The 2017 melting season

[magnamentis]

I suggest not to give this "second-lowest July volume loss" thing too much meaning. Years with low volume tend to lose less, as a lot of the easy ice is already gone. 2012 was 3rd lowest, with a relatively small difference. It was still enough for 2012 to take back the lead, but it's not a huge move. In addition, IJIS loss during July and especially the 2nd half was relatively slow, and temperatures relatively cool with lots of clouds, so why expect PIOMAS to report an unusual melt? If anything, I consider this year's volume loss to be surprisingly resilient (and disturbing) in the face of such a slow July.
I would expect the 1st half of August to show above-usual melt, due to the recent storm with its flash melt, but if a GAC fails to arrive don't expect volume numbers to get back into the lead over 2012.

+1 so grateful you wrote, i deleted mine for known reasons LOL

[Greenbelt]

As the season gets later and the sun gets lower, the storms seem to matter more and more, pushing ice around in the water. The main operational models seem to indicate more storminess.

[Ice Shieldz]

Here you go 48hrs of ECMWF surface winds forecast. The high winds associated with the low expanded somewhat and shifted further south toward the periphery of the pack. Can't see how any slush ice there will survive this. Yesterday's forecast at 48hrs out called for winds maxing at 65km. Now only 20hrs out those winds are forecasted at 70km.

[Shared Humanity]

Strongest winds hit right where the ice is weakest. One thing to be thankful for is that the low is moving quickly and the winds don't stay this strong for long.

[Jim Hunt]

Here you go 48hrs of ECMWF surface winds forecast

My own view of the forecast cyclone, in amongst lots of other stuff:

http://GreatWhiteCon.info/2017/08/facts-about-the-arctic-in-august-2017/

There’s also now a lot of open water on the Pacific side of the Arctic, and Sunday’s cyclone is forecast to create a large area of 2 meter plus waves heading in the direction of the ice edge.

I expect that to have a noticeable effect on the already fragile sea ice by early next week, assuming the storm arrives as forecast!

[Tigertown]

It looks like the area of high concentration off the CAA is getting smaller and smaller and no doubt thinner. I think the chances of  late season extreme export are going up.

[epiphyte]

[edit - I wanted to follow-up with a minor qualification... note that the Schröder graphs from June have an odd distortion (rotation/expansion/compression increasing toward the edges) which fooled me somewhat w.r.t parts of ESS... so I must amend that to almost everywhere.  ]

Hope it's not too rude to follow-up on my own post months after the fact but I wanted to highlight the lesson for future reference. From what I can see, the negative melt-pond anomaly regions from the start of June appear to be an uncanny predictor of the end-of July picture...

... everywhere there were no melt ponds, there is now no ice !

I don't think the melt pond info is necessarily prognostic, especially where there is new snow over new ice, as is the case over most of the area currently exhibiting anomalously low melt-pond fraction.

Earlier today I posted on one of the extent prediction threads about this year's ice in Laptev, ESS, and the CAB north of both. Much of it is started off as open water off the coast in March this year or later, and was snowed on when it was only a few CM thick. It has not been subjected to the normal ridge-promoting onshore-offshore oscillation, but rather has been marching north, under an insulating blanket of snow, in mostly barely-freezing weather, to fill the void left by the MYI exiting the Fram, for the past three months.

Today was literally the first day this year that any of this ice has seen above freezing temps, and it already appears to be wilting fast. Tomorrow and the next day, it's going to get rained upon. I honestly will be quite surprised if any of that anomalously-low melt pond ice is still there at the end of July.

Neven isn't it soon time for you to do an article on Melting Momentum? Just from looking at the MODIS images it seems that we really don't have that much this year.

I've been in contact with David Schröder and he has confirmed (or rather his model) that this year (again) there is lesser melt pond formation than in years with record low minimums. Below is a comparison of 2012, 2013, 2016 and this year:



As you can see, there are even less melt ponds than last year. In fact, it looks very similar to 2013. David Schröder wrote to me in an e-mail:

We predict the September Arctic sea ice extent 2017 to be slightly larger than last year: 5.0 +/- 0.5 mill. km2. In spite of the lowest Arctic ice volume in April, the May melt pond fraction in May is the second lowest of the last 10 years in our simulation with the sea ice model CICE. Anomaly figures with respect to 2006 to 2015 are shown for May 2017, 2016, 2012 (highest pond fraction) and 2013 (lowest pond fraction). Stronger melting and pond formation only occurred north of Canada and some parts of the Beaufort Sea. Otherwise the pond fraction is very similar to May 2013. Below average air temperatures (with respect to 2008 to 2015) and increased snow depth seem to have had a stronger impact in reducing pond formation than the thinner ice in increasing pond formation.

Persisting snow on the ice is the best explanation for the fact that we haven't seen much melt ponding (ice turning blue on satellite images), despite ideal weather conditions during the past two weeks. Of course, as we have seen last year, preconditioning isn't absolutely decisive when it comes to the final outcome of the melting season, but this data seems to put a damper on prospects of breaking the 2012 minimum records. For now.

Addendum: What's nice about this, is that it confirms our suspicions back in February when we discussed what could be the reason there was such a divergence between PIOMAS and CryoSat. The number one suspect was snow cover, due to the relentless series of storms coming in from the Atlantic.

As I wrote at the time:

This is pretty big, IMO, as it tells us something about snow depth on the sea ice which may have consequences for the state in which the ice pack enters the melting season.

Maybe we're getting the answer. The last time there was such a divergence between PIOMAS and CryoSat was during the 2012/2013 winter. Remember, 2013 was cold and cyclonic, causing an unexpected rebound after 2012's spectacle. I don't think this year will be the same, if only for the fact that volume is at a record low because of a record mild winter, but 2017 could be 2016's rebound.

I think I'll be looking a lot at SSTs in weeks to come. And the weather, of course.

[greatdying2]

Here is the 3 day LAMB map ending today (Aug. 4th) and an animation (click to start).

[greatdying2]

After a few days of cloud, a region of the Beaufort showed through yesterday. It does not look good. Here is July 29, July 31, and Aug 4th:

[JayW]

July 31-Aug 4, high speed loop to allow the eye to see through the clouds.

Wrangel island in lower left, Beaufort lower right.

http://feeder.gina.alaska.edu

[Pavel]

Cumulative albedo warming potential is 4th highest but the blue anomaly mostly outside the Inner Basin

[oren]

After a few days of cloud, a region of the Beaufort showed through yesterday. It does not look good. Here is July 29, July 31, and Aug 4th:

July 31-Aug 4, high speed loop to allow the eye to see through the clouds.

Wrangel island in lower left, Beaufort lower right.

Thanks for these images. Here is an animation of the UH maps for the region. It doesn't look good at all and could easily go cliffy again.

[Bill Fothergill]

As the "flat Earth" brigade still like to gibber about loss of sea ice in the Arctic being somehow balanced out by gains in the Antarctic ...

[subgeometer]

140K drop on NIPR today

 This is when 2012 went through the floor so there's a long of ground to catch up but there's a lot of extent that's obviously going soon on the Pacific/Siberian side, and a lot more that looks like this stuff right by the pole around 86-88N towards FJL. Who knows, but I think in the next month the FDD anomaly last winter will show its effect.

We may have dodged a cannonball, for now, but lets not say it's just a flesh wound yet

[Pavel]

Another centure break in NSIDC extent. Meanwhile it's only 140k difference with 2012 on day 216

2012,    07,  30,      6.400,       
2012,    07,  31,      6.368,     
2012,    08,  01,      6.255,     
2012,    08,  02,      6.152,       
2012,    08,  03,      5.977

2017,    07,  31,      6.660,     
2017,    08,  01,      6.460,     
2017,    08,  02,      6.358,     
2017,    08,  03,      6.260,     
2017,    08,  04,      6.117,

[Bill Fothergill]

Here's what the 4th August values look like on the NSIDC's 5-day Charctic.

2017 pretty much bisects 2012 & 2016, with 2007 (not shown) effectively identical to 2017 at today's date.

The other years shown are those in which the minimum has already been surpassed. (N.B. The minimums for the 1981-2010 median, and for both 1991 and 1998, are still fractionally below the 5-day value for August 4th 2017. However, as the daily values for Aug 3 and Aug 4 2017 are already well below these other entries, they will be "blown away" when tomorrow's numbers arrive.)

[Jim Pettit]

Another centure break in NSIDC extent. Meanwhile it's only 140k difference with 2012 on day 216

...And almost certainly the closest 2017 is likely to be to 2012 until perhaps the end of September. the 2012 GAC's impact on NSIDC SIE numbers starts tomorrow, resulting in a five-day decrease of over 900k (including two double century breaks). That's--obviously--going to widen the gap between first and all other places.

Speaking of NSIDC SIE: with five to six weeks of decrease still to come, 2017's minimum is already lower than that recorded in every year from 1979 to 1998 (save for 1990 and 1995) plus 2001. It's also lower than the 1980s and 1990s average minimums.

[oren]

NSIDC with its coarse resolution may tell a certain story, but at the AMSR2 3.125km resolution, area is now almost 600k above 2012 for the same date. Only a huge and persistent storm could now catch up with the 2012 record. A volume record is more plausible but also receding at this point. I am moving all my bets to a 2nd place finish.

[Bill Fothergill]

Speaking of NSIDC SIE: with five to six weeks of decrease still to come, 2017's minimum is already lower than that recorded in every year from 1979 to 1998 (save for 1990 and 1995) plus 2001. It's also lower than the 1980s and 1990s average minimums.

Jim, did you experience a brief frisson of déjà vu as you were typing that?  See #3565 

[F.Tnioli]

Here you go 48hrs of ECMWF surface winds forecast. The high winds associated with the low expanded somewhat and shifted further south toward the periphery of the pack. Can't see how any slush ice there will survive this. Yesterday's forecast at 48hrs out called for winds maxing at 65km. Now only 20hrs out those winds are forecasted at 70km.

Much appreciated. I had a feeling this is serious the moment i've seen the first one you made, and it is. If i remember it right, in the 1st one, forecasted top wind speed was 64 km/h, yes? Now it's 70 km/h for 20h forecast. Real thing could then hit with up to 75 km/h winds. That's "only" +17% windspeed increase, but according to this, it would produce 27% increase of wave height - to 4+ meter waves (i calculated for 12h duration; longer duration would give up to 35%).

[F.Tnioli]

You must have very highly calibrated eyeballs and the latest version cerebrum software uploads to process such detail and arrive at such a conclusion.

Sort of. Some bugs in that software, though. May be i was wrong, yes, i said so at once up there. And, we now have Oren, Pavel, you and others all briefly stating what they think about PIOMAS report. I say even if i've been wrong with suspecting PIOMAS underestimated the July loss by some 5...10%, - again, very possible, - i still served as a catalyst for a good brief discussion about latest PIOMAS report. And in any case, i am sorry for those "software bugs", if that's what.

[meddoc]

My Idea would be to only take into account 50% Ice Concentration as Extent.
And then compare 2012- 2017 the Years.

It's too hot right now, for me to do it; but if some of You Statistic Skilled has the time and lust...

[jai mitchell]

Tweet with picture from the north pole shows ice conditions, says that it was raining earlier.

https://twitter.com/moraymo/status/893805339950411778

[F.Tnioli]

Looks quite good! Thanks for sharing it, Jai. This is expected, for now. Pole itself is holding up much better than some areas just couple hundreds miles away from it, last i checked. If the pole would be some deformed but solid ice field, or slush, or even open water by the end of this melting season - is much a gamble, i think, much depending on drift, weather and whether some storm would pass through near-pole regions before the definite start of the freezing season. IMHO.

[Tfisher]

Tweet with picture from the north pole shows ice conditions, says that it was raining earlier.

https://twitter.com/moraymo/status/893805339950411778

Also : "Got there on 50 Years of Victory, russian nuclear powered icebreaker.  Cut thru ice like butter...all thin ice."

[greatdying2]

My Idea would be to only take into account 50% Ice Concentration as Extent.
And then compare 2012- 2017 the Years.

It's too hot right now, for me to do it; but if some of You Statistic Skilled has the time and lust...

Nice idea. Or even better than an arbitrary cutoff, graphs of extent vs. concentration. Go go go!

[bbr2314]

If TOPAZ is correct, we see a continuing cliff through the 13th, and with the amount of thin ice remaining, that seems likely to continue through the end of the month...

[Ice Shieldz]

Much appreciated. I had a feeling this is serious the moment i've seen the first one you made, and it is. If i remember it right, in the 1st one, forecasted top wind speed was 64 km/h, yes? Now it's 70 km/h for 20h forecast. Real thing could then hit with up to 75 km/h winds. That's "only" +17% windspeed increase, but according to this, it would produce 27% increase of wave height - to 4+ meter waves (i calculated for 12h duration; longer duration would give up to 35%).

Wow I had no idea how a linear windspeed increase creates a supralinear increase in wave height. Makes sense given the nonlinear relationship between the speed of travel and the amount of wind resistance.

It appears that our current winds, according to ECMWF, ended up closer to 65kms. After the winds pass we will end up with a disrupted halocline, but to what degree i don't know and any melting caused by this disruption will help offset the net effect? In four days forecasts call for a longer-lived and larger low to return to the same vicinity where this system passed. Of course the duration of the forecasted system will have a greater influence on wave height than our current short-lived system, also the size of the system would produce a large fetch - the other factor in determining wave height. Being four days out, it's a crap shot to pin point its location but if it happens in the same vicinity we already have weakened ice and a disrupted halocline both of which could lead to more significant drops in extent.

On a side note - Back in mid-June occurred the longest lived system of the melt season pounding the the region from the Laptev well into the CAB. The winds lasted long and were strong enough to generate over two meter waves on the CAB side of the Laptev.  That was pretty significant especially given the distance of fetch was not that great. At the time I was wondering how far those waves would extend into the pack and what effect it would have on the ice later in the melt season. Perhaps that big Laptev bite we are seeing now is to a large extent due to that system?

[Jim Hunt]

it would produce 27% increase of wave height - to 4+ meter waves (i calculated for 12h duration; longer duration would give up to 35%).

Here's the current WWIII forecast for this evening (UTC):

http://GreatWhiteCon.info/2017/08/facts-about-the-arctic-in-august-2017/#comment-222642

2.5 meters seems to be nearer the mark. Peak period 7-8 seconds.

[Jim Pettit]

Jim, did you experience a brief frisson of déjà vu as you were typing that?  See #3565 

Ah, the dangers of near-simultaneous posting. Well, great minds think alike, and all that...

[bbr2314]

it would produce 27% increase of wave height - to 4+ meter waves (i calculated for 12h duration; longer duration would give up to 35%).

Here's the current WWIII forecast for this evening (UTC):

http://GreatWhiteCon.info/2017/08/facts-about-the-arctic-in-august-2017/#comment-222642

2.5 meters seems to be nearer the mark. Peak period 7-8 seconds.

That is going to do a *horrific* number on the ice and explains the TOPAZ forecast. The Pacific edge of the pack is going to be retreating by tens of miles per day, and even the relatively thick ice is now vanishing (i.e. check Beaufort satellite imagery past few days).

The above will work in tandem with the already-decrepit first yr ice across the Russian/Atlantic periphery to result in continuing major losses. With the amount of heat on the Pacific side, the question is how far the front advances before refreeze... and with very little thick ice in the way of those waves, we may have to wait until September.

This means it may not take a GAC on-par with 2012's to achieve the same effect, simply because the remaining ice quality vs. 2012 is mostly substantially worse.

[epiphyte]

Starting to look quite chilly inside that low - perhaps some of the precip will fall as snow.

I was about to say that could be a lifeline if it happens - but OTOH I guess it's more about halocline mixing and bottom melt for the next few weeks anyway - It could be that a layer of snow would just delay the refreeze - who knows?

[FishOutofWater]

A large channel in the CAA garlic press north of the main channel just started moving. It isn't moving all the way to the Arctic ocean, but a large mass of ice just shattered and started moving south.

[Ice Shieldz]

Jim, I see now by looking at your wave height graphic and the general direction of previous surface winds that the cumulative winds have been blowing over a long fetch in the same general direction - another reason for the high wave heights. Wondering how the 7-8 second peak wave period factors in terms of impacts on ice?

There is a significant difference between the current windspeed calculated by ECMWF versus GFS.  ECMWF is 11kms higher. This is via WindyTV which uses GFS at 22km resolution versus ECMWF at 9km. Also I noticed in NullSchool (GFS) a similar difference in wind speeds, although a couple kilometers higher than WindyTV GFS.

[slow wing]

July 31-Aug 4, high speed loop to allow the eye to see through the clouds.

Wrangel island in lower left, Beaufort lower right.

http://feeder.gina.alaska.edu

Thanks Jay, this is very helpful to my understanding of what is going on under these storms.
Yes, the ice is moved around, but is there any melt or breaking up of the floes?
I just can't see any, even in the regions where the ice is already degraded.
This gif has caused me to think the storms are having less impact than I had previously thought.

[Sterks]

... everywhere there were no melt ponds, there is now no ice !

Had the melt pond maps shown positive anomalies in May, and should the surface melting over the Arctic not been delayed 15 days and even 1 month in many locations due to a cold Arctic we would be witnessing a really unprecedented situation. We are witnessing a bad, not unprecedented sitution. A lot of ice to disappear North of 80° in the Pacific side? Happened in recent years, 2012, 2015, 2016, and if nothing stops it, 2017.
As for the advance of the Pacific front, I agree, there is a limit, the new exposed open water is very cold, only a continuation of the windy weather in the right direction can imo facilitate all the mechanical action as well as atmospheric transfer of the heat available toward the Central Arctic.
A quiet phase or a reverse dipole, say bye to rapid melting.
Too speculative at this point.

[Adam Ash]

...
Thanks Jay, this is very helpful to my understanding of what is going on under these storms.
Yes, the ice is moved around, but is there any melt or breaking up of the floes?
I just can't see any, even in the regions where the ice is already degraded.
This gif has caused me to think the storms are having less impact than I had previously thought.

Remember that in those images the smallest floe you can see is probably several kilometres across.  The initial wave and wind damage will be in the metre-scale floes first (possibly partly due to overturning) with impacts on larger floes occurring due to wave action breakup, I should imagine.

[slow wing]

Adam, thanks for your comment.
That is true - the initial wave and wind damage will be largely on metre-scale floes - although it could also break up larger floes.
Even so, I still don't see any regions where more water is showing.
Q1. There will presumably be some regions composed of lots of metre-scale floes?
Q2. Wouldn't such regions darken in this gif if the small floes are decomposing?

[Jim Hunt]

Jim, I see now by looking at your wave height graphic and the general direction of previous surface winds that the cumulative winds have been blowing over a long fetch in the same general direction - another reason for the high wave heights. Wondering how the 7-8 second peak wave period factors in terms of impacts on ice?

In surfing terms the 8 second period still counts as mere "wind slop" rather than a full blown long distance swell. For much more detail on all that see:

http://GreatWhiteCon.info/2015/03/sea-ice-and-swells-in-the-beaufort-sea-in-the-summer-of-2014/

The September 2012 swell was indeed rather more substantial than those we have looked at in 2014. With a height in excess of 20 feet, the period was too short to qualify as a full blown swell and it was in fact directed away from the ice edge rather than towards it.

For more on what a long distance 13.5 second period swell in the Central Arctic can achieve see also:

http://GreatWhiteCon.info/2015/07/is-time-running-out-for-arctic-sea-ice/

The swell caused the vast MY ice floe nearest the Amundsen to ride up one side of the swell and fracture as it crested the wave peak, creating smaller ice floes of width approximately one half of the wavelength of the swell. In a matter of minutes from the initial onset of swell propagation, all large MY ice floes in the region were fractured in this manner, yielding a new distribution of smaller MY ice floes ranging from 100–150 m in diameter.

There is a significant difference between the current windspeed calculated by ECMWF versus GFS.  ECMWF is 11kms higher. This is via WindyTV which uses GFS at 22km resolution versus ECMWF at 9km. Also I noticed in NullSchool (GFS) a similar difference in wind speeds, although a couple kilometers higher than WindyTV GFS.

There's also a significant difference between forecasts a mere 6 hours apart. What should one believe is actually happening?!

[Jim Hunt]

Some interesting images in the wake of yesterday's waves:

http://GreatWhiteCon.info/2017/08/facts-about-the-arctic-in-august-2017/#Aug-6

[Sterks]

The pack can still break a-la 2016 without their GACs. The PAC storm re-intensifies in D3-D4-D5, and per its location it may push all that dispersed ice to their disappearance, and open up new polynya. Very strong pressure gradients with storm center at 975 hPa, that is why I say without their GACs.
The forecast shows persistence of this storm in the area for three more days, but that is very uncertain.
The gif will need a click (days 3 to 5 of ECMWF)

[Sterks]

Meanwhile, this is the Hycom drift envisioned for tonight.*
*Edit: Past night

[Jim Hunt]

The PAC storm re-intensifies in D3-D4-D5, and per its location it may push all that dispersed ice to their disappearance, and open up new polynya.

As you may have gathered from my comments above, I'm afraid these days I take forecasts that far out with a very large pinch of salt. With that proviso, here's the current forecast for the peak of the next pulse of swell. Note the change of scale, and the peak period is up to ~9 seconds:

[gerontocrat]

It's getting a bit chilly up there in the high Arctic?

[Sterks]

It's getting a bit chilly up there in the high Arctic?

That should not stop bottom and lateral melting so long as the ice is dispersed, and the ocean is subject to waves, upwelling, and as floe breaking starts to make lateral melting important. However, yes, a cold atmosphere will be absorbing heat and slowing down melting.

[Shared Humanity]

Unless my eyes are deceiving me, the 2m temps are slightly above freezing in a broad swath north of the CAA and slightly below freezing for the rest of the Arctic. This is normal during the melt season, isn't it?

[Sterks]

It's colder than July
But it doesn't hurt to remember that melting temperature for FYI ranges between -1.5°C and -1.8°C

[Shared Humanity]

It's colder than July

And September will be colder than August.

[Thomas Barlow]

I think this is might the dip I've been expecting, due to peripheral ice melting out, I mentioned a few times (there is possibly more vulnerable peripheral ice than 2012 for example). Could end up lower than 2012 in a week or so (and low volume in the Arctic Ocean).

[Shared Humanity]

But it doesn't hurt to remember that melting temperature for FYI ranges between -1.5°C and -1.8°C

Which would mean that all of the Arctic has been subjected to temps conducive to melt over the last 5 days.