sunsettommy
September 1, 2017 7:37 pm

RACookPE1978, thanks for the clarification. I wasn’t sure if it was this year or last year.

Reply to  Jamie
September 1, 2017 3:21 pm

Yes. Arctic sea ice is lower now than in 1980. But is higher than 2007, 2011, 2012, 2015 and 2016. Soon it will probably be higher than 2008 and 2010, too.
Arctic sea ice fluctuates naturally in about a 35-year cycle. The cycle from 1977 to 2012 was declining, following the PDO shift in ’77. Now Arctic sea ice is gaining. No climatic trend lasts forever.
But even at the current low, Arctic sea ice extent is higher than it has been for most of the Holocene interglacial, ie the past 11,400 years. There is no human fingerprint on Arctic sea ice, except maybe for soot.

Reply to  Jamie
September 1, 2017 3:33 pm

Well Cryosphere Today is now Cryosphere Yesterday. Their sea ice extent graphs have been off the wall since one of the satellites when tits up.
They never bothered cross-calibrating another source of data and keep producing BS graphs using broken data. I don’t know whether this applies to the pics you posted but it needs to be checked.
If they cannot get the total ice area sorted out , there is little chance they have such maps based on sound data.

Reply to  Greg
September 1, 2017 3:34 pm

Anthony removed the CT ice extent from the sea ice page a few months ago since it had been broken of about a year and a half.

Reply to  Greg
September 2, 2017 2:17 pm

The steady deterioration of many sources of climate data available online, hints strongly at an approaching end of the current warming period.

Reply to  Jamie
September 1, 2017 3:37 pm

Jamie, that first graph is incorrect. It shows ice in Kara Sea which sats show is currently ice free. Same in East Siberian and Laptev Seas. Something amiss at Chryosphere Today!!

September 1, 2017 2:54 pm

The idea of a downward spiral is complete and utter fraud. They pick a high point for ice and work over a few decades – it is well known arctic temps in the 1930’s were almost the same if not warmer (NASA’s own data – ) , and we have tons of observations that the melts were similar to much worse – it’s all smoke and mirror bullcrap. Yes, when we have a natural warming cycle, we have more ice melt – especially at the polls as tropical heat moves toward the poles and is dissipated in a well known process. The arctic melt season this year was short lived, and inline with the average melt during this naturally occurring warm period. The bizarre thing, if you overlay a map of Ice now with ice cover from national Geographic in 1971, during the ice age scare, the amount of extra melt, is not at all alarming, considering natural warmer temps in the arctic ( – meanwhile, Greenland is through it’s melt season and is challenging records for gains in ice. The billions wasted on Arctic research alone is disgusting – it’s like an exercise in watching grass grow….cold periods more ice, warmer periods less ice.. WOW.. they have spent trillions since about 1989, to try and convince us the arctic is disappearing, and not simply cycling – if the evidence is so damn conclusive (a joke), then we can’t prevent it anyway – so why more F-ing studies?

September 1, 2017 3:02 pm

“The scientific method” – Well yes, that is making observations, drawing conclusions and formulating theories. Give NASA, say, a thousand years – by then they might have enough observations to be able to find a discernible pattern.

Reply to  AndyE
September 1, 2017 3:56 pm

I disagree. They seem to be narrowly focusing on forecasting minimum summer sea ice extent based on observations made during the previous winter/spring. This isn’t an exercise to predict decadal sea ice behavior for example. This seems like a relatively plausible endeavor.
What caught my attention was this quote:”The challenge of making summer sea ice forecasts allows scientists to test their understanding of the processes that control seasonal sea ice growth and retreat, and to fine-tune computer models that represent connections among the ice, atmosphere and ocean.”
Isn’t there a logical disconnect between this statement and the constant drone we hear as to the ability of the GCM’s to accurately simulate the way the real climate behaves? If the ability of the GCM’s to accurately represent the “connections among the ice, atmosphere and ocean” is something in need of testing for the purpose of a relatively trivial task of predicting how much ice will melt in a season, when you already know ocean conditions (el nino – la nina, atlantic oscillation, etc.) along with starting sea ice depth, etc., how can you simultaneously be confident that you have sufficiently modeled the “”connections among the ice, atmosphere and ocean” to accurately represent changes in climate over decades and in response to a change in a single input (CO2)?

Bob boder
Reply to  AndyE
September 1, 2017 5:21 pm

My Mark 1 eyeball looking at simple graphs has told me that low winter extents don’t ussualy mean that your are going to have a low summer extant. Yet NASA hasn’t figured it out nor has Tony McLeod, seaice and Griff, no matter how many times I have told them. Well Tony has probably finally learned but due to our bet he shouldn’t be around to tell us so.

michael hart
September 1, 2017 3:12 pm

NASA scientists seek to improve sea ice predictions.

Making them worse would be more of a challenge.

Reply to  michael hart
September 1, 2017 3:38 pm

Since they start by assuming that Arctic sea ice is going to keep on declining indefinitely, they’re sure to be wrong.

Reply to  michael hart
September 2, 2017 2:15 pm

Making them worse would be more of a challenge.

September 1, 2017 3:45 pm

Fire the prognosticators who get it wrong. Seriously why pay them otherwise?

September 1, 2017 4:12 pm

They have been doing the Arctic sea ice prediction since 2008, almost 10 years, and there hasn’t been much progress. In July the average prediction was 4.50 million km2 for the September average. For the August prediction it increased to 4.54.comment image
The problem is the ice is already above the 2016 level of a year ago, and very likely to end that way. So it is another year when most of the models have failed… again.
It is clear that the models that are performing better are the statistical ones, and this should be worrisome to the scientists. If the physical models are performing worse than the statistical ones, the logical conclusion is that they don’t understand very well what are the driving factors for the melting. That explains why the prediction exercise is not going too well after 10 years and the melting prediction is so poor.
Here is the correlation between AMO and September Arctic sea ice. AMO is probably one of the factors that influence Arctic sea ice melting, or both depend on the same factor, as the relationship has been confirmed by scientists. It is interesting that the correlation breaks down below 5 million km2, probably indicating that very low levels of Arctic sea ice are more at the mercy of weather elements than anything else. This also complicates very much Arctic sea ice predictions.

Reply to  Javier
September 1, 2017 4:15 pm

True. For models to work, they need to predict August and September WX a year in advance.
Low sea ice years are due to late summer cyclones. Also El Niños.

Bill Illis
Reply to  Javier
September 2, 2017 5:27 am

Right now, in 2017, the Arctic sea ice extent is a good 0.5M km^2 higher than 2016.
The basic numbers say the minimum (under the sea-ice-outlook methodology) will be around 5.2 M km^2.
That is quite a bit higher than one would have said a month ago but August 2017 has seen a very low rate, particularly late August when the ice extent actually increased for four days in a row.
Tough to say this low melt rate will continue throughout September but only the very highest sea ice outlook predictions above are going to be close.

Bill Illis
Reply to  Bill Illis
September 2, 2017 5:30 am

Sorry, I meant to say the September ice extent will be 5.2 M km^2, not the minimum daily number which could be around,4.8 M km^2 (noting there is 225,000 km^2 added due to the north pole hole in satellite coverage).

September 1, 2017 4:18 pm

“NASA scientists seek to improve sea ice predictions.”
Uh, oh. Wasn’t that what they said about surface temperature readings?

Steve Oregon
September 1, 2017 4:33 pm

“NASA researchers are working to improve their forecasts of the size of the Arctic sea ice cover at the end of the summer melt season……………..
“The hope is that the models can outrun the changes in the Arctic, that we’ll be able to catch up and produce more reliable forecasts.”
Color me curious but, WHY? Seems like a lot of tinkering busy work and expense with such little applicable value that no real purpose or need is even mentioned.
There are many other truly dire needs that are lacking attention and funding.
What would be the loss if all of the ice measuring, studying, monitoring of freeze/melt cycles and forecasting came to an abrupt halt?
Who would miss it but those living off the busy work?

Gary Pearse
September 1, 2017 4:41 pm

Why should I be nervous when Nasa and company want to improve sea ice extent measurement? They ‘improved’ sea level measurements when, during the Pause, SLR was flattening. They added in an ocean basin elastic rebound factor to account for decreasing volume of the sea basins. This means that official sea level is now some height above physical sea level. If they wanted to use sea level to determine how ice melt is effecting sea level, they should have SUBTRACTED this factor to obtain what ice melt is contributing, but they should not change a metric that doesn’t have a physical existence.
Oh, also the abandonment of high tech sea temperature buoys in favour of ships intake measurements and night sea air temperatures to ‘correct’ the record that gave us the Pause!
Oh, and they changed the Saffir-Simpson Scale which made Harvey a Cat 4 instead of an old fashion Cat 3, which they measured from an airplane that higher up has a stronger wind. They are supposed to measure at 12 feet above ground.
Oh, and although all state record high temperatures and number of days above 100F were set in the 1930s and 40s and the aggregated annual temperatures for the lower 48 were higher than 1998 and still were by 2007. This was fixed by GISS’s Hansen by pushing the earlier record down by about 1.5C!! Later it was found out that similar charts for Canada, Greenland Iceland, Siberia, South Africa, Paraguay… also supported the older US temperature pattern and they were similarly pushed down.

September 1, 2017 5:23 pm

“Improve sea ice predictions”?
What beneficial purpose does predicting sea ice minimums perform?
Reminds me of similar games; “blind man’s bluff” and “pin the tail on the, er, donkey”.
Only, when these games are played by people of mature years; they tend to be drunk or senile.
These characters are not practicing science. They’re playing model games, glory and rent seeking based on belief and wishful thinking.

“Meier said. But improvements in the models and statistics, together with a growing record of sea ice measurements, make forecasters optimistic about the future. “The hope is that the models can outrun the changes in the Arctic, that we’ll be able to catch up and produce more reliable forecasts.”

It is sad to see what ex-scientists become when they cease practicing science.

Reply to  ATheoK
September 1, 2017 5:25 pm

It makes work for federal bureaucrats.

Reply to  ATheoK
September 3, 2017 9:53 pm

“Forrest Gardener September 2, 2017 at 12:15 am


why you ask? Falsifiable predictions are the only real way to advance knowledge of the natural world. Restoration of the good name of science is worthwhile cause in its own right, especially given the damage fake science has caused in that area. Reliable predictions assist in trade and commerce. I’m sure I can come up with more.”

My questions were:

“ATheoK September 1, 2017 at 5:23 pm

What beneficial purpose does predicting sea ice minimums perform?”

How are “sea ice minimums” any sort of “falsifiable science”?
Sea ice minimums are a pretense at calculating the percentage of sea ice cover in a grid.
As has been demonstrated in the past, simple Arctic storms easily rearrange sea ice, drastically changing area and thickness.
That makes guessing a “sea ice minimum” an attempt tp prophecy random factors; all too often using “gut feelings and beliefs”, not following practical repeatable ice forecasts.
Guessing sea ice minimums may make for fun wagers, but the current state of sea ice science as practiced by the sea ice climastrologists is fakery; which is why the sea ice fakirs get away with their alarmism sophistry.
Sheesh! Even way back in high school, we measured ice by volume when calculating heat exchange; not by visible surface ice.

September 1, 2017 5:55 pm

How to improve sea ice predictions:
Step One: Stop assuming the heat is from an imaginary source and research where the actual heat is actually coming from.

James at 48
September 1, 2017 7:16 pm

In terms of actuals, were are very close to this year’s minimum point. The area of 100% concentration is nearly equal to the overall area.

Steve R
September 1, 2017 8:06 pm

This passage bothers me:
“We ran our forecast in June without the sea ice thickness and it came with a much higher prediction than when we actually introduced the thickness measurements. That suggests that adding in the thickness observations is changing our values.”
Obviously someone wrote the code to include the models response to the thickness measurements. Why would he sound suprised when the model responded? The question should be whether the modeled response is accurate or not?
Am I missing something?

September 2, 2017 12:13 am

Been reading about Russian nuclear icebreakers. Now they have a nuclear LNG tanker with full icebreaker capability that can move through 2 (two) meter thick ice. Saw in a video that slabs of ice get thrown on to unbroken ice. Thick ice takes longer to melt. Wonder what effect use of these vessels will be on future arctic ice.

Donald Kasper
Reply to  Richard111
September 2, 2017 3:03 am

NW passage commercial ships with ice breaker capability may destabilize and reorganize ice flows by creating weak joints with no study of the consequences. They get to just get up and go run through it.

September 2, 2017 12:43 am

“But predicting how the sea ice is going to behave in a particular year is tricky”
There are a whole lot of things that we don’t know and things we will never know but our ego has gone so far ahead of us that we have an unrealistic expectation that we should know everything. It is for this readon that we can can no longer say “we don’t know that”. The best that we can do is “it’s tricky”.

Mr Julian Forbes-Laird
September 2, 2017 2:27 am

While on the subject of ice… NANSEN data hints at imminent minimum which looks like placing 2017 #5

Donald Kasper
September 2, 2017 3:01 am

NASA scientists seeks to upgrade climate ice models from useless vomit to just plain garbage as a first step.

September 2, 2017 3:21 am

When have we ever seen their previous forecasts published?
They do this so often. By announcing improvements, they are saying all before were bad.
For whom are they published? Polar bears can’t read.
Who would be foolish enough to believe them, anyway?

Steve from Rockwood
September 2, 2017 4:41 am

NOAA should bite the bullet and purchase the dancing monkey with the dart board.

Stephen Skinner
September 2, 2017 5:59 am

Looking at the profile of the last glacial maximums and minimums it will be hard to forecast in the short term (10s of years) something that only becomes apparent over thousands of years. It has taken around 50,000 years or more to reach glacial maximums and the route there was extremely erratic. Based on the knowledge we have so far we are already cooler than 5/6,000 years ago indicating that the next cooling phase has already begun. It may not be possible to confirm that we are cooling for at least another 1000 years as short term temp fluctuations can be interpreted as long term trends. We didn’t do too well predicting what would happen before the 2008 financial crisis and there were huge amounts of ‘clever’ people involved, or they were considered clever by people who considered themselves clever.

Bob boder
September 2, 2017 7:26 am

What amazes me, when watching the graph for arctic and Antarctic sea when the graph of one takes a significant uptick over the course of a day or two there is almost always the exact opposite reaction on the other graph. It zero sense to me but I have noticed it dozens of times, more than I think could be covered by meat coincidence, can’t fathom any possible connection but it there no the less.

Bob boder
Reply to  Bob boder
September 2, 2017 7:27 am

Omg auto spell argh!!!!

William Astley
September 2, 2017 7:51 am

Arctic sea ice predictions? Come on man. This is surreal. We are so caught up in the pointless fight no one has bothered to work out the scientific and practical implications of the alternative paths.
The warmists of course ignore the fact that there has been a lack of warming for the last 20 years which is one of a more than a dozen independent analysis results that disproves their hypothesis that the warming in the last 150 years was primarily caused (CO2 is assumed to be responsible for more than 50% of the warming in the IPCC reports, 100% in all warmists discussions) by the increase in atmospheric CO2.
Basic fundamental signal analysis unequivocally shows the warming in last the 150 years is natural, caused by solar cycle changes.

The authors of the paper that is discussed in the above video clip explain their analysis and its implications.
They transferred the temperature data 1850 to present from the time domain to the frequency domain by a Fourier Transfer which is a standard technique to enable the researchers to determine the portion of the signal that is cyclically changing vs non-cyclic changes which in turn can be used to determine the nature of the forcing.
The increase and decrease in temperature is all periodic with the same periods as in the past. There is zero evidence in the analysis of a monatomic increase in temperature in the transformed data. If CO2 was the mechanism responsible for the warming there would be evidence of a monatomic increase at all frequency in the frequency domain which there is not.
The period of the oscillations in the planetary temperature data matches known solar cycle variance.
The analysis conclusion is all of the warming in the last 150 years is periodic, there is no evidence of any monatomic CO2 warming in the frequency domain analysis.
There are hundreds of papers that support the above conclusion.

September 2, 2017 6:07 pm

The best way NASA could improve Arctic Ice Extent predictions is to start adding 100,000 KM^2/yr in Ice Extent growth over the next 30 years to account for PDO and AMO both being in their 30-yr cool cycles form 2019.
This year’s Arctic summer Minimum will be roughly 1,000,000 higher than the 2012 record low, and will be ranked around the 6th lowest since 2007…
CAGW’s dire predictions of an “ice-free” Summer Arctic Minimum from 2012 has come and gone in peels of laughter and eye-rolling.
NASA and CAGW have become a laughingstock…

September 2, 2017 11:50 pm

National Aeronautics and Space Administration. WTF has that got to do with ice predictions (and volcanoes for that matter). Do these guys have a mission statement or do they just get to fart about with whatever they like?

JB Say
September 3, 2017 7:31 am

Why cherry pick the 1980s sea ice extent as “normal”.

September 3, 2017 8:04 am

“NASA scientists seek to improve sea ice predictions”

Reply to  Ryan
September 3, 2017 8:22 am

The gypsy wants her crystal ball back?

September 4, 2017 5:01 pm

well in australia we need we need to improve our weather predictions spring 5/9…?—Storm-Journal

%d bloggers like this: