The 2016 melting season

[werther]

PS Never mind IJIS. On MODIS this can be followed like it is. And as it is given forecasts, 9 Mkm2 will soon be a reality.
As soon as Hudson and Baffin Bay ice becomes irrelevant, it will be much easier to do a quite good CAD-aided assessment of extent. It's the CAB thats relevant, 4 Mkm2 of ice that has experienced lowest 'winter power' ever.

[oren]

So I'm looking at the NSIDC (first image), everything looks roughly as expected...
Then I'm looking at ASMR2(second image), and wow, Polynyas EVERYWHERE.

I believe the difference is NSIDC uses a 25 km resolution while AMSR2 uses 3.125 km resolution. Always prefer AMSR2 to NSIDC for this reason.

[Lord M Vader]

June 1 marks the first day in 2016 with below normal temps in the DMI graph A delayed melt onset might have a strong impact at the rest of the melt season!

[jdallen]

Hi together,
Short question:
Is it usual, that Hudson and Baffin Bay is melting from the north side? And if so, why?
Best Regards

Not unusual.

Ice compacted as the result of partial melting and compaction driven by winds coming out of the NW.

That got accelerated by decreased albedo in the open water capturing increasing sunlight.

[bbr2314]

June 1 marks the first day in 2016 with below normal temps in the DMI graph A delayed melt onset might have a strong impact at the rest of the melt season!

Have we not had melting this year yet? Are you on crack?

[seaicesailor]

June 1 marks the first day in 2016 with below normal temps in the DMI graph A delayed melt onset might have a strong impact at the rest of the melt season!

Well... I'd say to have had continuous above normal temps for almost half a year should have a strong impact on the rest of the melting season.
But it's true too surface melting delay now can really impact the minimum extent : ')
In any case DMI 80N anomaly is a tricky metric in both directions

[seaicesailor]

Looking at the HYCOM drift maps, extent loss may slow down or stay at average rates for the next days. However, given the heat blows, area losses may keep on faster pace.

Another interesting thing from the maps is the effect of the weak low pressure system over Beaufort floes. Last year it was very bad for the broken ice to be dispersed toward the coast by seemingly protecting cold lows.

[Nick_Naylor]

So I'm looking at the NSIDC (first image), everything looks roughly as expected...
Then I'm looking at ASMR2(second image), and wow, Polynyas EVERYWHERE.

I believe the difference is NSIDC uses a 25 km resolution while AMSR2 uses 3.125 km resolution. Always prefer AMSR2 to NSIDC for this reason.

With 25km grids, there's a lot of room for the ice to melt/shatter before grids fall below 15%. 3.125 km grids are way better.

This image gives you an idea of what 25 km  grids look like in the Beaufort.

[JimboOmega]

So I'm looking at the NSIDC (first image), everything looks roughly as expected...
Then I'm looking at ASMR2(second image), and wow, Polynyas EVERYWHERE.

I believe the difference is NSIDC uses a 25 km resolution while AMSR2 uses 3.125 km resolution. Always prefer AMSR2 to NSIDC for this reason.

With 25km grids, there's a lot of room for the ice to melt/shatter before grids fall below 15%. 3.125 km grids are way better.

This image gives you an idea of what 25 km  grids look like in the Beaufort.

Not sure what I'm looking at.  I see an image that's mostly ice, with some blue squares on top of it? Are those blue squares NSIDC open water? Or something?

[JayW]

MODIS today shows 1.1 Mkm2 sea ice in the Chukchi, East Sib and Laptev Seas in blue hues. Extensive melt ponding. In some places, FI the Buor Khaya Gulf, this happened 'in a flash'.

Temps remain extremely high on the whole stretch from Bering Strait to Taymir all through the next eight days.
I can't remember this region undergoing such massive warming this time of the year.

The large American-Arctic coastal ice-free zone-to-be: there's a dispersed zone of ice left East of Cape Lisburne (only about 30 kim wide) and the attachment NW of Barrow (now a mere 40 km wide and almost passable along the fast ice boundary).

In the Labrador Sea, Hudson Strait is open. In Baffin Bay, connection of the Northern Waters polynia and the open water on Greenlands west coast is a matter of days.

Given the ECMWF forecast (dipole forming now and holding for about 5 days), I think '16 losses will closely follow those seen in '12.

I agree.  The 500 mb anomalous ridging over the arctic this month has been impressive.  Combined with the lack of "winter power" as you say , I'm wondering if 2012 can keep up with 2016.   

Attached are the H5 anomalies for May 1-30, all the scales are the same, for the years 2016, 2015, and 2012 respectively.

[ktonine]

June 1 marks the first day in 2016 with below normal temps in the DMI graph A delayed melt onset might have a strong impact at the rest of the melt season!

DMI N80 temps carry significant information at least until day 100.  After that they become problematic.  By day 150 they are almost useless and will remain so until sometime after day 235.  At this time of the year it's really just a curiosity.  Little correlation can be found in historical temperatures on DMI N80 for temperatures between days 120 to 240 with either SIE or SIA.

[Nick_Naylor]

Not sure what I'm looking at.  I see an image that's mostly ice, with some blue squares on top of it? Are those blue squares NSIDC open water? Or something?

I filled in the squares that looked like they were under 15% ice in blue. The rest is "extent". Note that this is a rough approximation. Move the squares a bit & you can get a noticeably different measure.

[6roucho]

With 25km grids, there's a lot of room for the ice to melt/shatter before grids fall below 15%. 3.125 km grids are way better.

This is analogous to the fractal dimension of coastlines.

[philiponfire]

June 1 marks the first day in 2016 with below normal temps in the DMI graph A delayed melt onset might have a strong impact at the rest of the melt season!

What are you trying to prove? this is the second time in the last month you have come out with a totally ridiculous post suggesting that the ice isn't melting. It is, it has been, and it will continue to melt. Adding to the over 900,000 sq km of missing ice  too date when compared to 2012 ( using MASIE).

[abbottisgone]

With 25km grids, there's a lot of room for the ice to melt/shatter before grids fall below 15%. 3.125 km grids are way better.

This is analogous to the fractal dimension of coastlines.

...the exact same thing, basically!

[6roucho]

With 25km grids, there's a lot of room for the ice to melt/shatter before grids fall below 15%. 3.125 km grids are way better.

This is analogous to the fractal dimension of coastlines.

...the exact same thing, basically!

The difference being that coastlines tend towards infinity.

[Michael Hauber]

June 1 marks the first day in 2016 with below normal temps in the DMI graph A delayed melt onset might have a strong impact at the rest of the melt season!

Obviously anyone on this forum knows that some melting has been happening somewhere for months, and any comment about delayed melt onset is talking about for a specific area.  In the current context I'd suggest is surface melt for the majority of the central Arctic.  It is not just the DMI graph for north of 80N showing below normal temps, but Andrew Slaters graph for the wider Arctic - although only for one or two days, before a rebound to slightly above normal.

Its a little while yet before we can declare a delayed onset of widespread surface melt (late June or later IMO), but it looks like we won't be as early as 2012, which was one of the earliest years with surface melt commencing for much of the central Arctic Basin in the first week of June.  I judge this on 850hp temps, although surface temps I've seen posted seem to suggest more widespread melt.  I've always found 850hp temps more informative, but mostly because surface temps won't go above 0 until all the ice is melted and 850 hp temps can show where significant above 0 heat exists.  So I wait and see in this case what happens.

[abbottisgone]

With 25km grids, there's a lot of room for the ice to melt/shatter before grids fall below 15%. 3.125 km grids are way better.

This is analogous to the fractal dimension of coastlines.

...the exact same thing, basically!

The difference being that coastlines tend towards infinity.

Ha, nice: ya got me thinking for a few days on that I suspect!

[Rob Dekker]

Slater's projection for July is slowing down.

He is now projecting 7.13 M km^2 for July 20, which is reducing the gap with 2012 to only 309 k km^2 at that date.

http://cires1.colorado.edu/~aslater/SEAICE/

I did not expect that slowdown.
Maybe something to do with the sharp increase in ice concentration over the past few days ?
If so, we can expect a reversal of that trend over the next week, with a torch-event predicted that affects the Chukchi, ESS, Laptev.

[abbottisgone]

Slater's projection for July is slowing down.

He is now projecting 7.13 M km^2 for July 20, which is reducing the gap with 2012 to only 309 k km^2 at that date.

http://cires1.colorado.edu/~aslater/SEAICE/

I did not expect that slowdown.
Maybe something to do with the sharp increase in ice concentration over the past few days ?
If so, we can expect a reversal of that trend over the next week, with a torch-event predicted that affects the Chukchi, ESS, Laptev.

Perfect timing: what be this so called 'torch-event' ?

[Rob Dekker]

Perfect timing: what be this so called 'torch-event' ?

Mmm. This one by climatereanalyser :

which is starting tomorrow (today if you are living in Europe).
But I'm sure there are even better temp predictor sites around.

[Watching_from_Canberra]

JAXA site seems to be back online and up-to-date:
https://ads.nipr.ac.jp/vishop/vishop-extent.html?N

Reporting 10,405,086 km2, June 1, 2016.  Still 440,000 km2 below any other year.

[abbottisgone]

edit Neven:

snip

I'm fed up with the sarcastic/cryptic/idiosyncratic one-liners containing zero information, asking for others to do the work, and then ask for more.

[EthanOConnor]

With 25km grids, there's a lot of room for the ice to melt/shatter before grids fall below 15%. 3.125 km grids are way better.

This is analogous to the fractal dimension of coastlines.

Unless my intuition fails me, there is a signal to be extracted from the extent difference as calculated on different grids. Has this been discussed on ASIF?

Generalized, we can consider measured extent as a function of grid size, and deltas or higher order measures on the shape of that curve for a given day.

These measures would probe the (putative?) power law underlying the size distribution of areas of open water. Mapped visualizations of some coefficient pulled from this process could complement thickness and concentration maps, and areal averages could complement compactness measures.

Of course NSIDC and JAXA/ADS-NIPR differ by more than grid size. This technique should probably be applied to carefully resampled versions of the higher resolution dataset. Carefully because aliasing and related effects distort self-similarity through induced false or frequency-shifted periodicity.

This (and any use of high-resolution sea ice products) has to be tempered with the reality that the raw observables produced by the satellites don't have the resolution to support a 3.125km grid. For example, the AMSR-2 19GHz channel has 3dB footprint of 22x14km. 3dB means a 50% falloff, so high-contrast areas like ice/water boundaries will bleed into each other significantly at that pixel size. Guessing, I'd say stricter bounds like 6dB or 10dB as often used to define resolving power would result in a footprint larger than 30x20. The highest resolution channel is spec'd at 5x3km for 3dB.

All of that means that a 3.125km grid is sampling some output of an interpolation and deconvolution process which is non-unique; which is to say it is a synthetic process and resolution dependence measures could accidentally probe characteristics of the signal processing rather than the ice.

Alright, I figure that's my offtopic allowance for several days. Hopefully someone has looked at this already and has a graph ready for us to bring us back to 2016 melting season? 

[Jim Hunt]

A fairly clear view of the Lena Delta from Aqua yesterday:

http//GreatWhiteCon.info/resources/arctic-sea-ice-images/summer-2016-images/#Laptev

[magnamentis]

June 1 marks the first day in 2016 with below normal temps in the DMI graph A delayed melt onset might have a strong impact at the rest of the melt season!

true that but DMI once more does not correspond with Climate Reanalyzer that hasn't shown any below average but above which means that we have to look at the two sourced independently, never knowing which is true.

as a standalone DMI indeed was the first day below average and spoiled the picture (just kidding)

[Iakub]

A fairly clear view of the Lena Delta from Aqua yesterday:

http//GreathiteCon.info/resources/arctic-sea-ice-images/summer-2016-images/#Laptev

This is beautiful!

Three aspects are particularly striking:

- The long distance ( > 100km) scale of "rivulets" in the (essentially) amorphous ice. Apparently there is an effect that causes the front edge of the stream to melt more quickly. There is some interesting physics going on there, looking at the similarity to the fractal appearance of viscous finger intrusion flows (e.g. http://fluidsengineering.asmedigitalcollection.asme.org/data/Journals/JFEGA4/27184/025301j.4.jpeg).

- The persistence of a wedge of ice at the centre of the stem of the stream. I would have naïvely expected to see that this is melted first given the strongest flow of warm surface water.

- The dirty blotches at some of the lower-left tips. Is this similar to jdallen's  #1793  image? Ice, melting over warmer water? Why then at the end? Does this mean the waters haven't mixed over this distance?

So intriguing.

[Laurent]

Laptev bite is brewing ?

[Laurent]

East siberian sea (ESS) becoming darker ?
Nearly no snow on the sides !
http://go.nasa.gov/1Vz9oZx

[Laurent]

Something to compare, 2 weeks from 16 may to june 2 in the Chuchi sea. Here 2012, 2015 and 2016 next.

[Laurent]

2015

[Laurent]

2016, this year the ice is darker.

[seaicesailor]

2016, this year the ice is darker.

Spectacular animations Laurent!
Indeed is darker. If it goes as 2015, the open front should harm a lot. Even when the effective thickness is big. This is FYI too.

Didn't the shallowness of the Chukchi-ESS-Laptev seas have some implication on how fast surface water warms there due to absorbed heat?

[Tor Bejnar]

A fairly clear view of the Lena Delta from Aqua yesterday:

http//GreathiteCon.info/resources/arctic-sea-ice-images/summer-2016-images/#Laptev

This is beautiful!
...

- The dirty blotches at some of the lower-left tips. Is this similar to jdallen's  #1793  image? Ice, melting over warmer water? Why then at the end? Does this mean the waters haven't mixed over this distance?
...

Dirty blotches are dirty river water on top of ice.  The spring floods cannot all fit beneath the sea [edit: sea or river] ice and so some flows on top, melting the snow and leaving some mud.

I agree as to the beauty of the image!   

[Magma.]

Three aspects are particularly striking:

[three observations by Iakub]

So intriguing.

This is the delta of the Lena River, not sea ice. You are looking at snow-covered sediments crisscrossed by hundreds of wide braided streams. (In this image north is to the left, and the Siberian landmass to the right.) The Lena is one of the world's great rivers, with an annual flow similar to that of the Mississippi, but almost unknown due to its remote and inaccessible location in eastern Siberia.

[Iakub]

Three aspects are particularly striking:

[three observations by Iakub]

So intriguing.

This is the delta of the Lena River, not sea ice. You are looking at snow-covered sediments crisscrossed by hundreds of wide braided streams. (In this image north is to the left, and the Siberian landmass to the right.) The Lena is one of the world's great rivers, with an annual flow similar to that of the Mississippi, but almost unknown due to its remote and inaccessible location in eastern Siberia.

Ah. I see. Mystery solved. Thank you for the clarification.

(Ice free view attached.)

[Iakub]

If you want an idea of what the Arctic Ocean will be like under ice-free conditions in the fall, simply look at the Great Lakes for a halfway decent analog.
...

Would not Hudson Bay be a closer analogy than any of the large freshwater lakes?

[Andreas T]

There is a substatial research station in the Lena Delta:
http://www.awi.de/en/expedition/stations/island-samoylov.html
 Involvement of the AWI seems to have ended? latest mention of german research there is from 2014.
Do the russian speakers have access to information on research there?

[JimboOmega]

If you want an idea of what the Arctic Ocean will be like under ice-free conditions in the fall, simply look at the Great Lakes for a halfway decent analog.
...

Would not Hudson Bay be a closer analogy than any of the large freshwater lakes?

People always show those graphs of certain regions and how they compare to previous years, I can't find them.  NSIDC has http://nsidc.org/images/arcticseaicenews/20110202_Figure4.png this old graph, but... given the small number of years and out of date (most recent is 2011) it would seem almost cherry picked to support the idea.

But support it, it does.  If you look at where extent starts on the left, and when complete refreeze happens... they're perfectly matched - lower extent, later refreeze.  (Of course, the graph doesn't show how long it was totally open in the summer, which would be an even better proof)

[Csnavywx]

If you want an idea of what the Arctic Ocean will be like under ice-free conditions in the fall, simply look at the Great Lakes for a halfway decent analog.
...

Would not Hudson Bay be a closer analogy than any of the large freshwater lakes?

Yes, yes it would.

[Magma.]

Ah. I see. Mystery solved. Thank you for the clarification.

(Ice free view attached.)

You're welcome. It didn't occur to me to show imagery of the area in summer for comparison... an excellent idea.

[Andreas T]

...

People always show those graphs of certain regions and how they compare to previous years, I can't find them.  NSIDC has http://nsidc.org/images/arcticseaicenews/20110202_Figure4.png this old graph, but... given the small number of years and out of date (most recent is 2011) it would seem almost cherry picked to support the idea.

But support it, it does.  If you look at where extent starts on the left, and when complete refreeze happens... they're perfectly matched - lower extent, later refreeze.  (Of course, the graph doesn't show how long it was totally open in the summer, which would be an even better proof)

is this what you are looking for?
https://nsidc.org/data/masie/masie_plots.html
Can also be accessed via the ASI graphs page

[JimboOmega]

...

People always show those graphs of certain regions and how they compare to previous years, I can't find them.  NSIDC has http://nsidc.org/images/arcticseaicenews/20110202_Figure4.png this old graph, but... given the small number of years and out of date (most recent is 2011) it would seem almost cherry picked to support the idea.

But support it, it does.  If you look at where extent starts on the left, and when complete refreeze happens... they're perfectly matched - lower extent, later refreeze.  (Of course, the graph doesn't show how long it was totally open in the summer, which would be an even better proof)

is this what you are looking for?
https://nsidc.org/data/masie/masie_plots.html
Can also be accessed via the ASI graphs page

Nope - I mean, it could be, but I need to see it over a whole year. The only graphs I could find (I looked a little more, on the "regional graphs" page) - don't have resolution enough... or it doesn't  have years where they are open substantially early that I'd expect it to have a big enough effect.

[Neven]

Gentlemen, I have moved your comments/essays to the Year-round ice-free Arctic thread. This took me almost 10 minutes. Next time I may not have the time.

[Crocodile23]

Mmm. This one by climatereanalyser :

which is starting tomorrow (today if you are living in Europe).
But I'm sure there are even better temp predictor sites around.

I like the graphs of temperature anomalies of this site(that takes them from GFS model):
http://www.karstenhaustein.com/climate.php#forecast
http://www.karstenhaustein.com/reanalysis/gfs0p5/ANOM2m_arctic/ANOM2m_mean_arctic.html

GFS predicted average temperature anomaly(based on 1981-2010 climatology) in the surface(2 m) for the next 7 days:

[Neven]

Something to compare, 2 weeks from 16 may to june 2 in the Chuchi sea. Here 2012, 2015 and 2016 next.

Thanks for the comparison, Laurent. I don't think there won't be much snow left in Siberia next week (or maybe the week after that for the Yamal peninsula). I wish I had maps to compare to previous years, but this year seems to be 'average' when compared to the last decade, according to the snow cover graph for Eurasia:



Edit: I found maps on that NOAA/NESDIS site hosting the graph above. Below a comparison for day 153 (June 1st/2nd) 2010-2016. If I'm looking at the coasts only, I'd say that 2015 is second to 2012 (but has more open water, of course):

[Neven]

Below, day 153 comparisons, very close, difficult to interpret. Here's mine: 2016 may have some more snow in Eurasia, but most of it is well south of the Lena Delta, and 2010, 2012, and 2015 still have some snow along the American coast, whereas it's almost all gone now, in 2016.

The forecast for the coming week: all the heat and open skies are going to be over those snow-covered regions (see Crocodile23's comment above).

2010 vs 2016
2012 vs 2016
2015 vs 2016

[oren]

Reporting on regional breakdown of extent: Beaufort, Barents, Baffin and Greenland Sea all continue to track downwards at record levels. Other seas are tracking on the 2015 trajectory. The Beaufort seems to be following 2012 and hasn't stalled like 2015.

[Rob Dekker]

Thanks Neven, for these (daily) snow cover images from the first days of June.

I really wish that Rutger's Snow Lab would produce their May numbers.
Not just so that we have a better way than eyeballing pixels and graphs to determine land snow cover, but also because I really would like to issue my "albedo"-based SIPN ice extent projection (which relies on Rutger's data) for September 2016.

My goal on this topic is that over the next year, I'll find good sources of gridded daily land snow cover, as well as sea ice extent and area data, and run the statistics on the albedo effect of these sources, so that we can determine a daily profile on how the melting season will likely proceed and an estimate of what its final September extent likely will be, with uncertainty margin.

[frankendoodle]

So are we calling open water alone the North American coast on June 2nd? Because at this point a vessel could sail from Alaska to the Canadian Archipelago without an ice breaker. They SHOULD NOT but you know what I am saying.