Short to Medium Term Arctic Sea Ice Conditions Discussion

[werther]

Morning Vergent,

It wouldn't be a very large job to CAD both areas on your ASMR-2 rgb pics. I'm not sure though if the blue swaths represent the same as what I call "mesh-pack". As I recall, the actual structural situation 20 August last year was much more fractured than the blue parts in the ESAS sector of the CAB suggests.
I do think there was a fraction of MYI left out there, it shaped up in what A-Team started calling "Goats' Head". During winter, it traversed the pack, leaving all the ESAS sector FYI. In that sense, the lighter blues on your pic for this year might indicate some other properties (certain clouds? maybe traces of smoke?).
The Beaufort doesn't fit either; it shows a light blue swath West of Banks, where indeed a large intrusion of MYI occurred Jan-Feb this winter. But then, why doesn't it attach to the large, dark blue region north of the CAA?
Wrapping up; the '12 "mesh-pack" pinned out just West of Prince Patrick Island, showing clearly white, not blue on that pic.

[jdallen]

Yah, looking at both the Hycom concentration and thickness models have gone seriously sideways, concentration in particular, but the thickness looks bad as well.

Bremen looks rather odd as well, with huge areas of low concentration "appearing' through portions of the mesh pack.

Seems like the models collectively are suffering some sort of hysterical break down....

[Frivolousz21]

The blue on the RGB pictures isnt a representation of MYI.





It looks tied to atmosphere.

[Frivolousz21]

Look at how much the ice over the Laptev/Atlantic side is being hit by the 20KT WAA winds off the warm waters.



http://lance-modis.eosdis.nasa.gov/imagery/subsets/?mosaic=Arctic.2013233.aqua.4km


Expect massive drops the next 2-3 weeks.

We will see the ice line reach 85N in a couple days.

[Frivolousz21]

It's been well above freezing going on three days now.

This is expected for 3-4 more days at the very least.

[BornFromTheVoid]

BFTV,

I've been trying similar methods to look for a precursor to the 2007 to 2012 summer pattern, and to look for a reason why it failed this year. Could you give more detail as to how you came up with the shift in patterns you describe during the early summer?

Here' what I used back in May, but it's quite a bit different now.

The 8 criteria were
1.  An Arctic Oscillation of between -0.5 and -1.5 on average over the last 6 months November to April (using the April year for the composite maps)

2.  An NAO that has averaged below 0 over the period November to April

3.  An AMO that has averaged a value of between 0 and 0.25 from the November to April period

4.  A PDO that has averaged below 0 for the November to April period

5.  ENSO anomaly in the 3.4 region of between -0.5 and +0.5 over June, July and August (JJA)

6.  Northern Hemisphere Snow Cover over 1 million km2 below average in June

7.  A rising and positive QBO

8.  11 Year Average Sunspot count below 70


Here were the matches, I just used 6, 7 and 8 for the composite maps.


June prediction vs reality


July prediction vs reality


I didn't use any kind of weighting back in May, and the criteria were much more simple too (such as looking at the average AO value over the last 6 months, rather than the trend).
The Arctic would probably require slightly different steps to be taken too, like taking into account regional sea ice and snow cover. But that's something that can be worked on later.

Here's what the prediction for the Arctic would have looked like for June (using the weighting for the years also)


Not a perfect match, but could have been useful. Certainly very different to recent years.

[werther]

In a search for clear pics, I took out ASCAT for yesterday. This is what I got after some enhancing (taking out a lot of the midtones).



Now as I understand, ASCAT operates on moisture content reflectivity. You can clearly see the reflective “mesh-pack” stand out, the Pole-hole on the fringe, the bleak FYI into the ESAS.

What you can also see is the Atlantic sector almost completely fading out here. Just the washed-up, foam-like boundary ridges show up far North of Frantsa Yosefa.

To the south (right), two long loops of clouds stand out; one over the Fram, another over the Barentsz Sea. The mid-tone suppression faded almost all other clouds, so you can imagine the water vapour content these bands contain to appear on this pic.

While MODIS just makes us guess what’s going on under the relatively warm Southern cloud intrusion, ASCAT shows this event to be big within the context of this seasons’ melt.

[Jim Hunt]

I don't think this has been mentioned previously. There's a new Arctic Sea Ice News article, entitled "The Balding Arctic"

According to the NSIDC the sea ice in the Arctic is "A bit thin on top":

There are extensive areas of low concentration ice, even in regions close to the North Pole, atmospheric pressure and temperature patterns this summer have differed markedly from those experienced in 2012; cooler than average conditions have prevailed over much of the Arctic Ocean. By contrast, Antarctic sea ice is near a record maximum extent for mid-August.

[F.Tnioli]

...

... By contrast, Antarctic sea ice is near a record maximum extent for mid-August.

Which may be happening, at least in part, because Antarctic continental ice sheet's melt accelerates now. Because, in initial phases of new, accelerating, mode of melting, one of first things which happen in there - has to be forming of melt ponds and small lakes on top of ice sheet itself, during summer, and much of that water goes under and "lubricates" the bottom of coastal glaciers. Just like it happens in Greenland, same thing. At least, some glaciers are affected by this very much. Which leads to large acceleration of sliding of ice from land to ocean. And i think this effect - this accelerated movement of ice from land (Antarctic continent) into lower elevation (into the ocea) - i think it goes on well into winter, if not significantly year-round. It's just the sheer scale of Antarctic continent and masses of ice involved which makes it that more able to go on "dropping extra ice" even during winter; the bottom, where liquid water increasingly "lubricate" rockbeds, is well insulated, in many places, from cold air above, - and Earth's own heat is also a factor.

Not so long ago, one big piece finally went off: http://www.nasa.gov/content/antarctic-glacier-calves-iceberg-one-fourth-size-of-rhode-island/ . If things like this are happening more often / in larger scale, due to increasing amount of under-ice water in coastal areas of Antarctic, then it's no wander that global warming leads to growing ice extent around Antarctic.

This must and will change once coastal areas in Antarctic would lose all "easy to part" ice, though - together with increasing meltwater annual drop, at some point max (and min) sea ice axtent there will switch from stable/slowly_rising from rapidly shrinking, which is already the case in Arctic, which has times less total mass of ice (i mean sea ice plus land-based ice combined).

The process is noted, in some sense, even by sceptical science: http://www.skepticalscience.com/antarctica-gaining-ice.htm (green part, that is).

[Jim Hunt]

This is not reality, this is a model going to tomorrow tomorrow land.

What a difference a day makes - "The Naval Oceanographic Office switched from ACNFS with NOGAPS to ACNFS with NAVGEM on August 20, 2013". Before:



and after:



Which version do you suppose is closest to reality?

[wili]

Werther, the area you identify with your 'fat red line boundary' above seems to correspond fairly well with the purple highest-concentration area from this CT map:

http://arctic.atmos.uiuc.edu/cryosphere/NEWIMAGES/arctic.seaice.color.000.png

(Does anyone else see what looks like a bear's head facing east in the ESS?)

[werther]

Yes, it does in it's proportions, Wili.
But the CT high concentration area is easily compared to Greenland next door, so around 2 mkm2.
CAD produces just 1,1 mkm2 for the "mesh-pack" within the fat red line.

[jdallen]

This is not reality, this is a model going to tomorrow tomorrow land.

What a difference a day makes - "The Naval Oceanographic Office switched from ACNFS with NOGAPS to ACNFS with NAVGEM on August 20, 2013".

Which version do you suppose is closest to reality?

Can I choose "Neither"?  There's got to be a lot of head scratching going on at the NOO, trying to figure out how to bring their product to where it at least *resembles* what's happening. Their thickness maps dropped off the deep end as well.

Are everyone's models going through a nervous break down this year? 

[Neven]

(Does anyone else see what looks like a bear's head facing east in the ESS?)

It would go a long way in explaining the disappearance of the goat's head. 

Steven Goddard is going to milk that new ACNFS map until the udder has shriveled. 

[ChrisReynolds]

HYCOM thickness has been revised upwards back in series as well.

From one of my blog posts - now correct image used.


Latest version.
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn/nowcast/ictn2013072718_2013072800_150_arcticictn.001.gif

BFTV,
Thanks for the detailed reply, I've been up since 0400 and really need a night off, will read tomorrow.

[Nightvid Cole]

HYCOM thickness has been revised upwards back in series as well.

From one of my blog posts.


Latest version.
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn/nowcast/ictn2013072718_2013072800_150_arcticictn.001.gif

BFTV,
Thanks for the detailed reply, I've been up since 0400 and really need a night off, will read tomorrow.

This is not a revision. This is a comparison of a 2012 map to a 2013 one...

[Xyrus]

The new HYCOM makes the old HYCOM look like a paragon of modelling accuracy. I understand that it's a new and there will be errors, but it should at least have something to do with reality. HYCOM is showing 90% coverage on almost the entire ice pack when that is clearly not the case (MODIS, CT, etc.).

[F.Tnioli]

Could it though be not errors, but intended disinformation? I imagine that people who control HYCOM are well aware that:
 - Arctic ice will be gone in September in a few years, and in 2020s, several months of ice-free Arctic ocean will be annual norm;
 - Average Joe around the world is quite stupid, but not stupid enough to not realize that if whole Arctic melts, then it means massive warming is going on;
 - Average Joe does not fly to Arctic himself, nor has any trusted peers who do, therefore he's only able to know that Arctic melted if specialists tell him it did;
 - It is possible, with sufficient effort and money, to make some specialists to make data and reports which are very far from reality;
 - It is possible, with sufficient effort and money, to shut up the rest (at least in terms of any significant mass-media, large internet portals included).

Could it be that HYCOM "new" system is just one of first, relatively small, steps of mass-disinformation campaign, one which is only starting as of now, aimed to make public to keep thinking, in observable future, that Arctic sea ice is not a goner (while it actually would be)?

[Frivolousz21]

[werther]

Don't know what Friv was about to post above...

He posted today's MODIS on the "home brew..." thread and might intend to show his concern of what's been going on in the Barentsz-Kara sector here too.

He's right. As he reacted on Chris' call for refreeze six days ago. Major event over there. Like shown above using ASCAT.
We're on the brink of losing at least 300K under 15%. Partly melt of rotten debris, partly major compaction of the steerless floes.

This illustrates it's not over yet and what could have been possible in a melt conducive year...

ECMWF shows the steering to be cut off soon though.

[Frivolousz21]

[Frivolousz21]

The Euro is still pretty bad though.  No cold to write home about.  Lots of Russian side compaction.

[Frivolousz21]

jaxa shows it well.  They need to forgo any windsat and go to AMSR2 because there is open water to far North where windsat can't scan it.

This is far from over.

[ChrisReynolds]

HYCOM thickness has been revised upwards back in series as well.

From one of my blog posts...

This is not a revision. This is a comparison of a 2012 map to a 2013 one...

Sorry, correct image now in place.

[ChrisReynolds]

BFTV,

I've been trying similar methods to look for a precursor to the 2007 to 2012 summer pattern, and to look for a reason why it failed this year. Could you give more detail as to how you came up with the shift in patterns you describe during the early summer?

Here' what I used back in May, but it's quite a bit different now.

The 8 criteria were....

Thanks BFTV, that's very useful.

The general patterns of lows and highs is very good in your prediction vs reality. It might be worth trying something similar for the 2007 to 2012 years, and I have a long weekend without commitments.

[paulklem]

Over the last week the broken and dispersed ice pack has moved into the pack and parallel to the 180 meridian in relation to the pole, while ice floes have continued to open up west of the pole (looking at the Arctic Mosaic). This pattern is likely to persist for another 5-7 days. The hole that has opened up, will likely fill up with low concentration ice extent, widening in the process and moving to cover an area between 150E and 180E.

If the first couple of weeks in September revert to normal ice drift, then this low concentration ice will begin moving back toward the Fram, and the race will be on... Will the hole cover the pole, or re-freeze first?  We could see an area several hundred kilometers in area, with concentration less than 50% over the pole by mid-September.

The complete breakup of the pack above 85N, around the pole, tells the interesting story of this year's melt. Despite protests to the contrary, the fact that this area got hit repeatedly by strong storms this year, must be the cause. The weather wasn't good for insolation heat buildup during the top melt season; and so the heat that is opening up the pack, must be coming from the deeper water replenished by Ekman pumping. And the deeper water is holding heat from last year, and heat transported into the Arctic from the Atlantic.

[nukefix]

Now as I understand, ASCAT operates on moisture content reflectivity. You can clearly see the reflective “mesh-pack” stand out, the Pole-hole on the fringe, the bleak FYI into the ESAS.

It's a C-band scatterometer, so moisture content and surface roughness affect the backscatter a great deal.

To the south (right), two long loops of clouds stand out; one over the Fram, another over the Barentsz Sea. The mid-tone suppression faded almost all other clouds, so you can imagine the water vapour content these bands contain to appear on this pic.

Clouds should have only a negligible effect at C-band - something else is going on.

[BornFromTheVoid]

Thanks BFTV, that's very useful.

The general patterns of lows and highs is very good in your prediction vs reality. It might be worth trying something similar for the 2007 to 2012 years, and I have a long weekend without commitments.

I gave it a shot last night for predicting June 2007 based of the data up to April. The -ve AO appeared very strongly, but not the -ve NAO or any ridge over Greenland.
I'll try tweak some of the criteria during the day and see if I can get anything closer to reality.

[Nightvid Cole]

Now as I understand, ASCAT operates on moisture content reflectivity. You can clearly see the reflective “mesh-pack” stand out, the Pole-hole on the fringe, the bleak FYI into the ESAS.

It's a C-band scatterometer, so moisture content and surface roughness affect the backscatter a great deal.

To the south (right), two long loops of clouds stand out; one over the Fram, another over the Barentsz Sea. The mid-tone suppression faded almost all other clouds, so you can imagine the water vapour content these bands contain to appear on this pic.

Clouds should have only a negligible effect at C-band - something else is going on.

Then why are clouds plainly visible over open water areas on the images?

[nukefix]

Clouds should have only a negligible effect at C-band - something else is going on.

Then why are clouds plainly visible over open water areas on the images?

That must be the effect of winds on sea-surface roughness (at centimeter-scale), the main product from ASCAT is sea-surface winds after all. High winds -> high surface roughness -> high backscatter.

edit: so very low backscatter means either dead calm open water or a very thin layer of new ice, both of which create very smooth surfaces.

[ChrisReynolds]

Nukefix,

I think you're correct here, it's probably winds as the scatterometer does use sea surface roughness to deduce winds. In winter it shows the difference between first year and multi year ice because of changes in dielectric constant of the sea ice as it ages.

BFTV,

I'll do a blog post tomorrow about the ENSO and the 2007 to 2012 period. I got average plots for all of your individual factors, the ENSO years give a ridge over Greenland (but out to the Atlantic) with troughs around. This made me immediately think the summer atmospheric pattern might be due to ENSO. However I've not found a pattern in the ENSO data supportive of this idea for MAM and JJA averages. Furthermore taking June NINO3.4 SST as an example - that still doesn't support the idea of an ENSO role, 2007. 2008, 2010, and 2011 are in a cluster with 2013 in the quartile just below median, whereas 2012 and 2009 are in the top quartile. So at present I think the similarity between your two NINO 3.4 years and the summer pattern is coincidental.

Jim, thanks for stating this issue with HYCOM/ACNFS over at the forum:

Strictly speaking the US Navy haven't changed "the algorithm", merely "the atmospheric forcing" of an unchanged model, if their brief pronouncements on the matter are to be believed at least. That's been forced upon them by the Navy Operational Global Atmospheric Prediction System (NOGAPS) being phased out in favour of the Navy Global Environmental Model (NAVGEM).

It's worth re-stating that PIOMAS continues to use NCEP/NCAR as it's done since inception. I'm hoping they manage to get thickness data published  for the summer.

Re PIOMAS.
Michael Yorke has contacted me by email and pointed out that PIOMAS gridded thickness is available to two decimal places (the binary files are single precision):
ftp://psc.apl.washington.edu/zhang/IDAO/retrospection
If that's been discussed here I've missed it, which wouldn't be surprising as I don't read everything.

EDIT - you'll be looking for files: heff.txtXXXX.gz where XXXX is the year in question. The files give lat/lon location of the grid box followed by twelve months of effective thickness of that grid cell.

[wanderer]

http://www.iup.uni-bremen.de:8084/amsr2/arctic_AMSR2_visual.png

this looks worse than the data suggests

[Jim Hunt]

An animated GIF revealing (amongst other things) the increasing "Barents bite" over the last 5 days, assembled from the University of Hamburg's 3.125 km AMSR2 concentration images:

[paulklem]

Jim Hunt, that is an amazing image; could you put it up on the main thread Hole post comments.

I have been watching this section on the Mosaic carefully over the last month. The prevailing southerly wind driving the fractured ice side of the pack north parallel to the 180 meridian should continue this week, then by Friday/Saturday, a HP cell should move up and over the NP. The flow will reverse to a clockwise flow then, and the fractured ice area will reverse direction. The movement will depend on where the center of the HP sets up, and the proximity of LP cells, but these holes and polynyas should move near the NP, and open up more. We may even see big polynyas open up to the west of the NP as the wind pushed fractured blocks away from the relatively solid pack in the r04c03 tile in the mosaic.

[Jim Hunt]

Jim Hunt, that is an amazing image; could you put it up on the main thread Hole post comments.

Your wish is my command.  I've given up making predictions, but I'll watch with much interest to see how yours pans out!

Here's the current GFS prediction for Saturday, courtesy of MeteoCiel:

 

[helorime]

A pretty good view is shaping up on MODIS this morning. 

http://rapidfire.sci.gsfc.nasa.gov/imagery/subsets/?mosaic=Arctic.2013239.terra.1km

I do not know how much more melting can occur at the pole itself this late, but what is there... or rather not there right now is pretty dramatic.

[jdallen]

A pretty good view is shaping up on MODIS this morning. 

http://rapidfire.sci.gsfc.nasa.gov/imagery/subsets/?mosaic=Arctic.2013239.terra.1km

That is just soup - looks like the barrier between open water near Svalbard has blown out and we now have a large curving very low concentration area cutting into the central basin.  I've been struck all summer by the cloud cover at low altitude that always seems to appear over open water, and wonder what net energy transfer the condensation represents... transport to or from the ocean....

[Jim Hunt]

A pretty good view is shaping up on MODIS this morning. 

A sneak preview on Worldview provides some (rather hazy) visual confirmation of the picture painted by AMSR2:

[Shared Humanity]

A pretty good view is shaping up on MODIS this morning. 

http://rapidfire.sci.gsfc.nasa.gov/imagery/subsets/?mosaic=Arctic.2013239.terra.1km

That is just soup - looks like the barrier between open water near Svalbard has blown out and we now have a large curving very low concentration area cutting into the central basin.  I've been struck all summer by the cloud cover at low altitude that always seems to appear over open water, and wonder what net energy transfer the condensation represents... transport to or from the ocean....

I would think that open water would evaporate and form low clouds over the water. This is what happens in Lake Michigan. The effect in the summer and winter is lake effect precipitation. What does this say about energy transfer?

[jdallen]

A pretty good view is shaping up on MODIS this morning. 

http://rapidfire.sci.gsfc.nasa.gov/imagery/subsets/?mosaic=Arctic.2013239.terra.1km

That is just soup - looks like the barrier between open water near Svalbard has blown out and we now have a large curving very low concentration area cutting into the central basin.  I've been struck all summer by the cloud cover at low altitude that always seems to appear over open water, and wonder what net energy transfer the condensation represents... transport to or from the ocean....

I would think that open water would evaporate and form low clouds over the water. This is what happens in Lake Michigan. The effect in the summer and winter is lake effect moisture. What does this say about energy transfer?

I'll have to compare winds with water & SST's, I recon...

[Marc]

Jim,

"Hi Marc,

Quote from: Marc on August 18, 2013, 09:39:09 PM

    If ice loss is a result of heat transfer, and water has the capacity to carry more heat by a factor of 2-3, why is the air temperature a significant factor?  Especially considering the surface area is larger below the ice then above?



If you're interested in how sea ice melts I recommend taking a look at Donald Perovich's 2011 paper "The Changing Arctic Sea Ice Cover". If you're really keen check out some of the references too!"

Thanks for reference, I'm struck by a couple things, probably most clear in statement:

"Surface melt in
2007 was typical for the Beaufort Sea,
with melt starting in early June and
a total of 0.7 m of surface melt. The
difference in 2007 was the extremely
large 2.1 m of bottom melting. There
was a gradual buildup of heat in the
upper ocean in July and August. Bottom
melt rates increased throughout the
summer, reaching peak values of
0.1 m d
–1
in late August."

Was the melting in Beaufort analogous for the melting throughout arctic or just better measured?
The conclusion seems to be that increased melting overall was due to increased solar radiation from decreased cloudiness and albedo.  It would be hard to conceive that a 3 month melt period would inject enough heat to radically effect bottom melt.  Do you think it was heat from that season flowing under ice, or transfer of heat from other oceans?

Thanks for your thoughts and time

[Frivolousz21]

Euro is pretty favorable for ice loss.  Mostly through compaction.  It clears the Laptev region by a few degrees latitude or more face value.

[wili]

It looks like a high is now dominating most of the Arctic Ocean.

https://sites.google.com/site/arcticseaicegraphs/

Is that going to mean some melt, or is the sun already at a low enough angle that it won't make much difference?

[K.a.r.S.t.e.N]

I recently invaded the Blog (User Khaustein), pointing to 2m temperature anomaly forecast maps which are now available at my website. As suggested by Kevin O'Neill, I thought I am posting the link here once again, as I have now added the anomaly maps for the Arctic (as requested from Neven ):

www.karstenhaustein.com/climate

As an example, the current 7 day average anomaly forecast looks like this:



Most of what's available on the web site should be self-explaining. It starts with the GFS forecast maps (Global, Europa, cont.US, Arctic, Antarctica), followed by a comparison of the previous model runs for the current time step. The anomalies are computed relative to the NCEP/CFSR 1981-2010 climatology. They are updated every 6 hours. I also provide the lower resolved NCEP reanalysis 1 hindcast. The archive of the monthly means (complemented with GISS data) can be found at the main page, while the archive of the last 4 GFS runs, together with the archived analysis time steps for GFS-CFSR and NCEP reanalysis for the last six months (see Archive).

Detailed information to the maps can be found in the FAQ section, which I encourage to read, particularly the bottlenecks involved. Don't take the maps as face value: FAQ

Hope it is a useful tool for some of you.

[Jim Hunt]

Another "video" built from the 3.125 km AMSR2 images provided by the University of Hamburg. This one reveals what's been happening on the Pacific side of the North Pole over the last week:



The skies over that part of the Arctic seem to be gradually clearing, and this morning Worldview provides some visual confirmation of the picture painted by the microwaves:

[Neven]

Nice work, Jim!

I have now added the anomaly maps for the Arctic (as requested from Neven ):

Awesome, Karsten! I will incorporate this in the ASIG at some point.

Sorry for not being able to cover this melting season more extensively here and on the ASIB. Our house is going to be built in two weeks, which means there's a lot to do.

[TerryM]

Where is the cyclone at the mouth of Bering Strait headed & what damage will it do?
Terry

[Jim Hunt]

Where is the cyclone at the mouth of Bering Strait headed & what damage will it do?

The models seem to all agree it will track along the north coast to the Amundsen Gulf until it merges with something more substantial over the CAA:

http://www.meteociel.fr/modeles/gfse_cartes.php?ech=6&code=code&mode=0&mode3h=0&runpara=0&carte=1

[pearscot]

Does the end of the melting season usually occur sometime mid to late September? If that's the case it looks like what we have is roughly what we will end up with. That said, do you guys think think there will be any interesting changes during that short time frame? I will be interested in seeing how the 'Panama' hole expands and if the ice pack witness more separation.

[ChrisReynolds]

Pearscot,

Typically you'd  expect minimum around mid September. I don't think there will be any surprises now, the loss rates are levelling as they do approaching the minimum.