Support the Arctic Sea Ice Forum and Blog

Author Topic: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland  (Read 800371 times)

nukefix

  • Frazil ice
  • Posts: 461
    • View Profile
  • Liked: 9
  • Likes Given: 1
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #950 on: September 17, 2015, 11:44:56 AM »
Here is a gif between Sep 6 and 12.  I don't see any major changes in the calving front, but note the difference in motion of the bergs labeled A, B, and C.  I believe berg A does not move, but appears to because of the different resolutions, sattelite position, and my matching errors.
I find it amazing that the iceberg C moved 18.5km in 12 days, which is some 1.5km per day. There must be a strong current that side of the fjord pushing the iceberg outward so quickly.

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #951 on: September 23, 2015, 10:46:43 AM »
The significance of the record-breaking calving event of 15 Aug 2015 — by far the largest ever documented at Jakobshavn — depends on earlier preconditioning of the crevasse system and later response of the glacier to the loss of nearly a kilometer of calving front depth.

We're in fairly good shape in terms of preconditioning imagery; the main issue is correlating ice-to-be-calved condition to flow bands. Recall surface flow lines can be followed tens of km upstream at Jakobshavn. Flow bands are bounding pairs of flow lines, at Jakobshavn most usefully marking out ice origination sources. Less than a third of the calving front tracks back to the deep interior, the rest comes in from a few sources on the south side.

One explanatory scenario posits that calving was delayed for a week or two prior to the event (maybe having to do with still/trough/elbow position/late melt season) while preconditioning continued all the while. Calving then suddenly caught up with preconditioning on the 15th, giving rise to the extraordinary event.

This would leave very little preconditioned ice available for calving post-event and so predicts the calving front will simply march forward at the ~ 40 m/day rate of glacial advance with little new debris in the water. That's in fact what is observed in the first 3 post-event Landsats (1st animation). A cyclic repetition of this scenario in future summers is then predictable.

A more drastic scenario interprets the event as the initial marker of Jakobshavn entering a collapse phase (2nd animation). Here we have zero guidance from glacial modelling because these ignore the elbow, narrowing and have the ocean interface, temperature and bedrock profiles and till depth completely off.

'All models are wrong and frequently none of them are useful.'

Longer term, the data situation is very unfortunate: we’ve had weeks of clouds this September (just like 2014) when the glacier’s response can only be monitored with less satisfactory Sentinel, Radarsat, Aqua and Terra. These don't have sufficient resolution to map preconditioning features.

In addition, it appears to me there has been a rare screw-up with Landsat targeting resulting in the loss of critical data -- the 17 Sep 15 Landsats LC80080122015260LGN00 and LC80080112015260LGN00.

Although those two have been in the orbital pattern for almost three years now, the help desk is claiming Jakobshavn was never on the acquisition list for that date, pointing to http://landsat.usgs.gov//L8_Pend_Acq/y2015/Sep/Sep-17-2015.txt (vary the 17 to see path,row for other dates).

It’s a pity that the scientific community didn’t bestir itself enough to put a daily plane in the air from Illulisat ($200 an hour?) and monitor the glacier from under the clouds at photographic resolution. Maybe someone did via a south shore drone but is not disclosing prior to later publication. Fields like astronomy have a well developed hot line and response plan for transient events, not glaciology though.

It's possible that proprietary high resolution imagery like WorldView3 exists; hopefully someone has the budget for that. Otherwise we are in a race with time, clouds and shortening daylight with Landsat8 for any further post-event details. It's quite possible we will never be any real understanding of this event.

I'll attach all this imagery in a bit.

2015 276 08 11 LC80080112015258LGN00 08-OCT-15 Landsat8 coming?
2015 274 10 11 LC80100112015274LGN00 01-OCT-15 Landsat8 coming?
2015 267 09 11 LC80090112015267LGN00 24-SEP-15 Landsat8 coming tomorrow?
                                     20-SEP-15 Sentinel-1 coverage
                                     18-SEP-15 Sentinel-1 coverage
2015 260 08 12 LC80080122015260LGN00 17-SEP-15 Landsat8 data lost or never acquired
2015 260 08 11 LC80080112015260LGN00 17-SEP-15 Landsat8 data lost or never acquired
2015 258 10 11 LC80100112015258LGN00 15-SEP-15 cloudy
                                     21-SEP-15 AQUA clear
                                     14-SEP-15 AQUA clear
                                     13-SEP-15 AQUA clear
                                     13-SEP-15 Sentinel-1 coverage
                                     10-SEP-15 Radarsat-2 coverage
2015 251 09 11 LC80090112015251LGN00 08-SEP-15 Landsat8 clear
                                     08-SEP-15 AQUA clear
                                     07-SEP-15 AQUA clear
                                     06-SEP-15 AQUA clear
                                     06-SEP-15 Sentinel-1 coverage
                                     03-SEP-15 AQUA clear
                                     03-SEP-15 Radarsat-2 coverage                                   
                                     01-SEP-15 Sentinel-1 coverage
2015 244 08 11 LC80080112015244LGN00 01-SEP-15 Landsat8 cloudy
                                     31-AUG-15 Radarsat-2 coverage
2015 242 10 11 LC80100112015242LGN00 30-AUG-15 Landsat8 cloudy
                                     27-AUG-15 AQUA clear
                                     27-AUG-15 Sentinel-1 coverage
                                     26-AUG-15 AQUA clear
                                     25-AUG-15 AQUA clear
                                     25-AUG-15 Sentinel-1 coverage
                                     24-AUG-15 Radarsat-2 coverage
                                     24-AUG-15 AQUA clear
                                     23-AUG-15 AQUA clear
2015 235 09 11 LC80090112015235LGN00 23-AUG-15 Landsat8 clear
                                     22-AUG-15 AQUA clear
                                     20-AUG-15 Sentinel-1 coverage
                                     18-AUG-15 AQUA clear
                                     19-AUG-15 AQUA clear
                                     17-AUG-15 Radarsat-2 coverage
                                     17-AUG-15 AQUA clear
2015 228 08 11 LC80080112015228LGN00 16-AUG-15 Landsat8 clear
                                     16-AUG-15 AQUA clear
                                     15-AUG-15 AQUA clear
2015 226 10 11 LC80100112015226LGN00 14-AUG-15 Landsat8 clear
                                     14-AUG-15 AQUA clear
                                     13-AUG-15 AQUA clear
                                     13-AUG-15 Sentinel-1 coverage
                                     12-AUG-15 Sentinel-1 coverage
                                     10-AUG-15 Radarsat-2 coverage
                                     08-AUG-15 Sentinel-1 coverage
2015 219 09 11 LC80090112015219LGN00 07-AUG-15 Landsat8 cloudy
                                     06-AUG-15 AQUA clear
                                     05-AUG-15 AQUA clear
                                     04-AUG-15 AQUA clouds
                                     03-AUG-15 Radarsat-2 coverage
                                     01-AUG-15 Sentinel-1 coverage
                                     02-AUG-15 AQUA clear
                                     01-AUG-15 AQUA clear
                                     31-JUL-15 AQUA clear
2015 212 08 11 LC80080112015212LGN00 31-JUL-15 Landsat8 clear
2015 210 10 11 LC80100112015210LGN00 29-JUL-15 Landsat8 clear
« Last Edit: September 23, 2015, 11:17:41 AM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #952 on: September 25, 2015, 12:37:05 AM »
The next Landsat8 just came in, 16 days after the last cloud-free image of. 08 Sep 15. The new image LC80090112015267LGN00 is too cloudy to provide details but the outline of the calving front can be discerned.

I looked into the timing a bit. There is a 7 hour time difference between Landsat's Greenwich Mean Time and my location in Tucson, AZ which mercifully does not observe daylight savings time. Looking at their posted capture schedule, the Jakobshavn scene was to be acquired at 15:00 GMT by the Svalbard (SGS) ground station. So that would be 8:00 am here. It took 7 hours to get the image processed and accessible at Earth Explorer, which is decent enough.

I have to do the big download in the wee hours because of bandwidth limitations on the Excede satellite and will process band 8 at that time.

However you can see already that the odd angular shape of the calving has not changed much but has simply moved downstream, still in line with the first scenario above. At better resolution, it would be worthwhile to take a closer look for calved bergs since August 16th.
« Last Edit: September 25, 2015, 01:11:58 PM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #953 on: September 25, 2015, 01:30:55 PM »
On the high resolution Landsat8 LC80090112015267LGN00_B8 of Sept 24th, which has the same path,row geometry as Sep 8th but low angled solar illumination, the clouds are not too bad in the vicinity of the calving front. I've doubled down on the ground resolution to 7.5 m and added a faint yellow grid scale of 30 pixels which amounts to 225 m x 225 m boxes.

The grid is tilted at 42º so that it follows glacial flow lines after the elbow, ie four of these boxes in from the calving front is 900 m back, or a month of motion at 30 m/day, representing the ice that will be calved off were calving front to remain stationary (more likely, it will move westward).

Note calving is not at all perpendicular the flow but penetrates much deeper near the south wall. I've never seen the calving front cleaved strictly perpendicular to the flow but more typically has been arched deeper in the center (where the velocity is highest).

The current calving front line does not correlate either with flow bands (defined by upstream ice origins) or transverse velocity profile. The area 1-2 km upstream seems unusually cross-crevassedsuggesting it is preconditioned to calving. With the melt season ending, it's not clear whether we'll see that or not.

The alligator skin region is tinted reddish in the second image below, its eastern boundary partly obscured by clouds. The yellow over-tint indicates an odd region already present in the Aug 16th post-event Landsat that appears to have broken off from the calving front but not made its way into the fjord, perhaps because it is supported by a bedrock ledge below or is leaning back against the calving front in a quasi-stable manner.

If this cross-crevassed region represents a new and ongoing development, it would favor the second scenario (two posts back) in which Jakobshavn has crossed the Rubicon into an altogether new and unstable phase. However it is not so easy to compare images because of clouds, different sun angles, different melt extent, and possibly snowfall (or rain).

2015 290 10 11 LC80100112015290LGN00 17-OCT-15 Landsat8 next scene
2015 283 09 11 LC80090112015283LGN00 10-OCT-15 Landsat8 clear                         sun 13.7º
2015 274 10 11 LC80100112015274LGN00 01-OCT-15 Landsat8 cloudy                       sun 17º
2015 267 09 11 LC80090112015267LGN00 24-SEP-15 Landsat8 partly cloudy              sun 20º
2015 251 09 11 LC80090112015251LGN00 08-SEP-15 Landsat8 clear                           sun 26º
2015 235 09 11 LC80090112015235LGN00 23-AUG-15 Landsat8 clear                           sun 32º
2015 228 08 11 LC80080112015228LGN00 16-AUG-15 Landsat8 clear first post-event   sun 34º

2015 226 10 11 LC80100112015226LGN00 14-AUG-15 Landsat8 clear last pre-event
2015 212 08 11 LC80080112015212LGN00 31-JUL-15 Landsat8 clear
2015 210 10 11 LC80100112015210LGN00 29-JUL-15 Landsat8 clear


With Sentinel, we can follow the position of the calving front even in the dead of winter, though glacial velocity (which drops somewhat in winter) is not so easily measured.  Landsat scenes are still taken ± 43 days about Jan 1st, so there is about a three month hiatus there.
« Last Edit: October 13, 2015, 02:08:50 PM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #954 on: September 26, 2015, 01:21:07 PM »
Below is a sequence of Sentinel-1 and Radarsat-2 covering before and after of the record calving event. These are simply taken from the low resolution snapshots available from DMI and expanded 3x in resolution. It would have been better to download the higher resolution images but these don't seem to be in the public domain for Radarsat.

The images are fairly well aligned on rock control points but do not make for a good time series animation because it turns out DMI too has not fully resolved geo-coding issues with radar data. In other words, the images are rotated correctly to match Aqua etc but their geometry otherwise does not quite match. This makes for a jerky animation.

The other issue is that some days the fjord melange and calving area are quite bright but some days they are dark. This may have to do with the ice surface being rougher/smoother depending on daytime temperature melt affecting its radar reflectance.

The first image needs a click to display at its full 700 pixel width. I will add some full resolution Sentinels in a bit; there is an IW image for Sept 24th. Note the confluence area is more distinct in Sentinel than Landsat (2nd image, far right). The third image (which is rotated CW 22º) shows the calving region on Sept 25th.
« Last Edit: September 26, 2015, 02:12:40 PM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #955 on: September 27, 2015, 12:18:14 PM »
Once upon a time, Jakobshavn Isbrae was not grounded at the calving front but instead had a freely floating ice shelf. Later, like all but 7 of Greenland's many glaciers, it lost its ice shelf. Today Jakobshavn is grounded on till over bedrock in its entirety (with only frozen-on melange and related extensions in winter for temporary ice shelf).

It's easy to tell from aerial photos -- which go back 65 years -- when Jakobshavn lost its ice shelf. That's because ice shelves rift off large stable tabular pieces that remain top-side up whereas calving at a grounded terminus yields unstable  irregular blocks of ice that roll over on their sides.

The Landsat7 image below was taken in 26 May 2003, some months after a large section of the ice shelf disintegrated into dozens of tabular blocks, each retaining the surface pattern of the remaining ice shelf. Not one of these blocks subsequently turned over.

Thus prior to this event, Jakobshavn was not grounded at the bottom of the fjord but instead had a free floating ice shelf whose thickness could have been readily determined from the freeboard of the tabular cleavage products.

Jakobshavn no longer calves tabular pieces. Instead, the lower terminus is highly crevassed (unlike an ice shelf) and large blocks of ice fall from the front of calving front (rather than drift away from interior rifts in an ice shelf).

Since calving of a glacier at a grounded terminus is a wholly different process than rifting of an ice shelf, it makes no sense whatsoever to compare them. The former is gravitationally driven physics whereas the latter, at hydrostatic equilibrium, is driven by horizontal shearing (tides, currents, wind, pinning points, bottom melt). Because of this, calved ice contributes to sea level rise but ice rifted from an ice shelf does not.

Comparing like to like, the event of 15 Aug 2015 was indeed unprecedented by a wide margin. It's fair to say though that the relevant period of comparison is fairly short, perhaps only five years.

The significance lies not so much in setting a volume record but rather in what happens next: has upstream Jakobshavn become unstabilized? We may or may not see an effect this year before winter sets in.

Jakobshavn is currently an immense mass of rapidly moving ice forced to alter course to make the sharp bends of the elbow region. Having the ice stream front retreat above the elbow removes a prodigious amount of resistance, which will only be partially offset by new buttressing as sides flow in and compete for channel. This will cause the ice stream to speed up and draw in an ever-widening area of ice sheet beyond its current 6%.

AbruptSLR

  • Multi-year ice
  • Posts: 17519
    • View Profile
  • Liked: 715
  • Likes Given: 236
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #956 on: September 27, 2015, 07:09:35 PM »
The linked NASA article discusses how an extreme "Spring Tide" could have contributed to an iceberg from the Jakobshavn Glacier (see image) to sink the Titanic.  While this article focuses on the likelihood that the spring tide refloated deep draft icebergs, we should all remember that such extreme spring tides can also promote iceberg calving events (both in Greenland and in Antarctica):

http://sservi.nasa.gov/articles/did-the-moon-contribute-to-the-sinking-of-the-titanic/

Extract: "As Greenland icebergs travel southward, many become stuck in the shallow waters off the coasts of Labrador and Newfoundland. Normally, icebergs remain in place and cannot resume moving southward until they’ve melted enough to refloat or a high enough tide frees them. A single iceberg can become stuck multiple times on its journey southward, a process that can take several years. But the unusually high tide in January 1912 could have been enough to dislodge many of those icebergs and move them back into the southbound ocean currents, where they would have just enough time to reach the shipping lanes for that fateful encounter with the Titanic."
“It is not the strongest or the most intelligent who will survive but those who can best manage change.”
― Leon C. Megginson

Espen

  • Young ice
  • Posts: 3251
    • View Profile
  • Liked: 177
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #957 on: September 27, 2015, 07:28:54 PM »
The linked NASA article discusses how an extreme "Spring Tide" could have contributed to an iceberg from the Jakobshavn Glacier (see image) to sink the Titanic.  While this article focuses on the likelihood that the spring tide refloated deep draft icebergs, we should all remember that such extreme spring tides can also promote iceberg calving events (both in Greenland and in Antarctica):

http://sservi.nasa.gov/articles/did-the-moon-contribute-to-the-sinking-of-the-titanic/

Extract: "As Greenland icebergs travel southward, many become stuck in the shallow waters off the coasts of Labrador and Newfoundland. Normally, icebergs remain in place and cannot resume moving southward until they’ve melted enough to refloat or a high enough tide frees them. A single iceberg can become stuck multiple times on its journey southward, a process that can take several years. But the unusually high tide in January 1912 could have been enough to dislodge many of those icebergs and move them back into the southbound ocean currents, where they would have just enough time to reach the shipping lanes for that fateful encounter with the Titanic."

I really doubt this theory, mainly because sea ice conditions was way different back then (1912) and since the accident happened in the month of April, months (3) before the melt season.
If it was an iceberg it could be one from many glaciers not only Jakobshavn, it was either a freak accident (a late iceberg) or maybe even a drunk crew!     
Have a ice day!

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #958 on: September 28, 2015, 01:19:21 PM »
Quote
really doubt this theory.. sea ice conditions very different in 1912 ... accident happened on April 15th ... before the melt season
Likewise. If Jakobshavn had a floating ice shelf back then (which follows from the known position of the calving front in that year, historic accounts and fjord bathymetry, second image below), then it would have rifted off tabuar icebergs, unlike the jagged berg the Titanic hit.

The calving front did recede ~5 km between 1902 and 1913 but that is only 5 km2 per year at fjord width Thus iceberg production in that decade was very minor compared to what currently comes down Nares. The Landsat7 animation for the drawn-out event of 2002-03 shows shattering of tabular pieces.

The calving event itself would have been from previous summers as the drift route and April 15th date of the accident imply a mid-winter calving date, implausible for colder seas and air. In January 1912, I would expect the entire fjord to be strongly frozen melange.

The tides are poorly known even today inside major Greenland fjords and the whole of Nares Strait; see the icy seas post on that by Andreas Muenchow: very little constraining data, large-celled global ocean tidal models. A higher tide than usual seems to require special pleading on buoyancy conditions to float grounded icebergs west off the terminal morraine at Illulisat.

That is, large icebergs will remain stuck on the fjord east of the moraine sill rise; refloating just puts them back in the fjord. From the Gladish/Holland articles, I am recalling a maximal depth of 250 m for passage out into Disco Bay. This has the effect of screening Jakobshavn's output to bergs which have already experienced a lot of bottom melt.

Incidentally, Espen asked a while back who should be credited for researching historic calving front lines. That would be Bauer for everything up to 1964 and Podlech for 1964-2003, though many many people have redrawn this figure.

In terms of calving front time series, we've seen people talking about linear recession, area losses, and volume records. This all needs to be recast in terms of calving front location relative to the grounding line. That is, to lose a linear km of thin floating ice shelf in a broad fjord has a very different physical significance than a grounded glacier retreating a km into a deep narrow rock funnel with reduced buttressing.

While linear retreat has visual appeal as a simple numerical measure of climate change, it is actually a poor proxy when the physics of retreat has changed character (eg Jakobshavn). And it's good to recall that Store Glacier, bathed in the same ocean water as Jakobshavn, hasn't retreated at all since 1948 (though it has sped up and thinned during that time.)

Bauer, A. 1968: Missions aériennes de reconnaissance au Groenland 1957–1958. Observations aériennes et terrestres, exploitation des pho- tographies aériennes, détermination des vitesses des glaciers vêlants dans Disko Bugt et Umanak Fjord. By A. Bauer, in collaboration with M. Baussart, M. Carbonnell, P. Kasser, P. Perroud and A. Renaud. Meddelelser om Grønland 173(3), 116 pp.

Podlech, S. and A. Weidick (2004), A catastrophic break-up of the front of Jakobshavn
Isbræ, West Greenland, 2002/2003, J. Glaciol., 50(168), 153–154.
[source for lower 3 illustrations]
« Last Edit: September 28, 2015, 01:46:03 PM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #959 on: September 28, 2015, 04:26:01 PM »
The only reason this click-bait iceberg story is being trotted out yet again is because of last night's full lunar eclipse. It was very upsetting to our local coyotes, as are solar eclipses.

It's worth noting that the record tide paper was not published in a reviewed scientific journal; indeed as the editor of Sky & Telescope magazine was a co-author so approval was a foregone conclusion.

There was quite an interesting rebuttal published the same year by Lane Wallace, who went to quite a bit more trouble than the original authors to interview relevant Canadian experts: there was nothing exceptional about 1912 ice bergs.

http://www.theatlantic.com/technology/archive/2012/03/no-the-moon-did-not-sink-the-titanic/254291/

During the course of modern sea level rise re-analysis, tide gauge data from all over the world was thoroughly reviewed as far back as it went. I don't know where that is stored or how accessible mid-April 1912 might be. I am going to guess  that high tide was not especially remarkable in percentage terms, that is, it wouldn't stand out on a plot of the last hundred years or at most represent a few mm over a typical yearly highest tide.

The animation below of ice at the outlet moraine of Jakobshavn does show ice bergs hanging up there but not for long -- these days they don't need anything special by way of tides to make their exit into Disco Bay.

Quote
Reanalysis of 1985 aerial photos shows substantial submarine melting of the floating ice tongue of Jakobshavn. The thickness of the floating tongue determined from hydrostatic equilibrium tapers from ∼940 m near the grounding zone to ∼600 m near the terminus. Speeds on the July 1985 ice tongue are nearly constant at 18.5 m/d indicating negligible dynamic thinning [just like Petermann].

The thinning of the ice tongue is mostly due to submarine melting with average rates of  0.62 m/d between the summers of 1984 and 1985. The cause of the high melt rate is the circulation of warm seawater beneath the tongue with convection driven by discharge of subglacial freshwater from the grounding zone. Buoyancy-driven convection is responsible for a deep channel incised into the sole of the floating tongue [just like Petermann].
http://onlinelibrary.wiley.com/doi/10.1029/2009JF001632/full
« Last Edit: September 28, 2015, 05:29:03 PM by A-Team »

AbruptSLR

  • Multi-year ice
  • Posts: 17519
    • View Profile
  • Liked: 715
  • Likes Given: 236
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #960 on: September 28, 2015, 05:48:11 PM »
Whether NASA's speculative theory about icebergs and the Titanic is correct, or not, it is still valuable to think about the influence of strong Spring Tides on future calving events for Jakobshavn, as the meltwater discharge stream emerging from beneath Jakobshavn would communicate the increased buoyancy during a Spring Tide well upstream beneath the calving front.  This increased hydrostatic head beneath this marine terminating glacier could help to destabilize the calving front, resulting in an increased probability of iceberg calving events during future strong Spring Tides.
“It is not the strongest or the most intelligent who will survive but those who can best manage change.”
― Leon C. Megginson

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #961 on: September 29, 2015, 01:57:20 PM »
Quote
increased hydrostatic head beneath this marine terminating glacier could help to destabilize the calving front, resulting in an increased probability of iceberg calving events during future strong Spring Tides
Where might we find a quantified list of strong tides for upper Illulissat Fjord for the next few years?

It is better to make predictions now than scratch around retrodictively for a an explanatory fit after an event. If the issue is false positives (big tide but no calving), that would leave matters in correlation-is-not-causation land.

Experimentally, we are in far better shape with tidal hinging on Petermann ice shelf because of the differential GPS units emplaced there this August by A Muenchow's group. That approach is not feasible on Jakobshavn because of the rapidity of crevasse widening and height ablation.

Based on Rignot's sonar scans of calving fronts  just to the north, everyone expects a firehose of meltwater emerging from the ice/bedrock interface in season at Jakobshavn but the fact remains that there is not the slightest shred of experimental evidence for this, much less seasonal quantitation or extent of buoyant mixing of frontal sea water.

Indeed the thickness of ice, depth to bedrock and amount of till at the calving front Jakobshavn remain exceedingly problematic. The very fact of Jakobshavn being the world's most extreme glacier bakes in the experimental difficulties in characterizing it.

The scientific community has been extremely evasive on the immediate vicinity of the calving front. We have no real idea where the calving front sits today relative to sills and troughs. You will never see a published bedrock profile overlain on a high resolution 2015 Landsat, which cannot be represented by a single flowline in any event because of pits and hummocks (as in Morlighem's description).

Instead every article shows a thick hand-drawn line obscuring an undated blurry low resolution image. This is partly because radar tracks have NOT been successful in obtaining clear bedrock reflections and partly because actual data points have to be heavily augmented by interpolated unphysical kriging (or untested mass conservation modeling).

The two recent papers on Antarctic glacier instability (mechanism proposed in a 1974 Weertman paper) have put reverse slope beds in the limelight (eg Hansen paper). The graphic below shows the flowline profile assumed in the 2015 Hughes/Gogenini paper cited above by sidd in #951. (I've added topographic analysis to Fig.15.)

http://www.skepticalscience.com/print.php?n=2837

However I question how applicable this really is to actual Greenland glaciers. Neither Jakobshavn nor Petermann has monotonically declining bed elevation, while the flow of Greenland's largest ice stream NEGIS (Zachariae) disrespects basal topography altogether. There are many issues at play here other than mere bed grade.

Treating Jakobshavn as a classical reverse slope situation seems really problematic given its 19 distinct transitions between reverse and positive slope over only 60 km (neglecting till which is reported at AGU2015 as up to 100 m thick presumably at overdeepenings). I'll add the mean slopes (meters of drop/km of length) of these zones in a bit.
« Last Edit: September 29, 2015, 02:43:13 PM by A-Team »

AbruptSLR

  • Multi-year ice
  • Posts: 17519
    • View Profile
  • Liked: 715
  • Likes Given: 236
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #962 on: September 29, 2015, 04:16:10 PM »
The attached image of Ilulissat Icefjord by Gladish et al (2015), provides a reasonable conceptual model for discussing the hydraulic boundary conditions for the Jakobshavn Glacier calving front (including sub-glacial basal meltwater discharge).   While the following abstract from Djoumna & Holland (2014) makes it clear that during a spring tide the Ilulissat Icefjord acts like a tidal jet-fjord that introduces significant amount of ocean heat content and turbulent mixing to the fjord's water, indicating that spring tides not only increase hydrostatic pressure beneath the calving front but also can increase the water temperature in the fjord.

1. CARL V. GLADISH, DAVID M. HOLLAND, AQQALU ROSING-ASVID, JANE W. BEHRENS AND JESPER BOJE (2015), "Oceanic Boundary Conditions for Jakobshavn Glacier. Part I: Variability and Renewal of Ilulissat Icefjord Waters, 2001–14"

2. Djoumna, G.; Holland, D. M. (2014), "Oceanic Mixing Processes in Disko Bay-Ilulissat Icefjord System: Can We Quantify the Heat Loss from the Atlantic Water Layer?"

Abstract: "Ilulissat Icefjord (IIF) has been the site of few hydrographic observations. Recent hydrographic data from Disko Bay (DB) showed a significant warming from the below of the cold Polar Water entering DB from the mid-to-late 19901990s onward, and the sill at the fjord mouth prevented the modified West Greenland Irminger Waters (WGIW) to fill the IIF basin. Here we identify and attempt to quantify the fluxes associated with the small-scale processes that contribute to the upward diapycnal fluxes of heat, salt and from the modified WGIW to the surface-mixed layer using five years of summer data (2009 - 2013). The interaction between the WGIW and Egedesminde Dyb trough cutting across the continental shelf from the shelf break into DB creates the cold/warm layering of water masses which contribute to the formation of double diffusive thermohaline staircases. We found evidence of thermohaline staircase consists of series of sharp interfaces across which both TT and SS increase with depth separated by thick well-defined convective layers. We hypothesize that the warming of the PW layer in DB may have been caused by the upward heat fluxes from the AW driven double diffusive convection. Vertical heat fluxes estimated from laboratory-based flux laws for the diffusive regime of double diffusive convection were up to 0.20.2Wm 2^{-2}. The other major player responsible for the Atlantic water heat loss is shear instabilities in the internal wave and tides generated by barotropic tidal flow over the Egedesminde Dyb trough, the continental shelf and across the sill at the entrance of IIF. Using moored pressure data, we found that the fjord could be described as a wave-fjord during neap tide and turns into a tidal jet-fjord during spring tide, therefore a weak nonlinear response due to supercritical conditions with flow separation over the sill and a linear baroclinic tidal response due to the deeper right side of the sill could be expected. We found enhanced eddy diffusivity KρK_{rho} and dissipation rate ɛepsilon up to 1.1×10 41.1times 10^{-4}m2^{2}s 1^{-1} and 9.81×10 69.81 times 10^{-6}Wkg 1^{-1} respectively in the deeper water inside IIF extending over an area more than 1010km from the sill depth. We attribute this high level of mixing rate to an internal wave radiated far away from the sill where it was previously generated."
“It is not the strongest or the most intelligent who will survive but those who can best manage change.”
― Leon C. Megginson

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #963 on: September 29, 2015, 05:50:01 PM »
In view of the Rignot 2015 paper, the Morlighem paper, the Ryan/Box drone paper at Store, and the two shocking AGU2015 abstracts on Jakobshavn, I am done with looking at speculative models and cartoon depictions of Jakobshavn's calving front.

The Rignot paper found that the latest IBCAO bathymetry of nearby fjords was grievously in error (way too shallow, missing channels and sills). The side-looking radar also found multiple meltwater exits on one glacier, excavated caverns in the calving front wall at bedrock, and observed meltwater volume and exit velocity.

The drone paper established that hydro-fracturing plays no role whatsoever in central or northern Greenland calving front dynamics.

Glaciology has long been a descriptive science, an oxymoron to be sure but seemingly unavoidable in the case of Jakobshavn. Experimentalists have struggled for years to determine even the most basic glacier parameters: fjord geometry, temperature profile, rheology, strain, stratification, bed friction, till location and hydration, and meltwater channels.

My sense however is that we may finally turn the corner in the coming year. The data is doable; the papers above prove that. At that point, it might be timely to think about models.

I struggle to see what today's models -- we're talking here about dozens and dozen of papers, a 20:1 ratio to field work -- are accomplishing when their assumptions are unmotivated, simplistic and demonstrably wrong from the get-go. The inconvenient elbow -- there's not a single paper that even acknowledges its existence. The confluence, what confluence? drainfield capture, what's that?

Not one of these model papers has ever issued a testable prediction. Much less gotten it right.

The brouhaha over whether the 16 Aug 15 was a 'record' or not -- all I learned from that is that the scientists weren't keeping any records. Not one had thought through what that should even mean. They're totally reliant on us to look at the incoming satellite data. I've come to the conclusion they're totally reliant on us for its interpretation as well.

Meanwhile, Air Zufari makes cheap daily flights over the Jakobshavn calving front in a 5-seater Partenavia P68. These are actually flown by Copenhagen Air Taxi. Numerous prominent glaciologists are based in Copenhagen. So where are the daily or even weekly high resolution videos? Only one has surfaced, an undated May 2014 handheld shot out the plane window by a 13-year old.

Where is the drone data? Two thousand dollars, off the shelf fully instrumented, autopilot, 0.5 m resolution. Illulisat: hot showers, thai food, dorm rooms, nearby, grad students and capable volunteers by the dozens. Instead we wait for the next haphazard 15 m Landsat for velocities averaged over half a month (often 2x that) of the melt season, to be published fifteen months after the fact.
« Last Edit: September 29, 2015, 06:02:57 PM by A-Team »

AbruptSLR

  • Multi-year ice
  • Posts: 17519
    • View Profile
  • Liked: 715
  • Likes Given: 236
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #964 on: September 29, 2015, 06:56:48 PM »
The following link leads to 2015 & 2016 tide tables for various harbors in Greenland including for Ilulissat Harbor:

http://www.dmi.dk/en/groenland/hav/tidevandstabeller-groenland/
“It is not the strongest or the most intelligent who will survive but those who can best manage change.”
― Leon C. Megginson

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #965 on: September 30, 2015, 12:59:23 PM »
Quote
2015 & 2016 tide tables
Thanks, very helpful. I am very curious as to what Andreas Muenchow will actually find up in the Petermann fjord (which has no harbor). I'm guessing ship GPS is so accurate (given their supplemental mountain top station) that tides can be measured accurately when moored in still water adjacent to the calving front. Icyseas has some representative data for a tidal cycle in Nares Strait. I explained the decomposition of tides into astronomical components at a recent Petermann post.

Exactly how an incoming high tide will mechanistically lift the long Petermann ice shelf is up in the air, though there's older interferometric analyses. In terms of tides and ice shelf instability, research in Antarctica is way ahead of Greenland (as usual):

http://www.pik-potsdam.de/~ricardaw/publications/rosier_green14.pdf free full 2014
http://onlinelibrary.wiley.com/doi/10.1002/2013JC009605/full free full 2014 tides last 21 kyr
http://www.sciencedaily.com/releases/2006/12/061221075130.htm

Quote
Nature research reveals that tidal motion influences the flow of the one of the biggest ice streams draining the West Antarctic Ice Sheet. This unexpected result shows that the grounded Rutford Ice stream varies its 1 m/d speed by as much as 20% every two weeks. 

Glaciologist Hilmar Gudmundsson says 'We've never seen anything like this before. The discovery that the spring-neap tidal cycle exerts such a strong influence on an ice stream tens of kilometres away is a total surprise. For such a large mass of ice to respond to ocean tides like this illustrates how sensitively the Antarctic Ice Sheet reacts to environmental changes. Glaciologists now to rethink how the Antarctic Ice Sheets reacts to external forces. '

The variations in flow of the Rutford Ice Stream are related to the vertical motion of the ocean caused by the gravitational effects of the sun and moon. Every two weeks sees large spring tides followed by small neap tides. Scientists expected movement of the floating ice shelves, but not grounded ice streams.
At Jakobshavn itself, the's been placement of traffic cones on the calving front (for laser displacement) and a recent high precision study of movement during calving from a seismic perspective. Tidal effects will change as the ice stream retreats into the narrowing south branch elbow. Here's a 2007 study of tidal movement in 2004, the last year for an ice shelf. The tides at Illulisat Harbor matched very closely those at a station near the calving front (small white square, first image):

Quote
The photogrammetric survey provided flow trajectories for 4000 surface points with a time resolution of 30 min. These flow trajectories were used to compare the vertical motion of the glacier with the observed tides. The existence of a free-floating glacier tongue in 2004 was confirmed by these data. However, it occupied only a small belt, of at most a few 100 m width, in the central part of the glacier front. Horizontal motion did not depend on tidal phase, unlike some of the fast-moving ice streams of West Antarctica.

http://onlinelibrary.wiley.com/doi/10.1029/2006JF000601/full free full

Meanwhile, I rescaled the Jakobshavn profile shown in #963 so that its scale becomes 10 meters of bedrock change per km of flowline. (The original scale is 8.66:1.) Gimp's rectangle tool provides run over rise; that's converted to slope in units of meters per km in the figure for each segment of the bedrock profile to facilitate comparison to retrograde slopes in the Antarctic literature.
« Last Edit: September 30, 2015, 11:32:58 PM by A-Team »

nukefix

  • Frazil ice
  • Posts: 461
    • View Profile
  • Liked: 9
  • Likes Given: 1
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #966 on: September 30, 2015, 03:36:38 PM »
Meanwhile, Air Zufari makes cheap daily flights over the Jakobshavn calving front in a 5-seater Partenavia P68. These are actually flown by Copenhagen Air Taxi. Numerous prominent glaciologists are based in Copenhagen. So where are the daily or even weekly high resolution videos? Only one has surfaced, an undated May 2014 handheld shot out the plane window by a 13-year old.

Where is the drone data? Two thousand dollars, off the shelf fully instrumented, autopilot, 0.5 m resolution. Illulisat: hot showers, thai food, dorm rooms, nearby, grad students and capable volunteers by the dozens. Instead we wait for the next haphazard 15 m Landsat for velocities averaged over half a month (often 2x that) of the melt season, to be published fifteen months after the fact.
What do you wish to accomplish with the ultra high resolution data? Build models that could predict calving events and retreat well into the future?

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #967 on: September 30, 2015, 11:51:11 PM »
Quote
What do you wish to accomplish with the ultra high resolution data? Build models that could predict calving events and retreat well into the future?
For now, I would just like to have a record of the exceptional calving events from above, as the oblique perspective from the rocks is difficult to rectify and increasingly obscured by retreat; also the frequency of cloud-free coverage (Landsat) and resolution of Sentinel and Terra/Aqua are unsatisfactory.

At this point we have little understanding of what preconditions ice near the front to calve, whether or not there is standing water in the crevasses, what explains the consistent asymmetry of the calving front with respect to the ice stream, where the front stands in relationship to emerging basal meltwater jets, what the ocean/front looks like at the ice/bedrock/till interface, what non-averaged high temporal resolution velocity and acceleration fields look like, and where the front is relative to sill, slope and trough geometry.

Comparing the meagre status of Jakobshavn to what JC Ryan et al learned about the calving front at Store Glacier from just 3 drone overflights, it would be odd if co-author JE Box isn't sitting on drone data for Jakobshavn too:
www.the-cryosphere.net/9/1/2015/

The first image below shows a frame from the May 2014 airplane window youTube of Ruben Wernberg-Poulsen. It shows a moderately large calving event and the rocks from which 'Chasing Ice' and the June 2015 youTube were shot.

There's also quite a nice near-vertical shot of calved blue-black deeper ice which JE Box measures attributes to bubbles being compressed to air clathrate on his new blog. According to the phase diagram though, the blue ice kicks in at far too shallow a depth for von Stackelberg type II air clathrate cages (for N2 and O2). In Greenland ice cores, clathrate ice is too brittle to allow a stable iceberg upon decompression, not too mention too warm at the ocean interface. I posted a crude spectral analysis of blue-black ice after the June 2015 event.


http://jasonbox.net/what-massive-greenland-iceberg-calving-looks-like-from-the-air/

The second image is today's Sentinel. Like recent Terra/Aqua imagery, it reveals a puzzling blob adjacent to the calving front but does not provide enough detail to explain it. There's a Landsat due tomorrow but who knows if it will be cloud-free. Air Zufari only flies from mid-May to mid-Sept so there's nothing to be expected there. Note this region is not visible from the south rocks or webcam A on the north shore.
« Last Edit: October 01, 2015, 01:31:14 AM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #968 on: October 01, 2015, 01:07:59 AM »
Following up on Abrupt's tide table chart for 2015 and observational measurement in August 2004 establishing tides at Illulisat Harbor matches inconvenient calving front tides, I tabbed and loaded the DMI data into excel to see if anything interesting happened around the time of the 15 Aug 15 event.

There were some 2.6 m high tides bracketing that date but those are not so exceptional, there being 118 per year in the 2.6-2.9 range. So no smoking gun though it is hard to rule out a contributing role.

13/08/2015   3:04    0.7
13/08/2015   8:33    1.8
13/08/2015   14:21   0.5
13/08/2015   20:59   2.6
14/08/2015   3:36    0.6
14/08/2015   9:10    1.9
14/08/2015   15:00   0.4
14/08/2015   21:31   2.6
15/08/2015   4:06    0.6
15/08/2015   9:45    1.9
15/08/2015   15:36   0.5
15/08/2015   22:02   2.6
16/08/2015   4:35    0.5
16/08/2015   10:18   2.0
16/08/2015   16:10   0.5
16/08/2015   22:32   2.6
« Last Edit: October 01, 2015, 01:16:02 AM by A-Team »

AbruptSLR

  • Multi-year ice
  • Posts: 17519
    • View Profile
  • Liked: 715
  • Likes Given: 236
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #969 on: October 01, 2015, 01:23:48 AM »
So no smoking gun though it is hard to rule out a contributing role.

Consideration of a contributing role is all that I am asking for; however, a cross-correlation of the tides to all recorded calving events in 2015 (not just on August 15, 2015), would have more statistical meaning (although I am not volunteering to number crunch).
“It is not the strongest or the most intelligent who will survive but those who can best manage change.”
― Leon C. Megginson

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #970 on: October 01, 2015, 01:15:12 PM »
Quote
cross-correlation of the tides to all recorded calving events in 2015... more statistical meaning (I'm not volunteering)
I am just fascinated by our discovery this summer that the scientific community KEEPS NO RECORDS on the world's fastest glacier, draining 6.5% of the Greenland ice sheet (10% in a couple decades?). Jakobshavn is the single largest point source contributor to sea level rise. What else aren't they tracking???

Some 400-500 researchers have co-authored papers on West Greenland glaciers over the last 15 years and all we get is this dodgy graphic of yearly 'maximum' retreat? I see little value in a trend graph that glosses over the transition from ice shelf (tabular) to grounded (turn over) calving. Meanwhile Store Glacier just to north hasn't retreated since 1948 meaning retreat per se is a cherry-picking proxy.

Academics have access to all manner of slave labor up the task of monitoring EVERY AVAILABLE IMAGE-- grad students, work-study undergrads, winos standing in the day labor queue, high school science fair project teachers, distributed amateurs like astronomers use, immigrants lacking green cards and so forth.

I looked into what's involved. It's a database mining issue (satellite archive). Jakobshavn had lost its ice shelf by 2004 so we're really only talking about Landsat7 and 8 and overnight use of EarthExplorer's bulk downloader.

There's an obscure technical issue with Landsat-7's scan line corrector breaking down on 31 May 2003. However it still took perfectly usable images after that date. Even using a high-end consultant, the animation below shows these can be processed at $1.00 per image.

So, checking first the L8 OLI/TIRS and then the L7 ETM + SLC-off 2003-present box, we find that Landsat takes an image of the calving front zone when the path,row is 8,11 or 9,11 or 10,11 and returns a followup 16 days later except for ~90 days in winter.

Thus the database consists of [[(365-90)/16]*3]*11 = 567 images for the grounded calving front era. Supposing a third of these are too cloudy for calving front tracing, this still leaves 378 images, bringing the cost up to $378 if it has to be contracted out (cropping and minor geometry adjustments are easily scripted).

This amounts to almost weekly coverage over the eleven years; however Sentinel and Aqua/Terra supplementation would be necessary to really attain that during the May to September season. Here I would keep the calving front line traces in a separate Gimp layer and import to ImageJ which is set up to fit splines. (The bottom animation explores almost-daily Aqua images for June 2014 easily compiled from DMI._

The area between consecutive splines then reflects the known forward gain of the glacier minus any calving loss. The only trick here is to obtain that area using flowlines (and orthogonals) as the coordinate system for the integral. That would allow the asymmetry to be extracted later.

Having thus passed the image database off to a numeric excel calving event db, that is then served as an online tool that can be maintained, sorted, graphed etc and used by anyone in the scientific community, much as is done in 99% of the academic world (Shakespeare textual analysis, selenoprotein 3D structures, elliptic galaxy catalogues,...). DMI has already provided the tidal db and there would be others like melt day anomaly.

It seems to have been too much bother to ask Poulsen or the Air Zufari pilot for a date on the "May" calving event so, since our little db hasn't been made yet, I looked through the 2014 Landsat-8 pairs to see if it could be pinned down. However that didn't go anywhere since Air Zufari flights begin 15 May and the best cloud-free Landsat pair consists of May 9th and June 1st which further have unmatched path,row.

This highlights the mismatch between the twice-a-day tidal time series and the 23 day photo interval. Even if tides were causal rather than contributory, it would hardly show up as a correlation. Somewhere back in the 19th century someone wrote a paper explaining just how the correlation coefficient falls off as one distribution is hit with an averaging convolution. The prospects would be much improved with a daily Air Zufari video.

It's worth mentioning, even for the 100th time, that Jakobshavn is grounded. It is not floating, not even close. Not one drop of sea water has gotten under the glacier despite its bedrock being 1400 m below sea level. In fact pressurized meltwater is gushing out the other way, into the ambient sea water.

However, just like a filled reservoir behind a dam bolted into bedrock was belatedly found (1920's) to exert a buoyant lift on all that concrete, the sea water (and its tidal variance in level) does have an effect on the ice front at the calving front. At Jakobshavn, we have no real idea how that daily 40 m extension sits relative to bedrock (whose slope and till depth are unknown despite a hundred passes of ice penetrating radar): is it conforming to bedrock, cantilevered over it, buoyantly supported, melted out at depth by ocean turbulence, or hollowed out at the point of meltwater emergence on the south side asymmetry?

LC80090112014152LGN00 01-JUN-14 clear at calving front
LC80842322014150LGN00 30-MAY-14 very dark partial coverage
LC80080122014145LGN00 25-MAY-14 cloudy not useable
LC80100112014143LGN00 23-MAY-14 too cloudy
LC80090112014136LGN00 16-MAY-14 too cloudy
LC80080112014129LGN00 09-MAY-14 clear at calving front
LC80100112014127LGN00 07-MAY-14 clear at calving front
LC80070122014122LGN00 02-MAY-14 too cloudy

LC80 09 011 2014152LGN00 01-JUN-14 clear at calving front
LC80 08 011 2014129LGN00 09-MAY-14 clear at calving front
LC80 10 011 2014127LGN00 07-MAY-14 clear at calving front
« Last Edit: October 01, 2015, 03:10:51 PM by A-Team »

nukefix

  • Frazil ice
  • Posts: 461
    • View Profile
  • Liked: 9
  • Likes Given: 1
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #971 on: October 01, 2015, 03:52:07 PM »
This page has some Calving Front Location products for Jakobshaven from SAR data:

http://products.esa-icesheets-cci.org/
« Last Edit: October 02, 2015, 10:58:45 AM by nukefix »

nukefix

  • Frazil ice
  • Posts: 461
    • View Profile
  • Liked: 9
  • Likes Given: 1
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #972 on: October 02, 2015, 11:23:04 AM »
The second image is today's Sentinel. Like recent Terra/Aqua imagery, it reveals a puzzling blob adjacent to the calving front but does not provide enough detail to explain it.
Here is the S-1 IW from 24.9.2015, it looks like the "puzzling blob" is some irregular-looking huge crevasses close to the middle of the stream. Might this be the preconditioning for another rather large calving event?

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #973 on: October 02, 2015, 12:19:23 PM »
Nfix, nice spotting on the CCI Ice Sheets Products, these are a good start and fortunately for us seems to focus on Petermann and Jakobshavn. If only they could update them to 2015!

The 'surface elevation change' uses Envisat radar altimetery for dh/dt estimates over 2002–2012. They hope to add ERS-1, ERS-2, CryoSat-2 and Sentinel-3 to expand the observation period from 1991 to the present. It will be necessary to dig into their geostatistical technique (kriging/collocation) to distinguish experimental points from later interpolatio to a grid.

The initial result -- the surface of the Jakobshavn Isbræ drainage basin is dropping near the calving front but not  at higher locations -- was already known but hopefully they have a better product over a more comprehensive area and a real-time frame. Again, this has to be done to high accuracy to get at the velocity of lowering and any acceleration in that.

The 'ice velocity' product for Petermann for winter 1995/96 uses feature tracking at 35 day intervals of ERS-2 data and seems to be embedded in an icesheet-wide product. This might get interesting if it is supplemented with winter 2015/16 data and small error bars that allow differencing.

The 'calving front location' data shows the terminus of Jakobshavn between 2003-10 throughout the year according to 90 m resolution ENVISAT ASAR images. The sampling rate isn't stated but winter data would be new and welcome, though 0.1 km is large relative to most calving events. It doesn't say the calving front is read off the image or requires manual intervention.

The 'grounding line location' product for Petermann is derived manually from surface deformation, differential interferometric SAR, repeat altimetry, and from texture and shape in visible satellite images. Tidal flexure measurements would complement work going on right now with differential GPS stations on the ice.

http://products.esa-icesheets-cci.org/login/?next=/products/downloadlist/

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #974 on: October 02, 2015, 12:48:49 PM »
Quote
irregular-looking huge crevasses might indicate  preconditioning for another rather large calving event
Nfix, mighty fine IW Sentinel product you posted in #924 for 24 Sept 15. I've blown up a piece of that to native resolution below for the calving front region (2nd image) to contrast it to the miserable cloudy Landsat released yesterday (1st image, nothing to see here move along).

Landsat is not working out this fall because of clouds. We are just not getting an adequate return frequency (though Sentinel, especially the IW, is rather sparse too). If tomorrow's scene is cloudy too, there won't be another for a week; time is running out in terms of shortening daylight.

We are getting on to two months past the odd event with very little imagery documentation. I'm convinced that 30 days of Air Zufari overflights x 5 passengers with cell phone video means that good coverage exists. The question is whether anyone will post to youTube. And that's not good enough if date and search keywords aren't provided -- often annotation is really minimal.

Air Zufari / Copenhagen Air Taxi have the passenger manifests and customer contact emails. I have no idea who owns these companies (bjorn blombergs?) and whether they would be sympathetic to a request for scientific data from someone reputable like JE Box.

At this point, we aren't sure whether the 15 Aug 15 event was business not quite as usual or indicative of a whole new phase of Jakobshavn behavior. The key is this 'puzzling blob'. A second massive calving would not be consistent with BNQAU. However with winter coming on, the blob may stiffen and not calve even though it 'should' have. That would leave us in interpretive limbo.

The 3rd image shows a tentative interpretation of the Sentinel IW image. It's not likely that any two people would do this in the same way! It would help if the 03 Oct 15 Landsat is clear, though the calving front will have 'moved on' some 30 pixels northwestward.
« Last Edit: October 02, 2015, 01:28:31 PM by A-Team »

nukefix

  • Frazil ice
  • Posts: 461
    • View Profile
  • Liked: 9
  • Likes Given: 1
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #975 on: October 02, 2015, 03:51:39 PM »
Since big crevasses seem to be opening up in blob 1 is it because ice has just been pushed up the retrograde bed and the bed levels again? Just a hunch.

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #976 on: October 02, 2015, 08:34:39 PM »
Quote
up the retrograde bed and the bed levels again?
Those crevasses on the interior of blob 1 must be quite substantial ... maybe 30 m across and a km long? There is information on the strain field from their angle to the direction of flow. In the scenario of the ice flowing rapidly up and over a hump (sill), it seems there must be a trade-off between stiffness of the ice, partial suspensional buoyancy from sea water, mass conservation of volume, and possible non-attainment of relaxational equilibrium. Depending on quantitative values, this could lead to crevassing at the top of the sill as you hunched above, in effect from partial cantilevering.

Maybe sidd could dive into the bedrock DEM one more time to pull out only a few km about the current calving front so we get a better picture of the exact topography as co-located to a Landsat 15 m mercator. It's not so easy to find the actual raw bedrock depth data points as they've gotten lost in the reprocessing, kriging, re-griding and modelling. I'm recalling just a couple dozen data points per ice stream width, 5 km.  I've looked at all the radar cross-overs -- they never seem to hit bottom.

Meanwhile Terra today was not terribly informative, not boding well for tomorrow's Landsat.

AbruptSLR

  • Multi-year ice
  • Posts: 17519
    • View Profile
  • Liked: 715
  • Likes Given: 236
Re: Jakobshavn Isbræ / Sermeq Kujalleq / Ilulissat Icefjord
« Reply #977 on: October 04, 2015, 05:43:12 PM »
Do we have some people here who already have a clue on how to access and interpret seismic data?

It seems to me the Ilulisat station of the Greenland Ice Sheet Monitoring Network (GLISN) should have recorded portions of the Jakobshavn event, helping us with timing and substructure. Not sure the best way of going about this, but so far have drilled in a ways eventually getting daily Iluli seismographs for our date range:

Greenland Ice Sheet Monitoring Network
http://ds.iris.edu/ds/nodes/dmc/services/seedlink/
http://ds.iris.edu/ds/nodes/dmc/data/#requests
http://ds.iris.edu/mda/DK/ILULI
http://ds.iris.edu/ds/nodes/dmc/tools/stationmonitor/DK/ILULI/

Quote
IRIS established the Greenland Ice Sheet Monitoring Network (GLISN) of seismic and geodetic sensors to observe the dynamic behavior of the Greenland Ice Sheet as it interacts with the atmosphere, oceans, and solid Earth. GLISN offers the unique ability to record, understand, and respond to changes in the size and scope of the Greenland Ice Sheet’s behavior as it reacts to climate change. Glacial-earthquakes are increasing in occurrence, and their spatiotemporal patterns continue to evolve. Rates of crustal seismicity are also expected to grow as ice melts and withdraws, but the background rates are poorly known.

Models of the Greenland Ice Sheet are hampered by limited knowledge of its basal conditions including the geothermal heat flux, and characterization and interpretation of mass-loss signals observed using GRACE and GPS data are complicated by limited knowledge of the viscoelastic properties of the crust and mantle underlying Greenland.

Seismic imaging techniques using observations of the propagating seismic wavefield can characterize the subsurface structure, permitting inference of heat flux and the space- and time-varying response to surface loading and unloading. Similarly, analyses of seismic source signals due to ice and water motion, rock fracturing, and ocean loading provide constraints on processes including iceberg calving, deformation within the ice mass and at the bed, changes in the bedrock stress state, and variability in sea state and sea-ice cover. Such observations provide key input for the effort to understand the effects of surface melt and ice-ocean interactions on ice-sheet and glacier dynamics.

All data and metadata are freely and openly available through the IRIS Data Management Center. GLISN allows the Arctic science community full access to year-round recordings of seismic and geodetic signals produced by the Greenland Ice Sheet.
Here is what we are looking for (legend for first image):
Quote
Spectrogram (bottom) and waveforms (top) of of 14 s of seismic data showing one intermediate depth (marked with red box) and three shallow icequakes (Röösli et al., 2012). Spectrum of the intermediate depth icequake is characterized by high frequency content in the 20-80 Hz range; shallow icequakes are characterized by much lower frequency content in the 10-20 Hz range. Waveforms at the top are bandpass filtered in the 1-80 Hz range.
Here is one of a dozen or so seismic studies in West Greenland:
Quote
Glacier, fjord, and seismic response to recent large calving events, Jakobshavn Isbræ, Greenland
JM Amundson et al
http://onlinelibrary.wiley.com/doi/10.1029/2008GL035281/full
The recent loss of Jakobshavn Isbræ's extensive floating ice tongue has been accompanied by a change in near terminus behavior. Calving currently occurs primarily in summer from a grounded terminus, involves the detachment and overturning of several icebergs within 30–60 min, and produces long-lasting and far-reaching ocean waves and seismic signals, including “glacial earthquakes”. Calving also increases near-terminus glacier velocities by ∼3% but does not cause episodic rapid glacier slip, thereby contradicting the originally proposed glacial earthquake mechanism. We propose that the earthquakes are instead caused by icebergs scraping the fjord bottom during calving. The recent loss of Jakobshavn Isbræ's extensive floating ice tongue produces long-lasting and far-reaching ocean waves and seismic signals

I do not know when the Murray et al. (2015) procedure for seismically identifying major calving events for outlet glaciers will be available for general use; but its application to Jakobshavn would allow for clear identification of both the date and approximate size of recent past (in the seismically instrumented period) and current major calving events:

T. Murray, M. Nettles, N. Selmes, L. M. Cathles, J. C. Burton, T. D. James, S. Edwards, I. Martin, T. O’Farrell, R. Aspey, I. Rutt and T. Baugé (17 July 2015, Published Online June 25 2015), " Reverse glacier motion during iceberg calving and the cause of glacial earthquakes", Science; Vol. 349, no. 6245, pp. 305-308, DOI: 10.1126/science.aab0460


http://www.sciencemag.org/content/349/6245/305.abstract?sid=76ef658e-614c-409d-848f-b8e53f184c97


Abstract: "Nearly half of Greenland’s mass loss occurs through iceberg calving, but the physical mechanisms operating during calving are poorly known and in situ observations are sparse. We show that calving at Greenland’s Helheim Glacier causes a minutes-long reversal of the glacier’s horizontal flow and a downward deflection of its terminus. The reverse motion results from the horizontal force caused by iceberg capsize and acceleration away from the glacier front. The downward motion results from a hydrodynamic pressure drop behind the capsizing berg, which also causes an upward force on the solid Earth. These forces are the source of glacial earthquakes, globally detectable seismic events whose proper interpretation will allow remote sensing of calving processes occurring at increasing numbers of outlet glaciers in Greenland and Antarctica."

See also:
http://www.latimes.com/science/sciencenow/la-sci-sn-glacier-earthquake-iceberg-greenland-ice-20150625-story.html
http://www.npr.org/sections/thetwo-way/2015/06/25/417457888/study-reveals-what-happens-during-a-glacial-earthquake
http://phys.org/news/2015-08-glacial-earthquakes-sea-level.html
“It is not the strongest or the most intelligent who will survive but those who can best manage change.”
― Leon C. Megginson

Espen

  • Young ice
  • Posts: 3251
    • View Profile
  • Liked: 177
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #978 on: October 04, 2015, 07:46:36 PM »
Not much to report, expansion underway (southern branch):
Have a ice day!

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #979 on: October 05, 2015, 03:22:07 PM »
The image below shows how pre-existing icebergs were displaced down-fjord by the huge mid-August event. Click to enlarge.

The background Landsat shows day 228 16 Aug 15. I've colored the icebergs that are recognizable in the day 226 !4 Aug 15 scene and drawn an dotted yellow boundary line that approximates the boundary of 'old' ice. That ice is shown as an overlay in the same colors where it was on the 14th (close to the calving front, green line).

A black line shows 13.6 km displacement for one of the large icebergs (purple). The reason that number is so much larger than the distance of calving line retreat (0.89 km) is because calved pieces are tall but narrow -- 100 m of ice calved off the front wouldbe almost 1.5 km thick and end up stabilizing on its side after calving, contributing that much to horizontal displacement of fjord ice (rather than the 100 m).

Because Jakobshavn remains too cloudy for imaging in the visible (2nd image) but icebergs remain recognizable on Sentinel (especially in IW mode), it is possible to track subsequent calving by monitoring displacement of bergs first visible on August 16th. Here forward motion of the glacier not involving calving must be subtracted off to get calving displacement.

It appears that very little subsequent calving has occurred during the last 6 weeks -- the glacier has simply translated  forward, retaining contours and features of the calving front during that time. That's clear from the animation -- note that the five small bergs very near the calving front remain trackable and are not displaced by new calvings. The last frame however shows considerable product coming in from the northwest side.
« Last Edit: October 05, 2015, 04:53:04 PM by A-Team »

sidd

  • Young ice
  • Posts: 4939
    • View Profile
  • Liked: 345
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #980 on: October 06, 2015, 07:15:12 AM »
"Maybe sidd could dive into the bedrock DEM one more time ..."

would love to, but i am on the road for this month, with little computational horsepower ...

sidd

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #981 on: October 06, 2015, 04:22:50 PM »
Quote
"Maybe sidd could dive into the bedrock DEM one more time ..." but i am on the road for this month
I wonder if they even retained the raw data points or provided them in supplemental or elsewhere online, as is required in  science in order to make a paper reproducible.

(Very few glaciology papers fall into this category. Two months ago I contacted a lead author in disbelief that they had enough ice penetrating radar data to actually locate a pinning point under an ice shelf in northwestern Greenland. I got back an honest enough answer: "No. We just made up the data for the figure because it made sense to us to have a pinning point there." Ok, but could that not have been mentioned in the paper?)

The other big issue is the error bars on the raw data. That could well have a systematic as well as random component. If there is say ±100 m error on a 1400 m bedrock DEM, that makes sills, troughs, and contour lines a bit problematic. I would expect much better accuracy up-glacier at shallower depths, but much worse at the heavily crevassed calving front.

The Kansas radar group has recently stopped calling this 'depth to bedrock' but rather 'ice thickness'. I suppose this is because of that AGU2015 abstract that measures depth of till using an old insufficiently processed seismic line.

Meanwhile, it would be great if Jakobshavn could do its final major calving so we could do a wrap for the melt year. I looked to see what the Sentinel schedule will be for the winter, trusting DMI's snapshots to get the schedule of coverage (since the ESA image browser is so funky).

While Sentinel's orbital repeat is steady at 12 days, coverage of the calving front is very erratic (going by DMI). It appears the repeat is not exact so quite often the scene containing Disko Bay does not get over far enough east to pick up the calving front.

It looks like there will be a Sentinel with calving front coverage later today as well on the 11th, 13th and 18th. This will be the second winter of coverage. The first Sentinel of the calving front (at DMI) is from 04 Oct 14. Did anyone ever post on 2014/15 winter calving? It's not so easy to see what is going on (mid-winter shot below). Possibly the instrument was still in the process of being commissioned.

Both Jakobshavn and Petermann are snowed in today according to Aqua/Terra. Although the calving front is still quite visible because of low sun angle shadowing, it is not clear how much detail will be left on Landsat and Sentinel.
« Last Edit: October 07, 2015, 10:35:13 AM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #982 on: October 07, 2015, 12:25:56 PM »
Pulling out all the Sentinel-1 coverage of Jakobshavn at the ESA image browser back to Dec 2014 and stripping those records into an excel-type database, I found 78 scenes of 4 types.

06   EW   GRDH   1SDH    2015  09 30   20 m by 40 m resolution
40   EW   GRDM   1SDH    2015  09 30   20 m by 40 m resolution
16   IW   GRDH   1SSH    2015  09 24   05 m by 20 m resolution
16   IW   SLC    1SSH    2015  09 24   05 m by 20 m resolution

The most useful to us are the 16 IW GRDH which nukefix has been reprocessing for us (see post #974)) to 15 m resolution to co-register with Landsat band 8 mercator from EarthExplorer. The file name tells the story: S1A_IW_GRDH_1SSH_20150924_cal_EC_todB_15m_UTM22_Sigma0_HH_db.jpg

Unfortunately the satellite has rarely operated in IW mode over Jakobshavn. There was a long hiatus between late March and September this year, ie no IW coverage of the melt season and main calving events. I don't know what the plans are for fall or winter. Additionally ESA seems to have lost the 08 Feb 15 image.

By great good fortune, we have a Sentinel IW and a mildly cloudy Landsat band 8 for the same day, Sept 24th (animation below). This allows us to better interpret both in terms of calving front and iceberg appearance. The co-registration is not perfect enough however to combine them into an RGB color image (blue taken as blank).

2015 09 24*
2015 09 12
2015 03 28
2015 03 24
2015 03 16
2015 03 12
2015 030 4
2015 02 28
2015 02 20
2015 02 16
2015 02 08 SLC only
2015 02 04
2015 01 15
2015 01 11
2014 12 22
2014 12 18

The next most useful to us are the less abundant EW GRDH which are ordinary tiff file pairs for the two polarizations. As commonly posted here (eg #980), these are rotated by 22º CCW relative to Landsat and otherwise distorted but ok for the immediate vicinity of the calving front. ESA only began taking these high resolution scenes in September:

2015 09 30
2015 09 25
2015 09 20
2015 09 18
2015 09 13
2015 09 06

Sentinel data is classified by type (GRDH, GRDM, SLC) and by acquisition mode (EW Extra-Wide swath and IW Interferometric Wide swath). All the Greenland data is Level-1. We are not concerned with SSH (sea surface height nor SWH significant wave height.

GRD (Ground Range Detected, H high and M medium resolution) consists of ‘focused SAR data that has been detected, multi-looked and projected to ground range using an Earth ellipsoid model. Phase information is lost. The resulting product has approximately square resolution pixels and square pixel spacing with reduced speckle at the cost of reduced geometric resolution.’

SLC (Single Look Complex) consists of ‘focused SAR data geo-referenced using orbit and attitude data from the satellite and provided in zero-Doppler slant-range geometry. The products include a single look in each dimension using the full TX signal bandwidth and consist of complex samples preserving the phase information.’ So far we have not done anything with the phase information.

Here is the full set of Sentinel-1 images of Jakobshavn up today in csv semi-colon format. Only the variable section of the url is shown; to make a valid download link, it needs to be inserted within https://scihub.esa.int/dhus/odata/v1/Products(‘#####')/$value

EW,GRDH,1SDH,2015 0930,5bfee856-b884-494b-9f44-afbd0ec221cc;EW,GRDH,1SDH,2015 0925,6ea7e5a6-4690-470f-8eb7-583b3e940020;EW,GRDH,1SDH,2015 0920,42cb4e91-f9d9-4c95-9f74-c52298421ef9;EW,GRDH,1SDH,2015 0918,6de0dc0d-7072-4ac6-9662-10626f989e71;EW,GRDH,1SDH,2015 0913,d6c1795d-a17e-405f-ba30-9200f90aea88;EW,GRDH,1SDH,2015 0906,573d2c63-48e0-4174-bbe2-be07bd791e3b;EW,GRDM,1SDH,2015 0930,b7b1dff6-8a01-4ead-ae45-56fc7c0ec14a;EW,GRDM,1SDH,2015 0925,96620160-f74b-4470-9704-c60854e60ca1;EW,GRDM,1SDH,2015 0920,646214be-b68b-4124-b989-fba7538e3e94;EW,GRDM,1SDH,2015 0918,530fb556-0a24-47a6-abda-7fa921f5e3ee;EW,GRDM,1SDH,2015 0913,fee2f106-cb68-4c11-a7de-9d0fc4459bb9;EW,GRDM,1SDH,2015 0906,77b2e367-f13f-40a4-84ac-aa876ef6d514;EW,GRDM,1SDH,2015 0901,fd5a0457-602b-4a6a-ad37-bece669790d3;EW,GRDM,1SDH,2015 0331,1ad306cc-f883-4557-a533-0fc0dee6a722;EW,GRDM,1SDH,2015 0330,981fcfe8-088b-43b1-a478-ee2e54524020;EW,GRDM,1SDH,2015 0329,84511901-ae50-4ed6-b516-b4e6a43e0f12;EW,GRDM,1SDH,2015 0323,73c6a025-0a04-4426-a8a1-1e060806f8f7;EW,GRDM,1SDH,2015 0322,2ee0b9d9-390a-4252-9b29-19e1b5d33ca5;EW,GRDM,1SDH,2015 0321,25f9ec5b-d005-43b6-a429-98abd2bdefd9;EW,GRDM,1SDH,2015 0319,0018a331-96f1-4470-a406-207b8637eeec;EW,GRDM,1SDH,2015 0318,c7b503f1-3687-4f1b-a69b-fd28ab8e7dcd;EW,GRDM,1SDH,2015 0317,1b2cd2e4-42ca-4795-88a0-e6af7350c9a3;EW,GRDM,1SDH,2015 0311,8966d945-8cfd-4b66-96a5-edece1f8ce9f;EW,GRDM,1SDH,2015 0310,5f996c80-aba1-4029-a8d6-fe6e32dc6c78;EW,GRDM,1SDH,2015 0309,e43c3d23-1aab-4d07-a43b-a75746c53f5c;EW,GRDM,1SDH,2015 0307,dbff3838-a400-40d0-89b1-85c5599d4b45;EW,GRDM,1SDH,2015 0306,665b59b7-5844-4244-9526-5faea51f3d1c;EW,GRDM,1SDH,2015 0305,2e0856a9-30a2-4c9a-915f-4fb3459d7985;EW,GRDM,1SDH,2015 0227,cb1f7fb4-482e-4b35-9ff1-8f568b6556c2;EW,GRDM,1SDH,2015 0226,d015a427-83ed-4e1c-a8b3-e1b13bc4567b;EW,GRDM,1SDH,2015 0225,f5808940-850b-41ef-aa07-400067cd9f3c;EW,GRDM,1SDH,2015 0223,3cf56f2a-0208-4c9c-9a9c-3e6631ef8661;EW,GRDM,1SDH,2015 0222,f6067ec7-4a5f-423b-8d4f-2b8bfd6a0814;EW,GRDM,1SDH,2015 0221,f1defd03-05d6-40bc-8f53-4c7a8af12ff2;EW,GRDM,1SDH,2015 0203,21d75733-2983-40e1-a708-5b9bc21c3dd5;EW,GRDM,1SDH,2015 0116,6379aeaf-a803-43f5-b0de-6169e432d721;EW,GRDM,1SDH,2015 0110,396319ee-7ff9-400e-8710-990977cf9464;EW,GRDM,1SDH,2015 0109,dbd2e956-e158-4570-b1fb-7b897392aec5;EW,GRDM,1SDH,2015 0108,1bb5e3a9-466f-4633-b41a-22e0d56c4be9;EW,GRDM,1SDH,2015 0106,66e83b1e-9e18-4123-8f31-f1bb00241c6f;EW,GRDM,1SDH,2015 0104,b16237a1-0d23-4c80-b65c-f63824eb302f;EW,GRDM,1SDH,2014 1224,eb7f0bb5-8509-4685-aac1-55fdf59451d9;EW,GRDM,1SDH,2014 1217,dda47f62-cfab-433b-b400-47ac2586c282;EW,GRDM,1SDH,2014 1216,e3cea07f-8d3b-4b17-afc1-0e352ae60e1c;EW,GRDM,1SDH,2014 1215,3efb8d34-e272-4bbd-a80a-919f7f5d66e4;EW,GRDM,1SDH,2014 1213,d8c6db44-d123-43fa-9201-4550fbdb48ff;IW,GRDH,1SSH,2015 0924,e7eeb282-4d63-4bfe-bfce-b82ce8e0ef5d;IW,GRDH,1SSH,2015 0912,6a0b02c9-50a2-47b2-830b-d0707e5d5cac;IW,GRDH,1SSH,2015 0328,2f06cd65-bb60-49b4-a932-45d66f4fa429;IW,GRDH,1SSH,2015 0324,85630776-a95d-4707-b4bf-805ce7bdde6c;IW,GRDH,1SSH,2015 0316,5b51d113-21f0-4bbc-887c-0b72b9a669f7;IW,GRDH,1SSH,2015 0312,ce1e063a-a237-43f7-8e85-c31f4e487f11;IW,GRDH,1SSH,2015 0304,50064c58-6ad8-4510-93d6-bae0b432e4ef;IW,GRDH,1SSH,2015 0228,f7092300-c089-4c5a-83de-f15a7b6b2d55;IW,GRDH,1SSH,2015 0220,6b184e0e-b570-48f0-ad91-7b8f0824957c;IW,GRDH,1SSH,2015 0216,9a555ae1-7542-4099-abb6-8fd2281f8d2e;IW,GRDH,1SSH,2015 0208,xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx;IW,GRDH,1SSH,2015 0204,346c7590-e9ef-4c65-a0cf-5f096a3be84b;IW,GRDH,1SSH,2015 0115,066dced6-1044-46d0-8296-1b44c7fa4be2;IW,GRDH,1SSH,2015 0111,f14e54d4-e0bf-4011-bd97-b278f3d700df;IW,GRDH,1SSH,2014 1222,0df1c9bc-ee95-49b1-8118-e8161651ff5f;IW,GRDH,1SSH,2014 1218,9598b21f-f46f-4d93-a8df-62022c979977;IW,SLC,1SSH,2015 0924,92507aaa-03d9-462d-8425-61a6229e0940;IW,SLC,1SSH,2015 0912,8b156bb8-8893-4b50-be40-a36c8dbc1161;IW,SLC,1SSH,2015 0328,5bc4acdc-81c8-4553-b003-dc0079865276;IW,SLC,1SSH,2015 0324,85590167-8bb4-4dbe-ab87-3fd09d19ed58;IW,SLC,1SSH,2015 0316,1af97347-45f8-4845-9650-c6fc952e2b54;IW,SLC,1SSH,2015 0312,ad5a8ff7-c789-484f-b7be-282bd23115eb;IW,SLC,1SSH,2015 0304,1744eb63-ff9b-4466-a6fd-7385d69bab6d;IW,SLC,1SSH,2015 0228,e666dfc0-d3d9-4c65-ba66-b55867e4c2d2;IW,SLC,1SSH,2015 0220,2231e74d-5984-4da4-a33c-ae72699cecda;IW,SLC,1SSH,2015 0216,2d192f45-cb99-423e-96ab-da48dc0f3535;IW,SLC,1SSH,2015 0208,0a008aff-cf1e-4c9b-9fae-9787bf6baf42;IW,SLC,1SSH,2015 0204,18d545fe-081b-4471-8cac-64934edf72c1;IW,SLC,1SSH,2015 0115,674f2120-d4f1-42b2-bb1c-5606b442cdea;IW,SLC,1SSH,2015 0111,fa8bbc6e-15dd-4c84-bc5b-7afb5f0456d2;IW,SLC,1SSH,2014 1222,f91d3a10-b2cd-4ed5-a307-08c888de4965;IW,SLC,1SSH,2014 1218,a348bd03-c1d2-470f-9b63-8ed22d2741e8
« Last Edit: October 07, 2015, 12:50:54 PM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #983 on: October 09, 2015, 03:15:46 PM »
Some unsatisfactory imagery for early October has drifted in. There has been a substantial calving event and more seems to be in the works. However with Landsat no longer making certain scene acquistions. Modis having its contrast snowed out, and the Sentinel user interface mired in a 2013 prototype and unable to call up images that we know exist from DMI previews, there is little that can be done by way of documentation unless we get a clear Landsat this coming Sunday.

nukefix

  • Frazil ice
  • Posts: 461
    • View Profile
  • Liked: 9
  • Likes Given: 1
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #984 on: October 09, 2015, 03:55:54 PM »
The shadow of the calving front is enormous in the S1 image...I'm waiting for the IW from 7.10 to appear on scihub. Is the calving front perhaps close to the highest point on the fjordbed as judging from the shadow it looks like the calving front is higher than in other images?
« Last Edit: October 09, 2015, 05:04:49 PM by nukefix »

Espen

  • Young ice
  • Posts: 3251
    • View Profile
  • Liked: 177
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #985 on: October 09, 2015, 05:26:09 PM »
Some unsatisfactory imagery for early October has drifted in. There has been a substantial calving event and more seems to be in the works. However with Landsat no longer making certain scene acquistions. Modis having its contrast snowed out, and the Sentinel user interface mired in a 2013 prototype and unable to call up images that we know exist from DMI previews, there is little that can be done by way of documentation unless we get a clear Landsat this coming Sunday.

Confirmed! A rather large on, not down to Aug. 15 2015 though, but "what seems to be in the works" is almost where calving front was on Aug. 15!
« Last Edit: October 09, 2015, 06:15:24 PM by Espen »
Have a ice day!

Espen

  • Young ice
  • Posts: 3251
    • View Profile
  • Liked: 177
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #986 on: October 09, 2015, 07:08:57 PM »
October 7 vs. August 20 2015:
Have a ice day!

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #987 on: October 10, 2015, 01:28:53 AM »
Quote
The shadow of the calving front is enormous ... Is the calving front close to  highest point on fjord bed -- as judging from the shadow it looks like the calving front is higher than in other images?
Good question. How high is this calving front cliff, how does it vary across the south branch, does the elevation profile change with calving front position, can depth-to-bedrock be inferred from this? There is said to be a stability cutoff around 100 m.

The Terra image from DMI has its share of puzzling shadows as well (2nd image). There are consistently present and not attributable to clouds etc.

Meanwhile, the ice stream is presumably thinning near the front (dropping in masl) which would offset gain from going over a sill. Maybe we can get our hands on summer 2015 laser altimetry (ICESat and IceBridge). Or figure vertical change ourselves from IW pairs and Sentinel toolbox (or do we need Sentinel-1B for that)?

I did locate this 2m hillshaded DEM of the calving front made from 0.5 m WorldView. After som ISP problems I was able to download the large file. As a 'tiled tif' it cannot be opened by ImageJ or Gimp though a plugin of ImageJ called BioFormats opened it without delay It may need the tile just to the west too. North branch of interest as well (smaller boxed area).

Somewhere Ian Howat has posted the DEMs without the hillshading. That is another exercise, to download that and make elevation profiles corresponding to ice stream flowline trajectories up to the calving front and just beyond. The dates would not be late-summer 2015 by any means ... we would need access to WorldView-3 for that.

SETSM_ArcticDEM_18_39_2_1.tar.gz   06-Mar-2015 19:27    290M
http://repository.agic.umn.edu/imagery/stereoDEM/SETSM/ArcticDEM/18_39/2m/SETSM_ArcticDEM_18_39_2_1.tar.gz
« Last Edit: October 11, 2015, 04:56:32 PM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #988 on: October 10, 2015, 07:30:06 PM »
I took another look at the close of the Landsat season of 2014. About half of them were too cloudy but in the end 5 scenes were available in Sep-Oct-Nov. The very last scene acquired, the Nov 1st, was really excellent for revealing landscape shadows attributable to slight topographic variation, the sun elevation being 6.0º above the horizon and due south.

This was not ideal however for measuring height across the calving front which was running more or less north-south on that date. I'll take a look at some late summer dates in 2015 with more northeasterly calving fronts (but less ideal sun angles) in a bit.

September 28 remains our date of maximal retreat but because of clouds and the mediocrity of monitoring, an earlier date and greater extent remain possible.

Espen

  • Young ice
  • Posts: 3251
    • View Profile
  • Liked: 177
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #989 on: October 10, 2015, 10:02:03 PM »
A new large calving is underway, but still far "behind" August 16 2015:

Click on image to enlarge and animate!
Have a ice day!

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #990 on: October 10, 2015, 11:18:50 PM »
Whoa, bizarre in so many ways. Be sure to notify the Big Shots immediately so they can take proper credit for your discovery. (I blurred the 'Espen Olsen' bottom right to help with this and will prepare some interpretation so they can put some substance in their quotes this time.)

Just upglacier of that large dark streak that we have been looking at for a while on Sentinel and Aqua, to me it looks like Jakobshavn has re-entered a tabular calving mode, which I would call quasi-tabular. Jakobshavn is calving but the calved piece is not rolling back onto its side like we are accustomed too.

This might be attributed to a newly retrograde slope underneath causing the calved piece to stably lie back against the main ice stream. Alternatively the calved piece might be wider than deep, causing the present tabular configuration to be in equilibrium without tilting over. This could happen if it broke off only part way to bedrock, allowing it to rest on the remaining piece, or if bedrock is not so deep at the top of a sill. It's not plausible that frozen melange could be buttressing -- too much open water down the fjord.

I've processed the LC80090112015283LGN00 Band 8; it is the 7.5 m below (2nd image). This is probably worth the naturalistic pansharpening coloring method the NASA fellow shared with us to see if there is liquid water in the crevasses. We have some other recent path 9, row 11's for comparative purposes and a Sentinel IW coming in shortly. While the Oct 10th has some cloud issues on the sides, this is not enough to introduce alignment uncertainty.

These are definitely sun shadows of an ice cliff above sea level melange: it is just barely possible to see through them to ice underneath. The sun is at 13.7º and due south so we will be able to work out the cliff height profile from simple geometry. Some of this is just seracs which shouldn't really 'count' towards representative cliff height.

This Landsat scene may have caught Jakobshavn in the middle of a long-running event. There is clearly more to come. Major calving this late in season -- that would be a new development relative to 2014 (two posts back). I'll put together a 2013 series shortly as that is still Landsat-8.

The third image emphasizes the ice lying in the shadows. The reddish region seems like an earlier and smaller version of the quasi-tabular calving concept, the first forum comments on it were several weeks ago. It still has not broken free and continues to be pushed ahead by the main ice stream motion.

The 'natural' color does a nice job of distinguishing calved from uncalved regions (4th image, bands 543 as RGB). The dotted boundary was also apparent as a preconditioned calving anomaly in earlier Sentinel images (see above). Some thin clouds can also be seen -- we were very lucky here because the overall scene is dominated by a impenetrable cloud layer and the next Landsat is not due until the 17th.
« Last Edit: October 11, 2015, 12:54:00 AM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #991 on: October 11, 2015, 04:42:25 AM »
I took a look at the shadow heights cast by the calving front on the floating melange of the fjord (taken as sea level). Acccording to the file metadata going with the 10 Oct 15 Landsat, the sun is 13.7º above the horizon and due south (ie Landsat scenes are acquired at local noon).

Thus a 100 m flagpole would cast such and such a length shadow according to basic trigonometry associated with this sun angle. This length would correspond to so and so many pixels at the resolution considered.

Looking at raw band 8 (no contrast adjustment) at 5x bicubic interpolation, a mask delineating the shadowed region can be drawn, which serves to restrict a vertical grid (yellow) which samples the calving front shadow heights every 10 pixels to give 58 positions along the calving front from which the height above sea level follows when the shadows are lifted to an even footing (graph).

On the technical side, the vertical grid only works because of the mercator projection of Landsat and a solar azimuth of 180º. The shadowed region can be picked objectively in grayscale space because its tone is darker. The grid bars resulting from truncation by the mask can be lifted in register because various obscure commands are provided in gimp.

This allows the whole process to stay within the pixelated raster world of the original Landsat to minimize measurement error. I'll get to that after adding the height scale. seracs, roughness, flotation and so forth tomorrow. For now, just note that the relative height of the calving front has a broad central peak and falls off to the sides, consistent with airplane photos. (These can be rectified with some effort but usually lack an embedded scale feature.)

There is an optical illusion in the image below by which the vertical bars in the graph look longer than their counterparts overlaid on the image. They are not; the bars have identical heights.
« Last Edit: October 11, 2015, 02:37:42 PM by A-Team »

Espen

  • Young ice
  • Posts: 3251
    • View Profile
  • Liked: 177
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #992 on: October 11, 2015, 10:17:35 AM »
A-Team. Thanks for your good sense of humor.

But I doubt it will end up tabular, but you are right if it does that's a major change.

And the image from October 10 was caught in the middle of the act, so to speak!

Seen in the October 10, it looks like the ice just behind in the  next calving in the lower section of the southern branch is somehow lifted (red encircled).
« Last Edit: October 11, 2015, 10:25:46 AM by Espen »
Have a ice day!

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #993 on: October 11, 2015, 12:00:38 PM »
Quote
it looks like the ice just behind in the next calving in the lower section of the southern branch is somehow lifted
If so, the lift will not be a record. Even though nobody keeps records and nothing like a lift has ever been observed before. Better we wait until Monday afternoon to read what the Big Shots say in the newspaper?

Alternatively, I am thinking the ice between the two shadowed crevasses has tilted back. That would mean the calving front per se is too high and its shadows will prove too long (post #993). It's too low in the back so the second crevasse appears too deep as well.

Note that the shadows seen in Sentinel have a different physical origin than those of Landsat. The latter takes a nadir view of a landscape illuminated by back-reflection from a low sun. The former looks obliquely, illuminating the scene with its own microwaves; shadows are regions physically blocked from incident radar look angle by ground topography. Here we see both types of shadows and both are anomalously large.

The upglacial ice stream is too heavy to be lifted and that would fracture it, not seen. In the rotating block scenario, it amounts to only a partial calving where the block has not completed its feet-first outward rotation as we see in youTubes (see block animation below).

Note that pink-tinted block above has been sitting there since the mid-August event, just moving passively forward with the ice stream despite being only partially attached.

However, as noted a couple of days back on the Oct 7th Sentinel, there has been substantial ice calved off the rest of the front. This is very clear from the trajectories of the large icebergs, some of which have quite distinctive shapes. One of these is now halfway down the fjord and new icebergs are visible in the bay by the calving front.
« Last Edit: October 13, 2015, 01:23:11 AM by A-Team »

Shared Humanity

  • Young ice
  • Posts: 3964
    • View Profile
  • Liked: 416
  • Likes Given: 48
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #994 on: October 11, 2015, 06:54:01 PM »
Espen/A-Team

I  would like to thank both  of you for creating one of the most exciting threads on the entire Forum. I come here everyday.

While I  am just an observer, I  do think that this glacier is exhibiting entirely new behaviors, on both the main calving face and on the  north wall.

Espen

  • Young ice
  • Posts: 3251
    • View Profile
  • Liked: 177
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #995 on: October 11, 2015, 07:51:49 PM »
I believe we had another (EDIT: VERY) large calving at Jakobshavn since October 10 2015 (last landsat data), sorry about the image quality :(

Large amount of glacier ice debris is seen in fjord despite the poor image quality!

And something suspicious happened in right corner of the northern branch?
« Last Edit: October 11, 2015, 08:06:00 PM by Espen »
Have a ice day!

Espen

  • Young ice
  • Posts: 3251
    • View Profile
  • Liked: 177
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #996 on: October 11, 2015, 08:17:24 PM »
To those who thought it was only a one night stand, I can only say we are back to a new normality:

EDIT: Very close size wise to the Aug. 15 / 16 2015 calving!!
« Last Edit: October 11, 2015, 08:26:38 PM by Espen »
Have a ice day!

Espen

  • Young ice
  • Posts: 3251
    • View Profile
  • Liked: 177
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #997 on: October 11, 2015, 08:31:03 PM »
Very strange season yesterday Nunatakassap Sermia http://forum.arctic-sea-ice.net/index.php/topic,1388.msg64601.html#msg64601

and today/yesterday Jakobshavn, a real new normality!

2 record calvings in Jakobshavn in one season!!
Have a ice day!

Espen

  • Young ice
  • Posts: 3251
    • View Profile
  • Liked: 177
  • Likes Given: 0
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #998 on: October 11, 2015, 08:58:12 PM »
Here is an updated version (better image quality):

New large ice bergs are clearly seen in the fjord!
Have a ice day!

A-Team

  • Nilas ice
  • Posts: 2401
    • View Profile
  • Liked: 282
  • Likes Given: 19
Re: Jakobshavn Isbræ / Ilulissat Isfjord / West Greenland
« Reply #999 on: October 11, 2015, 09:13:07 PM »
Quote
Thank Espen/A-Team. I  do think that this glacier is exhibiting entirely new behaviors, on both the main calving face and on the  north wall.
This glacier has indeed been taking a walk on the wild side. For the record, I don't make this stuff up (can't speak for Espen).

The north side of the south channel has followed a conventional formula to date: fairly steep descent into a fjord + unblocked by retreat of south branch = rapid speedup of calving. Or has it? The south branch retreated nearly as far in 2013 and 2014 but I'm not recalling so much disintegration at so late a date.

On the main channel, the handful of futuristic publications envisioned more speeding up (but with physical limitations to that). That scenario is off the table if the glacier is taking up ominous new behaviors.

This last six weeks has really changed my thinking about discrete 'events'. Instead of creep creep creep dramatic calving creep creep creep big calving creep creep creep, it seems better not to set boundaries that define events in isolation. Each stage of the glacier now seems to imply the next, here a long sequence drawn out over six weeks and counting.

I foresee more of the same over the next two weeks. However with colder weather and the end of meltwater come slower stiffer ice and perhaps a pause -- but not an end -- to the new behavior

The Terra image for the 11th has cloud and resolution issues. The processing also seems to have changed, affecting the blues (normally Disco Bay is blue-black). In support of Espen's read of the better Aqua, the jet being flushed out the fjord mouth is quite pronounced and even discolored as though ice with till had been calved.