Support the Arctic Sea Ice Forum and Blog

Author Topic: Bottom melt in Central Arctic ?  (Read 28862 times)

Rob Dekker

  • Nilas ice
  • Posts: 2386
    • View Profile
  • Liked: 120
  • Likes Given: 119
Bottom melt in Central Arctic ?
« on: June 25, 2013, 09:36:03 AM »
So far, measuring by SIE and SIA, it looks like 2013 is following a path of 'recovery' from the devastating 2012 melting season.

However, as several posters on Neven's main pages have pointed out, the ice in the Central Arctic is in dire straights, after continued pounding and dispersion due to persistent lows over the Central Basin.

It remains to be seen if the dispersion of ice and the many polynias currently present in the Central Basin will be persistent and will collect enough heat to basically "burn a hole in the North Pole" area this year, or if the ice there will compact, and still has enough resilience to withstand the remainder of the summer, to prevent an ice-free North Pole come September.

Interesting in the discussion about the resilience or vulnerability of the Arctic Basin is the observation of Ice Mass Balance buoy 2012J, which seems to show significant bottom melt over the past month :




This buoy is currently located at 87.75 N, 24.30 E, which puts it smack in the middle of the "damage zone" inflicted by the persistent lows which dominated late spring and early summer 2013, and that Neven rightfully names, the Persistent Arctic Cyclone (PAC) of 2013.

There has been some questions about the integrity of the data from this buoy (for starters, it's thermometer went haywire earlier this month, and it's bottom sounder is unreliable) but the data from the thermistor string insists significant bottom melt is occurring there, at least 30 cm over the past month, even though atmospheric temperatures are still close to salt-water freezing (-2 C).

Purpose of this post is to investigate if the bottom melt recorded by 2012J is real, and if so, what is causing it.

For that, I consulted the Ice Tethered Profiler buoys from the good people at Woods Hole Institute. These buoys record temperature and salinity below the ice, from the surface down to 700 meter below the ice.

There are three ITP buoys in the NP area (Nansen basin) which may shed light on 2012J's melting profile. ITP 57, 58 and 61.

ITP 57, at 86.9271° N, 88.2188° E, is closest to 2012J, and just like 2012J also smack in the middle of the slush zone that was affected by the PAC 2013.

The buoy status graphs shows that ITP 57 has been rocking-and-rolling over the past month, presumably as a result of PAC 2013 :



But here is the interesting part :
There area where 2012J, ITP 57, 58 and 61 hang out has a rather shallow halocline.
For example, the halocline under ITP57 is only 25 meters below the surface.
And now it seems that with the Ekman pumping and general turbulence created by the PAC 2013, the salinity right under the ice has increased quite dramatically. For ITP57, salinity increased from 31.5 psu a few months ago to close to 34 psu currently. ITP58 shows an increase from 31.5 to 34 and ITP61 shows increased salinity from 28.5 to 33 psu over just the past month.



Now of course, with increased salinity under the ice, we can expect water temperature below the ice to drop, and the heat to go to bottom melt, since the melting point of ice will be lower. But how much bottom melt does this increased salinity cause ? That depends on how fast and how deep the cold fresh melting water will get mixed with the lower salty waters.

In the case of all three ITPs (57,58 and 61) it seems that the upper 100 meter layer of water has cooled something like 0.2 C over the past month. That suggests that some 82 MJ of energy was transferred from the saltier water to the ice, which would represent something like 25cm bottom melt.

Incidentally that is pretty close to what we see has happened with 2012J.

Also, some posters suggested that maybe 2012J is close to a polynia, and thus may absorb solar energy that causes bottom melt as well. However, if solar energy caused the bottom-melt, then we would expect the water to "freshen" right below the ice. Instead, ITP 57, 58 and 61 all show increased salinity at the surface, which suggests the bottom melt is dominated by salinity increase, not insolation into nearby polynia.

So the hypothesis I propose is that currently bottom melt is occurring in the Central Basin, due to salinity increase below the ice, caused by increased turbulence by the PAC 2013, which stirs up the shallow halocline in the region.

In favor of this hypothesis is data from buoys in the area, notably 2012J, ITP 57, 58 and 61.
Notably also : Against this hypothesis is data from IMB buoy 2012E and 2012B in the area, which show not much bottom melt (yet).

I'm curious to read your thoughts.
« Last Edit: June 25, 2013, 11:15:50 PM by Neven »
This is our planet. This is our time.
Let's not waste either.

Rob Dekker

  • Nilas ice
  • Posts: 2386
    • View Profile
  • Liked: 120
  • Likes Given: 119
Re: Bottom melt in Central Arctic ?
« Reply #1 on: June 25, 2013, 09:41:51 AM »
Rats. These HTML tags did not work.
This is our planet. This is our time.
Let's not waste either.

Neven

  • Administrator
  • First-year ice
  • Posts: 9471
    • View Profile
    • Arctic Sea Ice Blog
  • Liked: 1333
  • Likes Given: 617
Re: Bottom melt in Central Arctic ?
« Reply #2 on: June 25, 2013, 09:43:22 AM »
Rob, I'll fix them tonight when I get back. You need BBCode instead of HTML tags (use the buttons in the editor). Welcome to the ASIF.

Excellent thread BTW.  :)
The enemy is within
Don't confuse me with him

E. Smith

Rob Dekker

  • Nilas ice
  • Posts: 2386
    • View Profile
  • Liked: 120
  • Likes Given: 119
Re: Bottom melt in Central Arctic ?
« Reply #3 on: June 25, 2013, 09:52:57 AM »
Rob, I'll fix them tonight when I get back. You need BBCode instead of HTML tags (use the buttons in the editor). Welcome to the ASIF.

Excellent thread BTW.  :)

I'm sorry Neven. And thank you for offering to correct HTML to BBCode.
I'll do better in the next post.

Still hope for comments on my hypothesis and analysis from the great people here on the forum.
This is our planet. This is our time.
Let's not waste either.

werther

  • Grease ice
  • Posts: 747
    • View Profile
  • Liked: 31
  • Likes Given: 0
Re: Bottom melt in Central Arctic ?
« Reply #4 on: June 25, 2013, 10:16:39 AM »
Thanks for analysing those buoys’ data, Rob!

I’ve been cadding on the MODIS r04c04 tile again, last week-end (FI producing the small NPole sector map on the 'ice-free'-thread), and, through the fog, saw no indication for concentration on the rise. On the contrary, measuring several floes and polynia’s, I got the impression that the intricate pattern of open water is, slowly, accumulating.
Even without PAC2013.
A large part of the CAB is under dense fog/low clouds now. That might be a reason why UniBremen has AMSR2 filling up pink/red again. CT area remains a basis for trend-comparison with the earlier years. But it is better not to read too much in the daily variance, for it has the same limits as UB.
Further, it is obvious that the ice in r04c04 has never been this dispersed this early in the season. Even last year, the ice over there got into June as part of  the well-known ‘mesh-structured’ pack.
I remember r05c03/r05c04 (theCAB Chukchi sector) showing the same properties July last year. As I recollect it, none of the usual indicators showed anomalously low SIE/SIA until GAC2012 came in and churned the whole region into open ocean.
I had a look on ASCAT today for another input. Hard to interprete, but an enormous 'Laptev-bite'
weakness seems to stand out...

anonymous

  • Guest
Re: Bottom melt in Central Arctic ?
« Reply #5 on: June 25, 2013, 10:55:55 AM »
Great work, Rob. I'd like to point to this report looking under every stone, trying to get hints, starting:

Observations of the ocean, atmosphere, and ice made by Ice-Ocean Environmental Buoys indicate that mixing events reaching the depth of the halochne have occurred in various regions in the Arctic Ocean. Our analysis suggests that these mixing events were mechanically forced by intense storms moving across the buoy sites.

There are a few more reports and papers documenting regional bottom melting, mixing events and eddies disturbing the halocline. It is a pity that only a handful of buoys cover an area of millions of square miles. I wonder how we might get more evidence except waiting for 25cm make a difference to the satellites.

The Navy model also suggest massive bottom melting caused by PAC in a region without buoys to confirm. The PIOMAS people have published a modelling paper (last year?) describing how the storm in the late season forced melting from below. Perhaps next gridded PIOMAS update will tell more. 

Richard Rathbone

  • Nilas ice
  • Posts: 1730
    • View Profile
  • Liked: 387
  • Likes Given: 24
Re: Bottom melt in Central Arctic ?
« Reply #6 on: June 25, 2013, 11:39:36 AM »
The argument hangs together well.

The mass and heat balance numbers fit.

Can you put numbers on the turbulence increase expected and show its large enough to cause the observations you ascribe to it?

I think you have a sound argument for a local effect causing a local melt, but you need to put some numbers on the amount of mixing that PAC 2013 would generate to turn that into a sound argument for 2012J being representative and 2012E and 2012B being outliers rather than vice versa.

Pmt111500

  • Guest
Re: Bottom melt in Central Arctic ?
« Reply #7 on: June 25, 2013, 12:45:12 PM »
"Against this hypothesis is data from IMB buoy 2012E and 2012B in the area, which show not much bottom melt (yet)."

At least the images (despite the radial blurring and blueing or what it is towards the edges of the images, one can see some variation in the floe' surface) do not show meltponding in here which in my opinion tells the upper layer of the ice is still somewhat under melting point. the snow layer doesn't look very thick, but is hard packed (looks like there's been no rain, hard packed snow insulates but it'll go fast), has all loss of snow (if there was much) been through sublimation? Once the floe hits an area of rain melting the insulating layer of (10-15cm?) snow, expecting the Perovich image to happen image url: http://forum.arctic-sea-ice.net/index.php?action=dlattach;topic=327.0;attach=2185;image.
Of course the floe may be out in the Fram by then.

Tried to get rid of the radial blurring manually, this is clearly a job for a programmer. attached NP2 cam photo (forgot the details already) from recent images folder. Original Image credit: NPEO (North Pole Environmental Observatory)
« Last Edit: June 25, 2013, 12:51:12 PM by Pmt111500 »

Jim Hunt

  • First-year ice
  • Posts: 6268
  • Don't Vote NatC or PopCon, Save Lives!
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 893
  • Likes Given: 87
Re: Bottom melt in Central Arctic ?
« Reply #8 on: June 25, 2013, 04:20:19 PM »
This parallels a discussion we're having over at "What the Buoys are telling us".

Here's a fairly crude chart of the readings from 2012J's thermistor chain. One interpretation is certainly that "significant bottom melt is occurring there, at least 30 cm over the past month". Is that the only interpretation however?
"The most revolutionary thing one can do always is to proclaim loudly what is happening" - Rosa Luxemburg

ggelsrinc

  • Frazil ice
  • Posts: 437
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Bottom melt in Central Arctic ?
« Reply #9 on: June 25, 2013, 05:30:36 PM »
So far, measuring by SIE and SIA, it looks like 2013 is following a path of 'recovery' from the devastating 2012 melting season.

However, as several posters on Neven's main pages have pointed out, the ice in the Central Arctic is in dire straights, after continued pounding and dispersion due to persistent lows over the Central Basin.

It remains to be seen if the dispersion of ice and the many polynias currently present in the Central Basin will be persistent and will collect enough heat to basically "burn a hole in the North Pole" area this year, or if the ice there will compact, and still has enough resilience to withstand the remainder of the summer, to prevent an ice-free North Pole come September.

Interesting in the discussion about the resilience or vulnerability of the Arctic Basin is the observation of Ice Mass Balance buoy 2012J, which seems to show significant bottom melt over the past month :

<p><a href="http://imb.crrel.usace.army.mil/irid_data/2012J_thick.png" target="_blank"><img src="http://imb.crrel.usace.army.mil/irid_data/2012J_thick.thumb.png" alt="" width="350" height="200" border="2"></a></p>

This buoy is currently located at 87.75 N, 24.30 E, which puts it smack in the middle of the "damage zone" inflicted by the persistent lows which dominated late spring and early summer 2013, and that Neven rightfully names, the Persistent Arctic Cyclone (PAC) of 2013.

There has been some questions about the integrity of the data from this buoy (for starters, it's thermometer went haywire earlier this month, and it's bottom sounder is unreliable) but the data from the thermistor string insists significant bottom melt is occurring there, at least 30 cm over the past month, even though atmospheric temperatures are still close to salt-water freezing (-2 C).

Purpose of this post is to investigate if the bottom melt recorded by 2012J is real, and if so, what is causing it.

For that, I consulted the Ice Tethered Profiler buoys from the good people at Woods Hole Institute. These buoys record temperature and salinity below the ice, from the surface down to 700 meter below the ice.

There are three ITP buoys in the NP area (Nansen basin) which may shed light on 2012J's melting profile. ITP 57, 58 and 61.

ITP 57, at 86.9271° N, 88.2188° E, is closest to 2012J, and just like 2012J also smack in the middle of the slush zone that was affected by the PAC 2013.

The buoy status graphs shows that ITP 57 has been rocking-and-rolling over the past month, presumably as a result of PAC 2013 :

<p><a href="http://www.whoi.edu/itp/images/itp57loc2.jpg" target="_blank"><img src="http://www.whoi.edu/itp/images/itp57loc2.jpg" alt="" width="350" height="200" border="2"></a></p>

But here is the interesting part :
There area where 2012J, ITP 57, 58 and 61 hang out has a rather shallow halocline.
For example, the halocline under ITP57 is only 25 meters below the surface.
And now it seems that with the Ekman pumping and general turbulence created by the PAC 2013, the salinity right under the ice has increased quite dramatically. For ITP57, salinity increased from 31.5 psu a few months ago to close to 34 psu currently. ITP58 shows an increase from 31.5 to 34 and ITP61 shows increased salinity from 28.5 to 33 psu over just the past month.

<p><a href="http://www.whoi.edu/itp/images/itp61dat3.jpg" target="_blank"><img src="http://www.whoi.edu/itp/images/itp61dat3.jpg" alt="" width="350" height="200" border="2"></a></p>

Now of course, with increased salinity under the ice, we can expect water temperature below the ice to drop, and the heat to go to bottom melt, since the melting point of ice will be lower. But how much bottom melt does this increased salinity cause ? That depends on how fast and how deep the cold fresh melting water will get mixed with the lower salty waters.

In the case of all three ITPs (57,58 and 61) it seems that the upper 100 meter layer of water has cooled something like 0.2 C over the past month. That suggests that some 82 MJ of energy was transferred from the saltier water to the ice, which would represent something like 25cm bottom melt.

Incidentally that is pretty close to what we see has happened with 2012J.

Also, some posters suggested that maybe 2012J is close to a polynia, and thus may absorb solar energy that causes bottom melt as well. However, if solar energy caused the bottom-melt, then we would expect the water to "freshen" right below the ice. Instead, ITP 57, 58 and 61 all show increased salinity at the surface, which suggests the bottom melt is dominated by salinity increase, not insolation into nearby polynia.

So the hypothesis I propose is that currently bottom melt is occurring in the Central Basin, due to salinity increase below the ice, caused by increased turbulence by the PAC 2013, which stirs up the shallow halocline in the region.

In favor of this hypothesis is data from buoys in the area, notably 2012J, ITP 57, 58 and 61.
Notably also : Against this hypothesis is data from IMB buoy 2012E and 2012B in the area, which show not much bottom melt (yet).

I'm curious to read your thoughts.

Good work.

This is just the first thought that popped into my head, but what if all those buoys are right? What if the conditions for large bottom melt only is in areas where a buoy was close enough to open water to detect it? Let me illustrate!

Let's say for the sake of discussion that the conditions for the 2012J bottom melt exist in half that area and the conditions for 2012E and 2012B bottom melt exist in the other half of that area. There are only three buoys, so one in eight times all buoys would be located in the large melt area and one in eight times no buoy would find large melt. Three in eight times one buoy would find the melt and three in eight times two buoys would find the melt. My point is the conditions for large bottom don't have to equally exist throughout the whole area for it to be significant. With such a small sample of data, the bottom melting could happen within a large portion of that area and produce similar results of only one buoy seeing it.

Is there good enough data to examine and compare the sea ice conditions around these buoys?

TerryM

  • First-year ice
  • Posts: 6002
    • View Profile
  • Liked: 893
  • Likes Given: 5
Re: Bottom melt in Central Arctic ?
« Reply #10 on: June 25, 2013, 08:23:49 PM »
Rob


Very nicely argued!
If I understand correctly Ekman pumping draws AW up into contact with the ice, then with melt water cooling & mixing down to 100 M we have a convection current constantly gnawing away at the bottom ice forcing more brine against the ice & continuing the rapid bottom melt.


This could help explain the amount of damage done last year by the GAC2012.


It's a situation similar to the way a polynia perpetuates itself but without the heat loss to the atmosphere. If this is happening how does the halocline eventually restore itself?


Terry




Jim Hunt

  • First-year ice
  • Posts: 6268
  • Don't Vote NatC or PopCon, Save Lives!
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 893
  • Likes Given: 87
Re: Bottom melt in Central Arctic ?
« Reply #11 on: June 25, 2013, 09:58:58 PM »
Rob - Having checked a few more CSV files it looks like ITP 61 is actually colocated with 2013B, and a variety of other kit also. Here's another thermistor summary which seems to confirm a reduction in water temperature just under the ice, but no bottom melt to speak of (yet).
"The most revolutionary thing one can do always is to proclaim loudly what is happening" - Rosa Luxemburg

jdallen

  • Young ice
  • Posts: 3410
    • View Profile
  • Liked: 650
  • Likes Given: 244
Re: Bottom melt in Central Arctic ?
« Reply #12 on: June 25, 2013, 10:38:01 PM »
Melt rates are pretty low at that temp - looks like -1.5C or so, which is about 3mm/ day or so.

Up the temp half a degree, and its about 13mm/day.  Up another half a degree and it will be about 25mm/day.  Get the water temp to zero and it will be"Katie bar the door!" Time, with between 45-50mm/day potential melt.

Now, a crucial oversight in our thinking comes to me... Wouldn't the buoys be placed by preference on MYI?!  If so, that could explain the lack of bottom melt we see them registering, as that wouldn't START to melt in earnest until seawater got at least a degree higher.  Until then, and until the internal temp of the ice stabilized, the melt we see would be more the result of chemistry.

Now with FYI, that would be a much different story... These temps, pumping and increased salinity would certainly be having a much sharper impact.  BUT, I suspect most of the instruments are not looking at FYI...!
This space for Rent.

Rob Dekker

  • Nilas ice
  • Posts: 2386
    • View Profile
  • Liked: 120
  • Likes Given: 119
Re: Bottom melt in Central Arctic ?
« Reply #13 on: June 26, 2013, 09:53:03 AM »
Guys, thank you so much for some great feedback.

Couple of responses, in no particular order :

TerryM "If I understand correctly Ekman pumping draws AW up into contact with the ice, then with melt water cooling & mixing down to 100 M we have a convection current constantly gnawing away at the bottom ice forcing more brine against the ice & continuing the rapid bottom melt."

Correct, as long as turbulence persists, Ekman pumping will bring warm/salty water to the surface, causing bottom melt, and the fresh cold water will be driven back down.
But since salty water is still heavier than fresh water, mixing (and thus bottom melt) stops almost right away when the turbulence ends (when the storm overhead passes). After that, it can take a week of two, but since salty water is heavier than fresh water, the salt water will sink and fresh water will rise, and stratification restores.

In the case of GAC 2012, we do not have enough buoy data to assess the true amount of bottom melt that is storm caused, although I have gone on record using ITP data that suggests that not just the upper, but even the lower halocline was disrupted during that storm down to some 500 meter,
http://neven1.typepad.com/blog/2012/08/arctic-summer-storm-open-thread.html?cid=6a0133f03a1e37970b0177440b63e5970d
and that there should have been at least some 100 W/m^2 bottom melt going on due to Ekman pumping during the storm. That may sound like a lot, but it lasted only for a week, and only for the 7-8 day duration of this storm, so it should have caused 20 cm bottom melt or so. That's pretty good for the first week of August, similar to clear skies over fragmented ice, but not overly spectacular.

Also, the PIOMAS team did a model run to see what the direct (bottom melt) effect of the GAC 2012 must have been, and ended up with "only" 150k km^2 lost due to bottom melt due to Ekman pumping, salt and heat distribution under the ice :
http://onlinelibrary.wiley.com/doi/10.1002/grl.50190/abstract
which seems a bit on the low end to me, but is not inconsistent with the ITP data and rough calculations.

Either way, it seems that a storm needs to be very large and very strong and long-lasting to cause enough Ekman pumping to rival heat absorption of clear-sky summer insolation over fragmented ice in August.
For June, there is no argument : clouds dominate energy uptake, and even a persistent cyclone like PAC 2013 does not even theoretically approach bottom melt due to direct sunshine into polynias and top-melt due to sunshine on melted snow and bare ice.

So I take my own argument with a grain of salt.
This is our planet. This is our time.
Let's not waste either.

Rob Dekker

  • Nilas ice
  • Posts: 2386
    • View Profile
  • Liked: 120
  • Likes Given: 119
Re: Bottom melt in Central Arctic ?
« Reply #14 on: June 26, 2013, 10:30:36 AM »
jdallen "Melt rates are pretty low at that temp - looks like -1.5C or so, which is about 3mm/ day or so."

I don't think that you can determine bottom melt <i>rate</i> by looking at the temperature of the water right under the ice alone.

In general, the temperature of water right under the ice will always in balance with the melting point of ice at the salinity level of the water. If it's higher, melt will occur, if it's lower, freeze will occur, but all that only in a thin layer of water right under the ice.

If you want an instant snapshot of the melt rate, then you would need to do a micro-temperature and velocity analysis of the water moving right under the ice. Exactly what Flux Buoys are doing :
http://www.oc.nps.edu/~stanton/fluxbuoy/

Unfortunately, the only Flux buoy deployed in the NP area is AOFB 28 (part of the NPEO 2013 instrument set) but this buoy apparently gave up already on June 4, and no bottom-melt flux calculations have been made with the couple of weeks of data this buoy collected.
Incidentally, we DO see an increase in salinity in the surface (right under the ice) over these few weeks, even with this buoy.

In absence of Flux buoy data, to determine the amount of melting the PAC 2013 caused, I looked at how much 'cooling' occurred in the upper 100 meters after the salinity increased in that layer. My reasoning is that the heat that used to be there must have gone to bottom melt, and ITP data suggest this can easily be 30 cm of bottom melt.

Long story short : You cannot determine bottom melt by looking at water temps under the ice, since they will always be at the temp you expect given the salinity of the water. Similar to how you cannot determine top melting by looking at DMI atmospheric temperatures, which will always be close to 0 C, melting temp of fresh-water snow.
This is our planet. This is our time.
Let's not waste either.

Rob Dekker

  • Nilas ice
  • Posts: 2386
    • View Profile
  • Liked: 120
  • Likes Given: 119
Re: Bottom melt in Central Arctic ?
« Reply #15 on: June 26, 2013, 11:07:42 AM »
Richard Rathbone "Can you put numbers on the turbulence increase expected and show its large enough to cause the observations you ascribe to it?"

and

"I think you have a sound argument for a local effect causing a local melt, but you need to put some numbers on the amount of mixing that PAC 2013 would generate to turn that into a sound argument for 2012J being representative and 2012E and 2012B being outliers rather than vice versa."

You are right Richard.
Unfortunately, it is very hard to calculate the "amount of mixing" caused by any storm, and especially one that is moving around and changing in intensity as much as the PAC 2013 did.
Not to mention the influence of other storms around.

So I can't give you a good answer, but I do have some humbling observations from at least one of the buoys I reported :

Jim Hunt : "Rob - Having checked a few more CSV files it looks like ITP 61 is actually colocated with 2013B, "

I realized that too, Jim. Both ITP 61 and IMB 2013B share the instrument set of NPEO 2013, which also features NP webcam 1 :
http://psc.apl.washington.edu/northpole/NPEO2013/webcam1.html

Since IMB 2013B does not show any bottom-melt, while co-located ITP 61 shows significant cooling of the upper 100 m ocean, I was scratching my head at this apparent contradition.

Until I analyzed the path that this instrument pack has taken.
ITP 61 shows increased salinity (and cooler temps) in the upper 100 m since about year day 145. Now it turns out (from ITP 61 track record) that these NPEO instruments just about moved from the West side, over the Lomonosov ridge into the deep of Nansen basin.

Now, obviously, Nansen basin supports a different ocean current than the Makarov basin, and thus 2013B's lack of bottom melt is easily explained : the 100 m water under the ice did not cool, instead the buoy moved over a different area, with different salinity.

This does not mean that 2012J's bottom melt is not caused by salinity changes due to the PAC 2013 as I suggest, but it does reduced the evidence for my whole argument to 2 buoys only. So let me analyze the track record of ITP 57 and 58, and see if their salinity changes may also be caused by them moving into Nansen Basin. If so, then I have no more evidence for my hypothesis, and will thus drop it.

Interesting still is that it seems that Nansen Basin contains more salt at the surface than usual this year. I've seen ITP data from Nansen Basin recording 32 and 33 psu, but never the 34+ psu that we see from e.g. ITP 58 :
http://www.whoi.edu/itp/images/itp58dat3.jpg
This is our planet. This is our time.
Let's not waste either.

TerryM

  • First-year ice
  • Posts: 6002
    • View Profile
  • Liked: 893
  • Likes Given: 5
Re: Bottom melt in Central Arctic ?
« Reply #16 on: June 26, 2013, 05:18:59 PM »
Rob
I'm thinking this through with FYI or MYI being melted and coming up with two very different scenarios.



1) Ekman pump draws warm saline water to ice interface


2a) FYI melts & releases cold brine which sinks through warm saline water
3a) cycle repeats as warm saline water is forced upward by displacement


2b) MYI melts releasing cold fresh water which finds it's depth in the saline water
3b) cycle halts as thermohaline stratification is restored
[size=78%].[/size]
If FYI was involved and the (a) sequence occurred it would leave the saline profile we're seeing rather than showing a rapid return to stratification. It's dependent on the melt water being saltier than the saline water that is doing the melting.


I recall that in 2011 a British expedition found FYI melt to be mixing the top 200 M so I'd expect bottom melt from Ekman pumping to have a similar effect. [/size]If the melt is taking place at a lower temperature than otherwise it seems as though the melted water/brine would drop even deeper than the melt that the 2011 expedition measured.[size=78%]

[/size]Is it possible that a storm over FYI would have a different outcome than a storm over MYI?[size=78%]

[/size]Terry[size=78%]

jdallen

  • Young ice
  • Posts: 3410
    • View Profile
  • Liked: 650
  • Likes Given: 244
Re: Bottom melt in Central Arctic ?
« Reply #17 on: June 26, 2013, 07:02:48 PM »
I don't think that you can determine bottom melt <i>rate</i> by looking at the temperature of the water right under the ice alone.

Absolutely.  There's a lot going on here.  *However*, the numbers I was quoting go back to a lab study examining heat transfer to 'berg ice and corresponding melt rates at different water temperatures.

http://www.igsoc.org:8080/annals/1/igs_annals_vol01_year1980_pg119-122.pdf

Certainly, there are some different dynamics in play with thinner floe ice, but this seemed a useful empiric examination of the general mechanics.

In general, the temperature of water right under the ice will always in balance with the melting point of ice at the salinity level of the water. If it's higher, melt will occur, if it's lower, freeze will occur, but all that only in a thin layer of water right under the ice.

Mostly agree, but in presenting the numbers I did, I was not presuming "static" conditions.  The act of melt does by its nature alter the conditions under which it starts.  Temperature at the ice/water interface, as well as salinity will change, in ways governed by the starting conditions and volumes involved.


Long story short : You cannot determine bottom melt by looking at water temps under the ice, since they will always be at the temp you expect given the salinity of the water. Similar to how you cannot determine top melting by looking at DMI atmospheric temperatures, which will always be close to 0 C, melting temp of fresh-water snow.

Yes, the melt is driven by energy exchange, not necessarily observed temperatures, and the phase transition alters the observed temperatures, whether air or water.

That said, observed temperatures over the broader volume of water or ice, *do* reflect the net enthalpy available to drive melt, or that which needs to be reduced to permit freezing.  The differential in temperature is predictive of how fast energy can move.  "Microclimates" do affect this, creating boundary layers which slow the transfer; but those "Microclimates" are themselves vulnerable and change rapidly.

Once again, the type of ice is very key.  At the very narrow margins of seawater temperature > 0 to -2C > MYI and FYI react very differently.  This is part of the point I was alluding to by wondering what type of ice the buoys have been placed on.

As TerryM points out, the nature of the ice itself also changes the energy exchange dynamic; MYI melt has dynamics which slow the rate and protect the ice by strengthening thermo and haloclines.  FYI will explicitly disrupt them, even without mechanisms like Ekman pumping.

Final note - I *do* think that the buoys are on MYI, and that is part of why we see them reporting low melt rates.  I further think that while they are giving us a read of local conditions, they aren't telling us what the FYI may be doing in the same area.  While MYI may not melt at -1.5C at all, over time, FYI can and will melt extensively.  The threshold for melt rate is very sensitive.  Half a degree can mean the difference between virtually no melt, and 20MM of melt a day.  As such, a relatively brief (24 hours or even less) fluctuation of sea water temperature can be highly localized, and have a very pronounced effect on the ice.
This space for Rent.

Juan C. García

  • Young ice
  • Posts: 3359
    • View Profile
  • Liked: 1279
  • Likes Given: 1127
Re: Bottom melt in Central Arctic ?
« Reply #18 on: June 26, 2013, 07:03:45 PM »
Maybe your 2013 "North Hole" post will mean something totally different...
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticict_nowcast_anim30d.gif
Hi Rob:
The link that wanderer put made me see the possibility of a 2013 bottom melt on Central Arctic, so I recommend seeing this link. 
Which is the best answer to Sep-2012 ASI lost (compared to 1979-2000)?
50% [NSIDC Extent] or
73% [PIOMAS Volume]

Volume is harder to measure than extent, but 3-dimensional space is real, 2D's hide ~50% thickness gone.
-> IPCC/NSIDC trends [based on extent] underestimate the real speed of ASI lost.

Rob Dekker

  • Nilas ice
  • Posts: 2386
    • View Profile
  • Liked: 120
  • Likes Given: 119
Re: Bottom melt in Central Arctic ?
« Reply #19 on: June 27, 2013, 08:55:27 AM »
jdallen said
Quote
Final note - I *do* think that the buoys are on MYI, and that is part of why we see them reporting low melt rates.  I further think that while they are giving us a read of local conditions, they aren't telling us what the FYI may be doing in the same area.  While MYI may not melt at -1.5C at all, over time, FYI can and will melt extensively.  The threshold for melt rate is very sensitive.  Half a degree can mean the difference between virtually no melt, and 20MM of melt a day.  As such, a relatively brief (24 hours or even less) fluctuation of sea water temperature can be highly localized, and have a very pronounced effect on the ice.

Thank you. Yes, I do think that these buoys have been placed on MYI, and yes, if salty, warm water bubbles up to the surface, it should affect (saltier) FYI more than (fresher) MYI.

So you are right that even if salinity increases under the ice surface that we may not see that reflected in melting on site, if the buoy is an IMB, simply because the energy goes to melting FYI nearby.

That is a good argument, and I wonder if we can find some evidence for that effect in the buoy data.
This is our planet. This is our time.
Let's not waste either.

Rob Dekker

  • Nilas ice
  • Posts: 2386
    • View Profile
  • Liked: 120
  • Likes Given: 119
Re: Bottom melt in Central Arctic ?
« Reply #20 on: June 27, 2013, 09:37:14 AM »
Juan C. García said
Quote
Quote
Maybe your 2013 "North Hole" post will mean something totally different...
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticict_nowcast_anim30d.gif
The link that wanderer put made me see the possibility of a 2013 bottom melt on Central Arctic, so I recommend seeing this link

Thanks Juan.
The purpose of HyCom is not to determine ice thickness accurately, but instead provide a reasonable estimate of ice encounters for US Naval vessels, so please take this model with a grain of salt.

For example, HyCom mostly overestimates ice thickness (for examle, in 2011 HyCom suggested 3 meter ice in the Beaufort, while the USS Healy was cruising along at 20 kts in that same area through 1 meter ice). Also, HyCom seems to have inherited it's predecessor (PIPS) attitude of asserting a "hole" in the NP area :
http://neven1.typepad.com/blog/2011/07/pips-is-back-more-info.html

It'd be more interested in the PIOMAS assessment of ice thickness in the Central Arctic.
This is our planet. This is our time.
Let's not waste either.

Peter Ellis

  • Grease ice
  • Posts: 619
    • View Profile
  • Liked: 33
  • Likes Given: 14
Re: Bottom melt in Central Arctic ?
« Reply #21 on: June 27, 2013, 10:53:24 AM »
Thank you. Yes, I do think that these buoys have been placed on MYI, and yes, if salty, warm water bubbles up to the surface, it should affect (saltier) FYI more than (fresher) MYI.

Some are on MYI, some are on FYI - if you look at the individual page for each buoy it will tell you. Note that the buoy page indicates whether it was MYI or FYI at the time of installation.  Every one of the 2012 buoys is on MYI by definition, since they were placed before this year's FYI started forming!

Peter Ellis

  • Grease ice
  • Posts: 619
    • View Profile
  • Liked: 33
  • Likes Given: 14
Re: Bottom melt in Central Arctic ?
« Reply #22 on: June 27, 2013, 11:01:39 AM »
Full run-down:

2012G - first-year ice when placed, so it's now second-year (perennial) ice
2012H - first-year ice when placed, so it's now second-year (perennial) ice
2012J - multi-year ice when placed
2012L - multi-year ice when placed
2012M - multi-year ice when placed
2013A - first-year fast ice (not sea ice)
2013B - unstated, but very likely multi-year given the starting thickness and location
2013C - multi-year ice when placed
2013E - unstated, but very likely first-year given the starting thickness and location

The installation notes for 2013B and 2013E haven't been put in yet.  However, both were installed at almost the same time in almost the same location, staging out of the Barneo ice camp this spring.  2013B was placed on ice over 2m thick, while 2013E was on ice around 1.4m thick.  If the latter is representative of the FYI (seems unlikely to be MYI), then the former is presumably MYI.

Jim Hunt

  • First-year ice
  • Posts: 6268
  • Don't Vote NatC or PopCon, Save Lives!
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 893
  • Likes Given: 87
Re: Bottom melt in Central Arctic ?
« Reply #23 on: June 27, 2013, 11:04:44 AM »
It's not quite that simple unfortunately Peter. 2013E says it is still "awaiting notes from field". In the absence of a functional bottom sounder it's guessimated to have been 140 cm thick in April, and it's still much the same today.

Is it FYI or MYI?

[Edit - Messages crossed!]
« Last Edit: June 27, 2013, 11:12:40 AM by Jim Hunt »
"The most revolutionary thing one can do always is to proclaim loudly what is happening" - Rosa Luxemburg

Steven

  • Grease ice
  • Posts: 957
    • View Profile
    • Arctic sea ice data and graphs
  • Liked: 481
  • Likes Given: 19
Re: Bottom melt in Central Arctic ?
« Reply #24 on: June 27, 2013, 02:55:53 PM »
Yes, I do think that these buoys have been placed on MYI, and yes, if salty, warm water bubbles up to the surface, it should affect (saltier) FYI more than (fresher) MYI.

Just a naive question.  The ice growth during the winter is mainly from the bottom.  Does that mean that at the start of the melting season, the bottom of the FYI and of the MYI have a similar salinity?  How does that influence the rate of bottom melt of FYI versus MYI at this time of the year?

In other words, has the salt migrated through the ice quickly enough, to make the bottom of the MYI fresher than the one of the FYI?
« Last Edit: June 27, 2013, 10:38:10 PM by Steven »

Jim Hunt

  • First-year ice
  • Posts: 6268
  • Don't Vote NatC or PopCon, Save Lives!
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 893
  • Likes Given: 87
Re: Bottom melt in Central Arctic ?
« Reply #25 on: June 28, 2013, 10:26:11 PM »
For comparison purposes, here's thermistor data from IMB 2006C which was in the Beaufort Sea through the summer of 2007. Both surface and bottom melt began on June 25th. 
"The most revolutionary thing one can do always is to proclaim loudly what is happening" - Rosa Luxemburg

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Bottom melt in Central Arctic ?
« Reply #26 on: June 29, 2013, 09:21:46 AM »
Jim,

In your plots, what is the x-axis scale?

Jim Hunt

  • First-year ice
  • Posts: 6268
  • Don't Vote NatC or PopCon, Save Lives!
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 893
  • Likes Given: 87
Re: Bottom melt in Central Arctic ?
« Reply #27 on: June 29, 2013, 11:43:36 AM »
In your plots, what is the x-axis scale?

Thermistor number.  Number 1 is in the air.  The last one is in the water. In the absence of further information the ones in between are open to interpretation, but they are nominally 10 cm apart.

I picked 2006C (albeit from the Beaufort) because it's famous as the longest lasting IMB buoy of all time, it looks at first glance like the buoy referred to in Perovich's papers, and it does come with some additional information. Thermistor 7 was on the ice surface at installation, so number 1 was 60 cm up, and number 45 was 380 cm down.

I've never followed the melt closely before, and I can't claim to have been through the archives with the proverbial fine tooth comb. However despite all the talk of a slow start this time around it does look as though 2012J, L and H have actually raced off the starting blocks quite quickly this year, to clearly visible effect in the latter case! 
« Last Edit: June 29, 2013, 12:05:52 PM by Jim Hunt »
"The most revolutionary thing one can do always is to proclaim loudly what is happening" - Rosa Luxemburg

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Bottom melt in Central Arctic ?
« Reply #28 on: June 29, 2013, 12:22:43 PM »
Thanks Jim,

It seems from what you say that we can only use this data in a qualitative sense.

TerryM

  • First-year ice
  • Posts: 6002
    • View Profile
  • Liked: 893
  • Likes Given: 5
Re: Bottom melt in Central Arctic ?
« Reply #29 on: June 29, 2013, 02:01:39 PM »
Chris


Thanks so much. I assumed everyone else understood the charts and that I had somehow missed the explanation. In retrospect my unwillingness to ask the question myself seems very foolish.


Terry

Jim Hunt

  • First-year ice
  • Posts: 6268
  • Don't Vote NatC or PopCon, Save Lives!
    • View Profile
    • The Arctic sea ice Great White Con
  • Liked: 893
  • Likes Given: 87
Re: Bottom melt in Central Arctic ?
« Reply #30 on: June 29, 2013, 02:33:53 PM »
Sorry Terry,

I was deliberately being slightly "economical with the truth".  It seems to me to be a not entirely trivial task to work out exactly where the top and bottom of the ice are from the raw thermistor readings alone. From your comment I guess you'd agree?

Nonetheless a lot of ice does now appear to be approaching its "melting point".


"The most revolutionary thing one can do always is to proclaim loudly what is happening" - Rosa Luxemburg

ChrisReynolds

  • Nilas ice
  • Posts: 1764
    • View Profile
    • Dosbat
  • Liked: 20
  • Likes Given: 9
Re: Bottom melt in Central Arctic ?
« Reply #31 on: June 29, 2013, 08:08:15 PM »
Thanks so much. I assumed everyone else understood the charts and that I had somehow missed the explanation. In retrospect my unwillingness to ask the question myself seems very foolish.

I frequently find myself reading comments and not quite getting what's going on. I only ask when stuff really interests me - the thermistor data did.

Jim,

A qualitative picture is still useful. Somewhere somebody posted a timeseries showing ice mass balance bouy data on thinning, some expressed amazement. Such massive thinning in July/August is what is shown by PIOMAS. Also HYCOM is showing thinning now because the thickness is dropping below 1.5m and the colour scale shifts around this point.
http://www7320.nrlssc.navy.mil/hycomARC/navo/arcticictn_nowcast_anim30d.gif