Support the Arctic Sea Ice Forum and Blog

Author Topic: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland  (Read 1200110 times)

Wipneus

  • Citizen scientist
  • Young ice
  • Posts: 4220
    • View Profile
    • Arctische Pinguin
  • Liked: 1025
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #550 on: February 20, 2015, 05:51:11 PM »
Believe it or not! Massive calving seen at the southern branch of Jakobshavn Isbræ:

Does this Sentinel "PolarView" from the 15th help throw additional light on matters?

The second one, from the 14th, is even less "crystal clear"!

These are image made with the sensor in EW (extra wide) mode. Lower resolution and especially worse quality than the IW images that also exist for Jacobshavn. The glare is worsened by the poor image processing that is the fate of these images.

I tried to apply some extra doses of love and care to get some decent images form the raw data.
The result shows beyond doubt that the calving was on February 14/15.

« Last Edit: February 20, 2015, 06:34:18 PM by Wipneus »

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #551 on: February 20, 2015, 08:50:34 PM »
Thanks guys for the additional documentation!
From what I can see, Radarsat data, Zachariae is moving big too at the moment, yes this will be an interesting new season. ;)
Have a ice day!

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #552 on: February 20, 2015, 08:55:35 PM »
Wow! The massive crevasses were evident earlier but I was thinking they were in a state of suspended animation due to buttressing from frozen mélange in the fjord. I'm going to guess this is an effect of warmer water in Baffin Bay (and so Disko Bay and at the calving front: the mélange is just not strong enough this year to hold back calving.

This was a 2-3 x larger event in terms of cubic km than we were seeing this summer. What effect would consistently less back pressure have on annual discharge? Perhaps lessen the slow-down in winter, thus raising the average monthly discharge and so the annual -- it takes a lot of May and June to re-accelerate Jakobshavn.

Meanwhile, here are the 15 m Landsats.

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #553 on: February 20, 2015, 09:18:46 PM »
Just to keep you informed, this is the retreat situation since 1851, the years mentioned are the minimums (at least 2013 and 2014)!

Please click on image to enlarge!
« Last Edit: February 20, 2015, 09:29:11 PM by Espen »
Have a ice day!

werther

  • Grease ice
  • Posts: 747
    • View Profile
  • Liked: 31
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #554 on: February 20, 2015, 09:29:08 PM »
Well, what can one add… thought something was moving as I eyeballed Sermeq kujalleq three days ago and saw this MODIS pic, cleared after the sun returned out there last week:



Next day I saw Espen calling a large calving and he’s right. I haven’t checked the new calving size, but it could actually have been about 7 km2 ( thickness 500-600 m1 so about 3.5-4.2 km3) that has subsided. Still, as spring approaches, the front is about 900 m1 short of the max. June ’12 retreat.
Sermeq Kujalleq is bound in a dynamic struggle to stay in contact with the sill at the entrance of the deep trench. It is just a matter of time until this contact will be lost and we might witness a rapid retreat.
« Last Edit: February 20, 2015, 09:36:31 PM by werther »

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #555 on: February 20, 2015, 10:00:42 PM »
Wow! The massive crevasses were evident earlier but I was thinking they were in a state of suspended animation due to buttressing from frozen mélange in the fjord. I'm going to guess this is an effect of warmer water in Baffin Bay (and so Disko Bay and at the calving front: the mélange is just not strong enough this year to hold back calving.

A-Team I dont think the fjord and what is in it, have any effect on the forces coming from Jakobshavn, we are in relatively warm water, anyway?
Have a ice day!

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: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #556 on: February 20, 2015, 10:53:39 PM »
A-Team/Werther - Thanks for the additional info, now added to my original article.

I waited till I figured they'd be up and about in Boulder, then I phoned the NSIDC to see if they had any comment. The recorded message said they're all away on what sounds like a team building exercise of some sort. Next I tried DMI, but I guess they'd already all headed off home for the weekend by then :(

Last time I called DMI I discovered that they had a team out in the field in Greenland. I wonder where?
"The most revolutionary thing one can do always is to proclaim loudly what is happening" - Rosa Luxemburg

lanevn

  • New ice
  • Posts: 91
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #557 on: February 20, 2015, 11:06:12 PM »
Were that ice  "floating" or it added now to sea level rise?

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #558 on: February 20, 2015, 11:17:57 PM »
Were that ice  "floating" or it added now to sea level rise?

The ice that was above sea level, more or less, will be added to the current sea level eventually, when melted.
Have a ice day!

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #559 on: February 21, 2015, 12:38:46 AM »
Quote
I don't think the fjord and what is in it, have any effect on the forces coming from Jakobshavn

Maybe get permission to take away the flying buttresses from a gothic cathedral to see if it holds up? The speedup started 15 years ago or so when JI lost its floating ice shelf. Now JI is losing its solid mélange which is the temporary winter counterpart to an ice shelf. Many many publications on this for JI, about 1 in 10 we have reviewed here. The blog search tool for 'Baffin" will find the classic Nature paper ... firewalled but I posted the key figures.

Domed ice sheets like Greenland want to sag flat. Frozen mélange on marine terminating glaciers is only one of many sources of resistance (buttressing) to that. The water near the fjord surface is not warm, just ever so slightly above the freezing point. Because the calving wall goes so incredibly deep, when it falls off and turns over sideways, it actually pushes even strongly frozen mélange westward a km or more during an event. Consequent convective mixing at the calving front leading to yet more melt has also been the subject of numerous studies.

Here is a 2015 paper on the Jakobshavn mélange that will cover all the earlier work as well:

Seasonal and interannual variations in ice melange and its impact on terminus stability, Jakobshavn Isbræ, Greenland
R CASSOTTO, M FAHNESTOCK, JM AMUNDSON, M TRUFFER, I JOUGHIN
http://www.igsoc.org:8080/journal/61/225/j13j235.pdf

We used satellite-derived surface temperatures and time-lapse photography to infer temporal variations in the proglacial ice melange at Jakobshavn Isbræ, a large and rapidly retreating outlet glacier in Greenland.

Freezing of the melange-covered fjord surface during winter is indicated by a decrease in fjord surface temperatures and is associated with a decrease in ice melange mobility and a drastic reduction in iceberg production.

Vigorous calving resumes in spring, typically abruptly, following the steady up-fjord retreat of the sea-ice/ice-melange margin. An analysis of pixel displacement from time-lapse imagery demonstrates that melange motion increases prior to calving and subsequently decreases following several events.

We find that secular changes in ice melange extent, character, and persistence can influence iceberg calving and therefore glacier dynamics over daily-to-monthly timescales which if sustained will influence the mass balance of an ice sheet.
« Last Edit: February 22, 2015, 01:07:22 PM by A-Team »

AbruptSLR

  • Multi-year ice
  • Posts: 19703
    • View Profile
  • Liked: 2268
  • Likes Given: 286
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #560 on: February 21, 2015, 01:14:33 AM »
Were that ice  "floating" or it added now to sea level rise?

The ice that was above sea level, more or less, will be added to the current sea level eventually, when melted.

Not to be a nitpicker but as soon as the glacial ice Volume Above Floatation, VAF, joins the mélange it contributes immediately to sea level rise, whether it has melted, or not; because it immediately displaces ocean water.
“It is not the strongest or the most intelligent who will survive but those who can best manage change.”
― Leon C. Megginson

Pmt111500

  • Guest
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #561 on: February 21, 2015, 04:27:25 AM »
Just to keep you informed, this is the retreat situation since 1851, the years mentioned are the minimums (at least 2013 and 2014)!

Please click on image to enlarge!


tried to combine this image with the bottom topography one on the previous page and got an ugly image that would state the calving front is about 3 km (little under 2miles, if you insist on using obsolete measures) from the last ocean bottom hill before the channel deepens towards the main Ice Sheet. Is this about right? Thus the ocean water should start to get under the GIS in few years.  This would transform parts of it into a very old (130000-year) sea ice, so the deniers might get excited by the sudden increase of old ice. But as I said I'm not too sure the image overlay is in correct position, so not posting it.

This should pretty much ensure, that if there's no top melt on GIS there will be bottom melt, no? Thus GIS might join the WAIS parts in the irreversible melting mode in a few years. the bottom topography map though does not show if there are additional barricades of rock and ice, so not saying +15m ASL is inevitable in 300-1500 years.
« Last Edit: February 21, 2015, 05:55:11 AM by Pmt111500 »

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #562 on: February 21, 2015, 01:19:25 PM »
Right, it is not so easy to locate the latest high resolution grayscale DEM of the Jakobshavn sill area in the south branch, even though the garishly colored hill-shaded angled perspective broad area view (like the one that keeps getting posted here) is ultimately bump-mapped up from this flat grayscale.

The short answer is cut to M Morlighem's web page and follow the link to NSIDC data archive for the netCDF files and open with Panoply freeware or the ImageJ2 plugin. That group has refined the 6 main Greenland maps into a self-consistent co-registered GIS dataset using conservation of mass considerations. This will very likely improve on the above Bamber 2013 in these over-deepened marine terminating glaciers (yet how is this to be experimentally validated since radar isn't doing well on these?)

http://sites.uci.edu/morlighem/dataproducts/mass-conservation-dataset/

Otherwise, it is quite difficult to co-register an overhead satellite nadir view with a perspective view with whose angles and viewing location aren't provided when, as here, the elevations have been shaded according to shadows cast by that particular view. That shading was not done sensibly in a HSV (hue saturation value) color space channel.

Since the map-generating software, palette, and command line used are almost never mentioned in glaciology articles (which seldom have substantial 'methods' sections), it would be difficult to monte carlo your way up from a grayscale DEM up to the perspective view, that is, invert the perspective view to the flat gray DEM so you could do a proper overlay.

Nor is it easy to locate the most recent 1D depth profile map of the channel bottom and locate the current calving front relative to nearest sill. Here the experimental error is important as it will be quite high on the deep or steep-walled areas which are especially problematic for ice penetrating radar.

The short answer is cut to Fig.3 of Joughin 2014 (enhanced below). That shows the JI calving front climbing out of a very deep trough and just about at the top of the second highest sill, to be followed by a shallow trough, the main sill, and then a long slide. They ending up tossing all the post-2008 DEM products and all the subsequent grid flights: Free full at http://www.the-cryosphere.net/8/209/2014/tc-8-209-2014.pdf

Quote
Bed elevations were interpolated from a gridded map of radar depth soundings produced by the Center for Remote Sensing of Ice Sheets (CReSIS)(Li, 2009; Van der Veen et al., 2011). Multiple versions of the DEM exist, but based on comparison with other data sets our preferred version is the one located at ftp://data.cresis.ku.edu/data/grids/old_format/2008_Jakobshavn.zip. 
The radar data from middle years is carefully reviewed by Gogenini 2014, http://www.igsoc.org:8080/journal/60/223/j14j129.pdf, noting the problems arising from terminal roughness, side reflections, and material at the bottom.

Quote
The ice thickness is ~850m at the calving front [which converts calving area to calving volume], increases to~1500m, then decreases to ~1200m between 5 and 7km. Beyond this, the ice thickness varies between 1400 and 1200m. We also observe two distinct echoes separated by 100–200m in the along-flow line echogram caused either by side-wall echoes and bed echoes or multiple interfaces at the bed.
Oddly, this article does not include the great many subsequent grid overflights with what one supposes are improved radar designs nor provide a bottom line map of the depth profile. Ice thickness would have to be subtracted from surface elevation to get at trough and sill.

I've looked at all the later down-channel flights but it is just not possible to read off the trough-and-sill profile at the calving front because the radar is not reaching bedrock (per echo2 image annotations by Cresis themselves).

The data on JI -- even though it is the most intensively flown grid in either Greenland or Antarctica -- is still quite sparse: points at 50 m intervals don't fill lines and lines at 5 km spacing don't fill areas -- Kansas is not flying peano curves. JI is an extreme situation and ice-penetrating radar struggles with flares off the steep walls and noisy crevasses near the terminus.

So who among us has time to compare and contrast the latest trough-and-sill profile from Morlighem with Joughin 2014 and Gogenini 2014?
« Last Edit: February 21, 2015, 08:28:34 PM by A-Team »

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #563 on: February 21, 2015, 03:26:52 PM »
Quote
Sermeq Kujalleq is bound in a dynamic struggle to stay in contact with the sill at the entrance of the deep trench. It is just a matter of time until this contact will be lost and we might witness a rapid retreat.

To quote Joughin 2014, pg 212, fixes in brackets:

Quote
Further summer speedup of Jakobshavn subsequently declined in 2013 as the terminus retreated to higher ground. By contrast, the 2009 mean speed was faster than the 2011, despite a more advanced 2009 terminus position. This difference likely reflects the fact that at these time scales, the other processes and feedbacks have a substantial influence on flow.

If the terminus has reached the bottom of an overdeepened region, then the terminus may be able to find a position of transient stability on the high spot farther upstream [located at positions 12 to 17 km in the inset below] as retreat to shallower depths yields slower speeds (Joughin et al., 2012a).

The relatively high surface slope region upstream of the overdeepened basin [1.5º, image below] where the present heights are [80 to 167 m] metres above flotation, may further slow retreat, since it should take more time to thin to near flotation. By contrast, low surface slopes and heights near flotation likely facilitated the rapid retreat since 2009.

While the high spot above the basin may slow flow, the terminus would still be grounded on a bed at least 900 m below sea level, likely yielding speeds well above balance that would maintain strong though potentially diminished thinning. As a consequence, the terminus likely will continue to retreat, albeit perhaps more slowly in the near term, until it again reaches bed depths similar to summer 2012 (1300 m below sea level) at 15 km farther upstream.

At the current stage, further retreat may largely be driven by ice dynamics with only weak coupling to climate forcing. Thus, a re-advance rather than retreat likely would require a period of several years to decades of cooling. Once past the high spot, the trough extends roughly 50 km farther inland at depths below sea level of 1200 m and more before eventually reaching shallower depths.

Thus, although variation in trough width of 3.5 to 5.5 km could modulate the rate of flow, once into this deepest part of the trough, extreme velocities (12 km/yr) are likely to persist as the terminus rapidly retreats. Furthermore, without the [need] to seasonally advance up a relatively steep bed slope as in the past several winters, such high speeds may be sustained year round.

Because the deep trough of Jakobshavn is extremely difficult to measure with conventional techniques, we cannot rule out that some or all of the high spot might be a gridding artifact, in which case retreat may occur even more rapidly. Similarly, we cannot rule out that some bed highs may not have been resolved that could provide additional points of transient stability.

I re-did the flotation (thinning) graphic below to fix the scaling. The flotation curve looks like the mirror image of the bedrock curve because it is essentially that adjusted for current surface elevation (affine transformation).
« Last Edit: February 21, 2015, 08:39:03 PM by A-Team »

Pmt111500

  • Guest
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #564 on: February 21, 2015, 03:58:10 PM »
Thanks, A-Team, this makes it certain my image was falsely located... still not certain by how much, but at least 3 km off to the fjord side... Good to hear there is detailed investigation here.

werther

  • Grease ice
  • Posts: 747
    • View Profile
  • Liked: 31
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #565 on: February 21, 2015, 04:10:12 PM »
Thanks, A-Team,
That quote seems to neatly sum up my own thoughts. I've been writing down and summarizing my views on the process today and will spare everyone the detail as your quote is illustrative enough. Here's a layered insert of the bedrock map I found several years ago on CE-journal, coming from I think a presentation by Konrad Steffen:



The black lines on both sides of the gorge represent the sea level on the bedrock. The red line indicates the maximum 2012 retreat. Where yellow fades into green lies -500, and green into blue -1000m. I've been thinking about how 'Baffin Water' could eventually get under the glacier. There's no large intrusion yet. But some specifics in the present situation and the properties of the deep gorge make it likely that this intrusion might become a fact in the next few years.
In that context I just remind the research done by ETH during '03-'04, indicating a layer of free flowing water right under the glacier at the bottom of the deep gorge. The layer was about 120m deep at maximum and extended for 40 km all the way up to the sill.
There's also the present situation, where the dynamic equilibrium during the last few years has created a sort of mixer bassin, about 40 square km, on the Eastern sill-slope, reaching from 300 to 900 m below sea level.
The big question is, will further retreat expose the end of the 40 km stretch of the above mentioned free flowing water layer and its supporting melt water drainage system enough for warmer, saltier ocean water to enter against the pression of the deep-/melt water flow?

If that happens, dig this. It took a century to retreat 30 km up to the head of Ilulissat Fjord. It took only 10 years to retreat another 15 km to where the calving front is oscillating against the backside of the sill. I suppose there are more than enough factors contributing to an even faster retreat as soon as the sill is cleared at last. Within ten years the calving front might be almost halfway into the gorge, leaving a new, deep fjord for ocean water to find a playground.

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #566 on: February 21, 2015, 04:12:39 PM »
A-Team how much of that 628 meter was lost from 2013 to 2014?
Have a ice day!

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #567 on: February 21, 2015, 06:43:13 PM »
I have been digging into my Landsat Archives, and I cant find evidence of any real calvings this early in the season, is this a new future behavior?
Zachariae is already a year around calving engine. 
Have a ice day!

nukefix

  • Grease ice
  • Posts: 546
    • View Profile
  • Liked: 27
  • Likes Given: 9
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #568 on: February 21, 2015, 07:07:14 PM »
What is the thickness of the glacier at the pinning point  a few km upstream of the calving front? Is it even close to floating or would it need to thin a lot more, which would require higher flow-speeds?

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #569 on: February 21, 2015, 08:46:29 PM »
Quote
What is the thickness of the glacier at the pinning point [sill] a few km upstream of the calving front? Is it even close to floating?
I extracted the flotation graphic from Joughin 2014 and added to #565 above. It is about 150 m too thick, nowhere near floating if we assume the bedrock depth profile is correct. Note the authors themselves are questioning the existence and positions of the sills.

I believe Werther has tracked down the source of an immense amount of journal confusion: the older Steffen paper where the sill is assumed downstream of the current calving front. I have no idea what this could have been based on, very early radar maybe. Cresis did not reach bedrock until 2006-2008.

Thinning of lower JI was discussed in a 2001 Steffen paper:

Abdalati W and Steffen K (2001)
Greenland ice sheet melt extent:1979–1999.
J. Geophys. Res.,106 (D24), 33983–33988
« Last Edit: February 22, 2015, 01:54:42 PM by A-Team »

sidd

  • First-year ice
  • Posts: 6774
    • View Profile
  • Liked: 1047
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #570 on: February 21, 2015, 11:22:26 PM »
I extracted the flotation graphic from Joughin 2014 and added to 564 above. It is about 150 m too thick, nowhere near floating.


The scale on the left for the blue line seems to have been chopped off. Could you indicate the units ?

werther

  • Grease ice
  • Posts: 747
    • View Profile
  • Liked: 31
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #571 on: February 21, 2015, 11:53:51 PM »
Well, I have to admit I'm losing it on A-Team's bedrock length-profile. Maybe I'm relying too much on the clear bedrock map by Steffen I posted earlier. I just don't see any important sill within the next 40 km from the present position of the calving front.
Taking that as fact, there's no 'pinning point' at all a few km upstream. A-Team's profile though presents a buckle out there up from -1300 to about -800. Could be; after all there's no real good data available.
Anyway, I agree with A-Team that glacier thickness is too large for floating out there at this stage. No sill 2100 m, over a sill like in the profile 1400 m. Too much to float, the deep water layer on the bottom of the gorge is under enormous pressure. Nevertheless, dynamic layering, shear and changes in the behaviour/area of the ablation zone could very well change the properties.

sidd

  • First-year ice
  • Posts: 6774
    • View Profile
  • Liked: 1047
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #572 on: February 22, 2015, 05:26:49 AM »
contour map from morlinghem nc file bed layer
upside down, north is down, sorry, don't have time to fix
 

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #573 on: February 22, 2015, 10:31:44 AM »
Now the numbers are upside down ;):

Have a ice day!

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #574 on: February 22, 2015, 10:44:01 AM »
Including the present calving front unless I am completely wrong?

Have a ice day!

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #575 on: February 22, 2015, 11:01:02 AM »
Quote
How much of that 628 meter was lost from 2013 to 2014?
Not very much, but I wonder if this will be the year that the calving front retreats beyond the sill. Though how would we know if its location and even existence aren't certain? I suppose by a fit to the physics. Yet all is not well there because it seems not to be kicking out any publishable predictions for 2015-20.

You raise an interesting point in that this article pins calving front, surface velocity, and thinning data on a somewhat arbitrary channel centerline coordinates whereas the south channel has ~5-6 km of width. The 2013-14 change line shown above is observed kattywompus to the centerline.

Note the calving front itself is quite heterogeneous in terms of ice origin (the northern third is the ice stream proper, the southern part comes in from the side), so we have a situation where part of the calving front could have already climbed out of the overdeepening. Chasing down the topography of this sill -- rounded knob, asymmetric teardrop, or flat spillway -- with respect to the centerline is really problematic given the data situation.

Quote
The scale on the left for the blue line seems to have been chopped off. Could you indicate the units ?
The scale is still there, barely (the m below the 167), it is meters. I do condense the graphic legends to make more room for resolution per the 700 maximal pixel width here. This one had to be radically rescaled vertically in gimp to get its scale readable. The authors did not furnish the underlying numeric data points in journal supplemental (despite utilizing it) so I could not cleanly re-graph it.

The background grid is really needed to subdivide the units which came overly large 500 m chunks in the original, whereas the flotation varies on a scale of meters.

These grids are easy to make in Gimp if you can remember where they have the tool buried: Filters --> Render --> Pattern --> Grid. First count the number of pixels at full range of the scale and decide how that range is best subdivided, eg 165 pixels factors as 3*5*11 so those plus 15, 33, 55 would work for grid spacing pixels (or rescale the 165 to something more decimally like 200).

The trick is not to do the grid within the graph but rather as a separate layer that is otherwise empty (all alpha). That way the color and thickness of the grid can be changed later, but most importantly placed unobtrusively behind the graphic (if the white spaced there can be alpha-ed) and dimmed with the opacity controller.

Quote
relying too the clear bedrock map by Steffen
This older map is an important discovery because it explains a lot of confusion in the journals about where the sill is located, before or beyond the current calving front. Presumably later maps based the 2006-08 experimental radar data are better because those years were the first to reach bedrock.

Glaciers almost always have overdeepenings (greater bedrock gouging) and those imply an upstream sill (lesser gouging). I'm wondering about the status of side-looking sonar sounding in the first 100 km of fjord, whether the bathymetry has gotten at all close to the retreating glacier, or whether it could with an autonomous underwater drone. (Radar cannot penetrate sea water so cannot measure depths once the ice is gone.)

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #576 on: February 22, 2015, 12:15:56 PM »
Quote
Including the present calving front unless I am completely wrong? Now the numbers are upside down
Nice work, sidd and Espen! Not so easy to place the calving front but it looks about right relative to above sea level exposed rock.

I wonder if Morlighem was limited to 100 m resolution by computational time, model accuracy, or by considerations of input experimental error. In terms of sill peaks, 100 m leaves something to be desired. There seem to be a few bugs such as the -6600 m at the start of the fjord. This bug is a boundary effect that arose elsewhere and is now propagating all over.

This map does not support the classical notion of a shelf or teardrop-shaped knob acting as sill. Instead, it is showing a bottom slanted downward to the north for many tens of km upstream, with maximal depth being a series of -1200 to -1500 m holes adjacent to the northern wall.

In cross-channel profile, the ridge gives a gently slanted structure reaching minimal elevation along the north wall and tapering to shallow on the south. This seems inverted relative to glacial gouging at least in the straightaways. However the high velocity ice stream, as seen clearly in 15 m Landsat8 flowlines, follows very close to the north wall and so differential erosion can be anticipated there. It's not so clear why the tapering extends so evenly to the south wall however.

Note that the surface velocity maps of the lower ice stream that we looked at months ago in great detail have the fastest ice in the middle, tapering symmetrically to the sides. This does not fit the picture of the most glacial excavation taking place on the north side. Most of that action may have taken place earlier when this glacier truly dwarfed what we have today.

If this interjection of physics (requiring mass conservation) to fine-tune initial radar characterization better represents the real situation in the Jakobshavn channel, then it is indeed a breakthrough. It would certainly be cause for deleting many previous modelling papers on this glacier.

However we don't know that -- if the 'inverse' method of Morlighem could be experimentally validated (by improved radar), why would we even need Morlighem's method? (This asks which bottom profile is most consistent with surface observables and basic glaciology.) The marine outlet glaciers have intensive, not sparse, coverage. It's not clear that holding back a subset of flight segments really tests the method.

The asymmetry in the steepness of the channel walls is very striking, indeed puzzling. On river erosion, the inside bend has the steep walls and the outside bend the diffused gradient, as seen here as the glacier rounds the Big Bend and approaches the calving front. Farther upstream there are places where this relation is reversed.

I looked at removing the numbers altogether and coloring between the contours, just the ones below sea level. The trick for that in gimp is color-picking on the blackest black and then growing the selection boundary by a pixel in all directions. However that takes out some of the contour pixels, requiring too much repair work. So I first cut the width down to 700 pixels so the calving front area will display here at its full resolution.

I can improve on the quickie coloring below but just wanted to toss something out for now. The land above sea level is  better done as a grayscale so as not to distract from the channel. This would free up some of the color palette for the channel ... though seems like 15 colors is sufficient.

The palette is made very quickly in gimp choosing first and last colors, doing a gradient across a palette rectangle, softening with the hue slider, overlaying a grid separating colors, and posterizing to discrete colors. Then a central palette color is picked with the eyedropper and used to fill in a color-picked central contour-bounded white space and work out from there. It's hard to tell what contours go with what numbers go but by working out concentrically from the initial choice, the elevation will always jump to the next band available, though at some point that reverses. Takes longer to describe than to do!

If sidd can produce this contour map at slightly higher resolution, the tighter contours will be get better separated and be feasible to infill farther upstream. It's very important not to jpg colors because that eradicates later color-picking.

Notice how the map is cut off on the west side. It can't be extended unless really old radar is around from a time before all this retreat of grounded ice began. If our sill has evaporated, I'm wondering whether the overdeepening west of it is still on solid ground (experimentally speaking).
« Last Edit: February 22, 2015, 09:20:16 PM by A-Team »

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #577 on: February 22, 2015, 12:36:29 PM »
And here is one from September 28 2014:
Have a ice day!

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #578 on: February 22, 2015, 01:21:02 PM »
Let's watch that spelling on Mathieu Morlighem, associated with the Rignot group and now faculty at UC Irvine. The original papers are here:

Quote
Deeply incised submarine glacial valleys beneath the Greenland ice sheet
M Morlighem, E Rignot, J Mouginot, H Seroussi, E  Larour
http://www.nature.com/ngeo/journal/v7/n6/full/ngeo2167.html

The bed topography beneath the Greenland ice sheet controls the flow of ice and its discharge into the ocean. Outlet glaciers move through a set of narrow valleys whose detailed geometry is poorly known, especially along the southern coasts. As a result, the contribution of the Greenland ice sheet and its glaciers to sea-level change in the coming century is uncertain.

Here, we combine sparse ice-thickness data derived from airborne radar soundings with satellite-derived high-resolution ice motion data through a mass conservation optimization scheme*. We infer ice thickness and bed topography along the entire periphery of the Greenland ice sheet at an unprecedented level of spatial detail and precision. We detect widespread ice-covered valleys that extend significantly deeper below sea level and farther inland than previously thought.

Our findings imply that the outlet glaciers of Greenland, and the ice sheet as a whole, are probably more vulnerable to ocean thermal forcing and peripheral thinning than inferred previously from existing numerical ice-sheet models.
Quote

*M Morlighem
A mass conservation approach for mapping glacier ice thickness.
free full http://onlinelibrary.wiley.com/doi/10.1029/2011GL048659/pdf
Geophys. Res. Lett. 38, L19503 (2011).

The traditional method for interpolating ice thickness data from airborne radar sounding surveys onto regular grids is to employ geostatistical techniques such as kriging. While this approach provides continuous maps of ice thickness, it generates products that are not consistent with ice flow dynamics and are impractical for high resolution ice flow simulations.

Here we present a novel approach that combines sparse ice thickness data collected by airborne radar sounding profilers with high resolution swath mapping of ice velocity derived from satellite synthetic-aperture interferometry to obtain a high resolution map of ice thickness that conserves mass and minimizes the departure from observations. We apply this approach to the case of Nioghalvfjerdsfjorden (79North) Glacier.
Let's hope sidd has time to pull out a few more of these contour maps for our favorite glaciers!

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #579 on: February 22, 2015, 01:30:49 PM »
Let's watch that spelling on Mathieu Morlighem, associated with the Rignot group and now faculty at UC Irvine. The original papers are here:

Quote
Deeply incised submarine glacial valleys beneath the Greenland ice sheet
M Morlighem, E Rignot, J Mouginot, H Seroussi, E  Larour
http://www.nature.com/ngeo/journal/v7/n6/full/ngeo2167.html

The bed topography beneath the Greenland ice sheet controls the flow of ice and its discharge into the ocean. Outlet glaciers move through a set of narrow valleys whose detailed geometry is poorly known, especially along the southern coasts. As a result, the contribution of the Greenland ice sheet and its glaciers to sea-level change in the coming century is uncertain.

Here, we combine sparse ice-thickness data derived from airborne radar soundings with satellite-derived high-resolution ice motion data through a mass conservation optimization scheme*. We infer ice thickness and bed topography along the entire periphery of the Greenland ice sheet at an unprecedented level of spatial detail and precision. We detect widespread ice-covered valleys that extend significantly deeper below sea level and farther inland than previously thought.

Our findings imply that the outlet glaciers of Greenland, and the ice sheet as a whole, are probably more vulnerable to ocean thermal forcing and peripheral thinning than inferred previously from existing numerical ice-sheet models.
Quote

*M Morlighem
A mass conservation approach for mapping glacier ice thickness.
free full http://onlinelibrary.wiley.com/doi/10.1029/2011GL048659/pdf
Geophys. Res. Lett. 38, L19503 (2011).

The traditional method for interpolating ice thickness data from airborne radar sounding surveys onto regular grids is to employ geostatistical techniques such as kriging. While this approach provides continuous maps of ice thickness, it generates products that are not consistent with ice flow dynamics and are impractical for high resolution ice flow simulations.

Here we present a novel approach that combines sparse ice thickness data collected by airborne radar sounding profilers with high resolution swath mapping of ice velocity derived from satellite synthetic-aperture interferometry to obtain a high resolution map of ice thickness that conserves mass and minimizes the departure from observations. We apply this approach to the case of Nioghalvfjerdsfjorden (79North) Glacier.
Let's hope sidd has time to pull out a few more of these contour maps for our favorite glaciers!

I will add Zachariae to that list! ;)
Have a ice day!

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #580 on: February 22, 2015, 02:34:45 PM »
According to this 2015 paper, we could predict the next calving event just by watching for melange movement on Sentinel, Modis, and Landsat pairs. Counter-intuitively, melange movement precedes calving. Less glamorously, we could look backwards a few weeks and see if the current one could have been anticipated.

Seasonal and interannual variations in ice melange and its impact on terminus stability, Jakobshavn Isbræ, Greenland
R CASSOTTO, M FAHNESTOCK, JM AMUNDSON, M TRUFFER, I JOUGHIN
http://www.igsoc.org:8080/journal/61/225/j13j235.pdf

We used satellite-derived surface temperatures and time-lapse photography to infer temporal variations in the proglacial ice melange at Jakobshavn Isbræ, a large and rapidly retreating outlet glacier in Greenland.

Freezing of the melange-covered fjord surface during winter is indicated by a decrease in fjord surface temperatures and is associated with a decrease in ice melange mobility and a drastic reduction in iceberg production.

Vigorous calving resumes abruptly in spring following the steady up-fjord retreat of the sea-ice/ice-melange margin. An analysis of pixel displacement from time-lapse imagery demonstrates that melange motion increases prior to calving and subsequently decreases following several events.
 
... In the late 1990s the lower glacier began to thin due to intense submarine melting of a 10km long floating tongue . The resultant loss of buttressing forces caused the glacier to accelerate and stretch, causing the floating tongue to disintegrate and initiating a sustained retreat.

... There is a pattern of seasonal variation in terminus position superimposed on the long-term retreat that can be attributed to variations in the strength of the proglacial ice melange, a dense pack of icebergs that persists year-round, extends >15km down-fjord of the calving front, contains distinct shear bands and is densest near the terminus. Thus the ice melange may act like a weak, granular ice shelf that inhibits calving by transmitting stresses from the fjord walls to the glacier terminus.

... Seasonal variations in calving are important because they influence a glacier’s long-term rate of retreat through nonlinear processes; longer warm periods that extend the calving season could lead to greater rates of retreat.

... We used MODIS sea surface temperatures (thermal infrared emission) to generate a 12-year proxy record of daily fjord ice conditions in Ilulissat Icefjord

... the 11 April 2009 calving event was the first event of that year; the onset of the break-up of the seasonal floating tongue show an increase in melange mobility in the days leading up to calving. A profile of speed shows the floating
tongue advected into the field of view on 22 March. The tongue progressed steadily down-fjord over the next 20 days. By 8 April, the melange in front of the tongue started to speed up, and on 11 April a large calving event caused the floating tongue to disintegrate and disappear from the profile.

... The speed of the melange increased >40% over mean winter values in the days leading up to calving. Following the event, speeds were much slower as the advancing glacier terminus pushed the post-calving melange down-fjord and compacted it.

... The ability of the melange to prevent ice at the terminus from overturning decreases as the mobility of the melange increases, leading to more frequent calving during periods of enhanced mobility.

... Our record of melange conditions demonstrates that winters with persistently low surface temperatures, taken to indicate rigid and immobile ice melange, had few calving events which permitted the glacier terminus to advance several
kilometers and form a floating tongue.

... Though an ice melange may not have sufficient strength to directly impact the glacier force balance, it can impact glacier velocities indirectly by regulating calving.

...The ability of melange to influence calving seems to have diminished in recent years as conditions warmed and the glacier has retreated into deeper water.

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #581 on: February 22, 2015, 02:42:19 PM »
Quote
Let's hope sidd has time to pull out a few more of these contour maps for our favorite glaciers!
I will add Zachariae to that list!
I could sure use Petermann. As with Zach and Niag, there is a real question whether (or how) Morlighem's method takes the upheaval fields into consideration. I would say it doesn't.

That could be good news because then the discrepancy between Morlighem's replacement of bedrock kriging using physics-driven interpolation and what we can see from a rather dense grid then gives some clues as to what modifications the underlying basal sliding model needs to accommodate properties of the upheavals.

This situation is different than Jakobshavn and other deep canyons in that radar doesn't seem to have any problem getting to bedrock even through giant upheavals. However we don't know the composition of these upheavals so not the speed of microwave light through them so consequently not the true depth of bedrock (though it is a good fit to adjacent layer cake without any adjustments). It may be that the ice is moving too slowly lower Petermann for Morlighem's method to have major import.
« Last Edit: February 22, 2015, 05:13:43 PM by A-Team »

mspelto

  • New ice
  • Posts: 25
    • View Profile
  • Liked: 0
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #582 on: February 22, 2015, 09:15:51 PM »
On Feb. 12th there is some rifting near the calving front, these rifts are structural weaknesses that indicate calving is imminent.  Melange movement ahead of calving generally reflects changes in sub glacial water output.  The amount of calving here is not impressive except for it being February as you have all noted.  Given that the velocity of JI is faster than in the past, that alone makes such an event more likely.

sidd

  • First-year ice
  • Posts: 6774
    • View Profile
  • Liked: 1047
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #583 on: February 22, 2015, 09:25:39 PM »
"I wonder if Morlighem was limited to 100 m resolution ..."

The 100m contour spacing was chosen by me, the nc bed layer can be contoured to arbitrary interval, but the 100m contour already takes hours on the laptop i am travelling with ...

i do have access to heftier hardware, but not the time.

The contours were generated from qgis, but grass or gdal from command line will work also, if someone wants to try.

sidd

Shared Humanity

  • Guest
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #584 on: February 22, 2015, 11:53:58 PM »
Including the present calving front unless I am completely wrong?

This map certainly explains the calving that we are seeing along the north wall as the calving face retreats as this portion of the ice sheet is also grounded below sea level. It also explains the broad area of rifting that can be seen as this portion of the ice sheet accelerates to the now open water. It looks like further retreat by the glacier may open up another source of calving along the north wall.

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: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #585 on: February 23, 2015, 11:08:07 AM »
It's a while before they wake up in Colorado, but DMI are back at their desks this morning. Their Greenland field trip was to Qaanaaq, as part of the EU ICE-ARC project:

http://www.ice-arc.eu/2015/02/03/twilight-measurements/



I'll let you know if anyone there is willing and able to comment on the Jakobshavn calving.

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

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #586 on: February 23, 2015, 01:42:56 PM »
Quote
can be contoured with Qgis to arbitrary intervals but the 100m contour already takes hours on a laptop
 
Thanks so much for doing this, sidd. Sounds like a high risk operation that has the potential for tying up the computer all day only to wish later different settings had been chosen.

You are apparently the first person to ever do anything with the nc data. Nothing with remotely the detail of this map has ever been published. This is the case with so much of Greenland glaciology -- grievous under-analysis, or no analysis at all, of data. True, it is 'implicit' in the data file but that is like saying dark matter and black holes are 'implicit' in a photograph of an elliptic galaxy. 

I mopped up the color infill with special attention to sills between the potholes (which approach 1600 m below sea level). Having the channel in two dimensions paints a very different picture of the Jakobshavn south channel than what people have been modelling from (a one dimensional centerline that is neither centered nor representative) -- no wonder nothing has been predictable.

Going by this map and Espen's calving front overlay, it would appear that the centerline is currently sitting near the bottom of the overdeepening trough with 200m to climb (1.5 km?) before reaching a minor sill and beginning a 400 m descent into the first pothole. It is not so easy to accurately overlay Joughin 2014's centerline because of their unsatisfactory coordinate system.

I put the map into two tiles (no overlap) so it will display here at resolution.
« Last Edit: February 23, 2015, 02:41:20 PM by A-Team »

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #587 on: February 23, 2015, 07:50:26 PM »
Thanks Sidd and A-Team, these new maps explains much more than I already knew about Jakobshavn. ;)
Have a ice day!

sidd

  • First-year ice
  • Posts: 6774
    • View Profile
  • Liked: 1047
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #588 on: February 23, 2015, 07:53:33 PM »
"... has the potential for tying up the computer all day only to wish later different settings had been chosen ..."

ah, not so bad, but sometimes i forget that the thing is running in the background, and shut off the machine ... gaaa

"You are apparently the first person to ever do anything with the nc data."

No chance. Everybody knows about nc files, that and PDS (planetary data set) format have been around for a long while, and the tools are opensource (altho some of the data sets have to be run thru a fortran program to unpack, but still not so bad.) I really can't believe that Morlinghem or an associate hasn't done this already. There are error layers in that nc file, but i havent played with them yet.  That will fuzz out the contours.

Hell, i have dumped the nc files to text and then used gnuplot on them sometimes. But don't tell anybody.

sidd

sidd

  • First-year ice
  • Posts: 6774
    • View Profile
  • Liked: 1047
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #589 on: February 23, 2015, 07:55:29 PM »
 I wanted to ask, does anyone here run a unix ? i can guide em thru a nctools/grass/qgis/gdal setup. Then someone else who has more time can babysit this ?

Anyone ?

sidd

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #590 on: February 23, 2015, 09:12:44 PM »
sidd, is it a hassle to just output the image as a grayscale DEM, never mind the contours? I can take it from there in terms of re-paletting / 3D / bumpmap / perspective / hillshading / rendering /fly-throughs. I am also set up for cut / display /spling of transects.

It seems to me that a simple 8-bit file would suffice for the ice stream's path: the elevation range is 0 to -1600 m which would be 256 bins of 6 m each after thresholding out above sea level and contrast renormalization. I'll bet though they are carrying it at 16 or 32 bit which are ok for me too and probably better starting points if there have to be manipulations.

If they also have the error map as a grayscale, I can see how that sits relative to the sills and potholes. It can be embedded as a hue color channel over the DEM itself or used as a blur mask. The other thing to do assuming gaussian error is resample the DEM many times per the error distribution and see how robust the sills and troughs are. That could contribute the saturation channel for the HSV.

I recall Morlighem did put out a comparison somewhere with Bamber 2013 in the form of a pixelated postage stamp in a non-quantitative palette, so that's something else we can revisit.

Quote
these new maps explains much more than I already knew...
Likewise. This is really the 'missing map' for Jakobshavn. Nothing quite like reading a straightened-channel one-slice one dimensional modelling paper while having this map open.

Below is a mask for the area of interest, its overlay on the 21 Feb 15 Landsat LC80080122015052LGN00 (didn't have good markers for scaling and coregisration), and the B5 channel of that Landsat (note the contributing ice streams, fastest at top).
« Last Edit: February 23, 2015, 09:46:34 PM by A-Team »

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: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #591 on: February 23, 2015, 09:13:29 PM »
I wanted to ask, does anyone here run a unix ?

Anyone ?

Does Scientific Linux count? If so, then Moi! Should we repair to the Developer's Corner?
"The most revolutionary thing one can do always is to proclaim loudly what is happening" - Rosa Luxemburg

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: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #592 on: February 24, 2015, 12:29:35 AM »
NSIDC say they have no comment to make on the recent calving, and suggested that I try Jason Box instead. So I did:

https://twitter.com/GreatWhiteCon/status/570001605086588929
"The most revolutionary thing one can do always is to proclaim loudly what is happening" - Rosa Luxemburg

sidd

  • First-year ice
  • Posts: 6774
    • View Profile
  • Liked: 1047
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #593 on: February 24, 2015, 07:33:13 AM »
ok taking the discussion to developers corner

https://forum.arctic-sea-ice.net/index.php/topic,1165.0.html

sidd

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: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #594 on: February 24, 2015, 09:43:54 AM »
I have received a reply from Jason Box, who for those unfamiliar with the name is a Professor at the Geological Survey of Denmark and Greenland, and begetter of the Dark Snow Project. Here it is:

Quote
It's an interesting finding.

A cautious response: even if this calving were abnormal, we will likely see an advance in the next weeks that will fill the void. Why?

A.) this glacier flows fast and

B.) now with less flow resistance there will likely be an acceleration making the void filling happen even faster.

Here are annual end of melt season area changes measured by PROMICE.org. These are being updated. I will ask Karina Hansen today to update for 2014 and 2015. We could have that result in a few hours.

name ID lat lon region region ID width 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13
Jakobshavn 14 69.1788 -49.7329 W 105 11.4 -2.75016 -2.4734 -16.357427 -45.617225 -21.235024 -10.015401 -4.155551 -3.829812 -3.202087 -2.174176 -8.725405 -6.69317 -13.742999 -2.825888

In context of the ongoing retreat, I would speculate that this retreat could make further retreat more likely because the acceleration from B.) would cause 'dynamic thinning' that through a positive feedback would reduce glacier bed friction facilitating further thinning and acceleration. This feedback is an amplifier and not runaway but being activated would precondition Jakobshavn glacier for further retreat.

Sorry about the formatting!

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

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #595 on: February 24, 2015, 11:55:59 AM »
Quote
Sorry about the formatting!
Nice initiative, Jim. Might want to delete on of the duplicate posts. By 'end of melt season area changes", does he mean retreat of the calving front of Jakobshavn, measure not by a line as we have been doing but by the area in km2 between successive retreat lines? Net area works better for last year when the 2013 vs 2014 advance lines crossed.

That could be converted to volume by using the 850 m thickness, but as he points out, this does not give the total volume of annual ice discharge, only the balance between glacier advance and retreat.

The annual volume discharge is tantamount to drawing a line up-glacier that denotes all the ice that will make it to the front that season and figuring the volume using the thinning. (It's not so clear how to treat sub-glacial water discharge as some of that is pass-through meteoric water and some of it is melted ice.)

Nice resource there at PROMICE.org but I am not seeing where the Jakobshavn data is being stored. Have they gotten back to you with the 2014 and 2015 numbers?

Reformatting the areas, some very big years and no discernable trend other than continuing retreat (numbers are consistently negative). Some of the big years are attributable to loss of buttressing from the now-gone floating ice shelf, others may be do to melange being stronger/longer in very cold winter, still others could be impacts from more lubricating water in warm summers (though JI is already modeled as sliding).

name          Jakobshavn
ID                 14
lat                69.179
lon                49.733
region             W
ID               105
width             11.4

1999/00          -2.750
2000/01          -2.473
2001/02         -16.357
2002/03         -45.617
2003/04         -21.235
2004/05         -10.015
2005/06          -4.151
2006/07          -3.830
2007/08          -3.202
2008/09          -2.174
2009/10          -8.725
2010/11          -6.693
2011/12         -13.743
2012/13          -2.826
2013/14          -?.???
2014/15          -?.???

andy_t_roo

  • New ice
  • Posts: 61
    • View Profile
  • Liked: 1
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #596 on: February 24, 2015, 01:09:27 PM »
I recently found http://people.ee.ethz.ch/~luethim/pdf/script/pdg/chapter5.pdf -- on the physical properies of glaciers, and the numerical simulation of velocity and stress fields under various geometries.

Some of the thoughts here might aid discussions on future flow velocities of the various glaciers.

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: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #597 on: February 24, 2015, 01:50:04 PM »
Nice initiative, Jim. Might want to delete on of the duplicate posts.

This forum has a mind of its own! Duly deleted. This just in from Karina Hansen via Jason Box, who comments:

Quote
In the attached prepared by Karina Hansen you will see a light yellow polygon illustrating the end of melt season 2014 ice Jakobshavn front position retreated from the Feb 2014 (pink line) and Feb 2015 (green line) positions. The Feb 2014 and Feb 2015 positions are roughly the same with 2014 Feb being further retreated than Feb 2015.

The calving appears to be not abnormal at least in terms of where the front position is and therefore is not a shock but rather illustrates my points A. and B. of this morning's message.

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

A-Team

  • Young ice
  • Posts: 2977
    • View Profile
  • Liked: 944
  • Likes Given: 35
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #598 on: February 26, 2015, 05:28:59 PM »
By popular request, here is the extraordinarily clear, low angle 15 meter B8 band of Landsat 2015-02-21. Hopefully Espen can overlay the calving front over maximal retreat of last year.

PATH = 8 ROW = 12
LC80080122015052LGN00
SUN_AZIMUTH = 170.69201137
SUN_ELEVATION = 11.04439574

Espen

  • Young ice
  • Posts: 3705
    • View Profile
  • Liked: 420
  • Likes Given: 4
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #599 on: February 26, 2015, 07:29:00 PM »
Here it is:

Have a ice day!