Support the Arctic Sea Ice Forum and Blog

Author Topic: MOSAiC news  (Read 36093 times)

Tor Bejnar

  • Young ice
  • Posts: 3115
    • View Profile
  • Liked: 409
  • Likes Given: 200
Re: MOSAiC news
« Reply #250 on: November 15, 2019, 07:54:53 PM »
This is a big improvement over the happy face absurdities from the PR dept back on shore. I am not familiar with the term Presseisrücken, perhaps what we are calling ice dynamics induced by pressure from ice movement but maybe it is a technical term for pressure ridges.
...
An internet search turns up (new to me website) Linguee.com  Looks like "Presseisrücken" means "pressure ridges".
Screen print:

Arctic ice is healthy for children and other living things.

Stephan

  • Grease ice
  • Posts: 804
    • View Profile
  • Liked: 273
  • Likes Given: 141
Re: MOSAiC news
« Reply #251 on: November 15, 2019, 08:26:39 PM »
As a native German speaker I can confirm this translation.
It is too late just to be concerned about Climate Change

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #252 on: November 15, 2019, 09:58:32 PM »
Mosaic Pbuoys up to nov15. Rough Polarstern position from follow mosaic today in red
edit: smaller buoy dots, shorter tails
edit2: bigdots longer tails
« Last Edit: November 16, 2019, 01:20:51 AM by uniquorn »

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #253 on: November 16, 2019, 12:21:00 AM »
Quote
Rough Polarstern position from follow-mosaic today in red
Follow_Mosaic seems not to known where the Polarstern has been nor where it is now nor where it is going next. The ship about to hit some serious divergence in the wind vector field which will affect the ice similarly.

Like! (like bigger dots, longer tails)

Not sure what you mean by Pbuoys. These?

300234062881930*   SVP.07   2019P16    07.10.19   84.98   134.50   SVP   UPTEMPO   AWI   11.10.19
300234062887920*   SVP.08   2019P22    07.10.19   84.76   132.21   SVP   SVPB-AWI   17.10.19
300234066080220*   SVP.09   2019P122   10.10.19   85.74   128.71   SVP   SVP   EU   21.10.19
300234066081220*   SVP.10   2019P123   07.10.19   83.83   133.05   SVP   SVP   EU   21.10.19
300234066082230*   SVP.11   2019P127   07.10.19   85.90   115.25   SVP   SVP   EU   21.10.19
300234066084230*   SVP.12   2019P126   11.10.19   85.32   125.17   SVP   SVP   EU   21.10.19
300234066085220*   SVP.13   2019P128   07.10.19   86.21   110.64   SVP   SVP   EU   21.10.19
300234066087220*   SVP.14   2019P125   11.10.19   85.61   133.38   SVP   SVP   EU   21.10.19
300234066089220*   SVP.15   2019P124   29.09.19   85.78   129.60   SVP   SVP   EU   21.10.19
300234067509680*   SVP.16   2019P88    07.10.19   85.18   132.33   SVP   IT-B   AWI   22.10.19
300234067608220*   SVP.17   2019P90    07.10.19   85.64   132.39   SVP   IT-B   AWI   22.10.19
300234067700680*   SVP.18   2019P91    07.10.19   85.40   133.78   SVP   IT-B   AWI   22.10.19
300234067700760*   SVP.19   2019P92    07.10.19   85.18   126.43   SVP   IT-B   AWI   22.10.19
300234068121990*   SVP.20   2019P188   07.10.19   84.65   134.18   SVP   IT   OSU   17.10.19
300234068123990*   SVP.21   2019P191   05.10.19   84.60   129.99   SVP   IT   OSU   17.10.19
300234068125990*   SVP.22   2019P193   05.10.19   84.71   134.16   SVP   IT   OSU   17.10.19
300234068129990*   SVP.23   2019P192   07.10.19   84.58   130.83   SVP   IT   OSU   17.10.19
300234068218450*   SVP.24   2019P152   29.09.19   84.15   124.63   SVP   SVP   PRIC   17.10.19
300234068220020*   SVP.25   2019P190   07.10.19   84.63   133.49   SVP   IT   OSU   17.10.19
300234068221000*   SVP.26   2019P189   05.10.19   84.64   130.85   SVP   IT   OSU   17.10.19
300234068222000*   SVP.27   2019P194   05.10.19   84.65   130.34   SVP   IT   OSU   17.10.19
300234068226020*   SVP.28   2019P196   07.10.19   84.94   132.28   SVP   IT   OSU   17.10.19
300234068312200*   SVP.29   2019P156   28.09.19   81.85   118.45   SVP   SVP   PRIC   17.10.19
300234068312210*   SVP.30   2019P157   26.09.19   81.08   116.74   SVP   SVP   PRIC   17.10.19
300234068316210*   SVP.31   2019P155   29.09.19   83.62   123.58   SVP   SVP   PRIC   17.10.19
300234068910590*   SVP.32   2019P200   07.10.19   84.94   133.37   SVP   IT   OSU   17.10.19
300234068910810*   SVP.33   2019P198   07.10.19   85.00   132.62   SVP   IT   OSU   17.10.19
300234068915790*   SVP.34   2019P203   07.10.19   85.01   133.16   SVP   IT   OSU   17.10.19
300234068917830*   SVP.35   2019P206   07.10.19   84.91   132.74   SVP   IT   OSU   17.10.19
« Last Edit: November 16, 2019, 12:31:28 AM by A-Team »

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #254 on: November 16, 2019, 01:03:35 AM »
Due to unfortunate lack of coding skills I hack macid's code for the Pbuoy animations which use the meereis csv format here
All Pbuoy data is calculated but some is out of range of the animation above.
The times vary but most are valid to 15/11 0300
The followmosaic PS location (to 2 decimal places) from an unknown time is added as a rough guide.
Hopefully they will provide a time with the location at some point.

macid

  • New ice
  • Posts: 26
    • View Profile
  • Liked: 13
  • Likes Given: 16
Re: MOSAiC news
« Reply #255 on: November 16, 2019, 02:04:43 AM »
There's a static array with the buoy names in the code that be quickly edited in case you want to remove some odd ones or add new, I tested with P & S Type buoys for csv format handling, both do drift speed I think.

I think the most accurate location of the PS you can get by putting a dot between the closest buoys..

*edit: checking the code, line 31 makes it only download P types, have to hack it to do others yep

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #256 on: November 16, 2019, 03:38:23 AM »
Quote
most accurate location of the PS you can get by putting a dot between the closest buoys
Good suggestion. Seemed like one of the buoys was actually fixed on the corner of the selected floe. Still, the floe is ~3000 m whereas 4 dp GPS is ~11 m.

Quote
The follow_mosaic PS location (to 2 decimal places) from an unknown time is added as a rough guide. Hopefully they will provide a time with the location at some point.
Questions should be directed to the AWI press officer. No contact information is provided however. Does that answer your question?

Quote
AWI press officer Sina Löschke will contribute to the Polarstern-blog from Bremerhaven, Germany. It is going to be her part to answer questions asked by readers and help whenever the expedition team needs a writing hand.

They are really moving. Purely zonal movement (east): 2.3 degree of longitude gained at fixed latitude 86.2 in the last 55 hours, with increasing gale force winds (14+ m/s), dropping pressure and rising temperatures:

  86.2  120.4 19-11-16 09:00   17  260     -6.0  1005.1
  86.2  120.3 19-11-16 08:00   17  260     -5.9  1004.9
  86.2  120.1 19-11-16 07:00   16  250     -6.3  1005.6
  86.2  120.0 19-11-16 06:00   15  240     -7.1  1006.2
  86.2  119.9 19-11-16 05:00   14  250     -7.8  1007.0
  86.2  119.7 19-11-16 04:00   12  240     -8.0  1007.3
  86.2  119.6 19-11-16 03:00   12  240     -7.3  1007.6
  86.2  119.5 19-11-16 02:00   12  250     -6.9  1008.0
  86.2  119.4 19-11-16 01:00   13  260     -6.9  1007.8
  86.2  119.3 19-11-16 00:00   13  250     -7.4  1008.2
  86.2  119.2 19-11-15 23:00   11  250     -8.1  1009.1
  86.2  119.1 19-11-15 22:00   10  250     -8.6  1009.2

« Last Edit: November 16, 2019, 12:01:27 PM by A-Team »

be cause

  • Grease ice
  • Posts: 892
  • Citizenship .. a Lurker gets asylum
    • View Profile
  • Liked: 262
  • Likes Given: 219
Re: MOSAiC news
« Reply #257 on: November 16, 2019, 11:05:26 AM »
 .. chances increase that they will get much closer to the pole .. b.c.

ps. thanks for the info flow ..
2007 + 5 = 2012 + 4 = 2016 + 3 = 2019 ...

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #258 on: November 16, 2019, 12:35:46 PM »
A third Sentinel-1AB came in this morning for yesterday Nov 15th. It shows what appears to be a white pressure ridge appearing between 04:16 and 05:54 right along the axis of the Polarstern. To the west (right) a long dark lead has opened further and shifted towards the Mosaic floe.

Other instabilities can also be seen but we await today's S1AB for confirmation and further developments over the multi-day course of the cyclone which will be massively disruptive for the Mosaic floe and the equipment deployed on it. PolarView is not processing images so far today, perhaps because it is early on the weekend.

The ship added another two tenths of a longitude degree just in the last hour from gale force winds. They originally moored at 85.1, 133.8 on Oct 5th.

  86.2  120.6 19-11-16 10:00   17  260
  86.2  120.4 19-11-16 09:00   17  260
« Last Edit: November 16, 2019, 12:53:03 PM by A-Team »

gerontocrat

  • First-year ice
  • Posts: 6824
    • View Profile
  • Liked: 1728
  • Likes Given: 22
Re: MOSAiC news
« Reply #259 on: November 16, 2019, 01:49:41 PM »
Imagine...

So you've volunteered and lobbied like mad to be on the project - what an opportunity for a young scientist!

You've got a bit of Arctic experience - in the summer's 24 hour daylight.

But today,
- 2 months of darkness is getting to you,
- it's about -7 celsius outside, + windchill from 50 kmh wind,
- the ship is creaking and groaning,
- the ice is creaking, groaning, grinding, crack! a new lead.

The boss needs you to go on the floe to rescue some gear before it heads 4,000 metres down.
The polar guards have said the bears are a bit active today - lots of holes in the ice means a better chance of getting a seal.

Life is wonderful.

Stuff that for a game of soldiers.
"Para a Causa do Povo a Luta Continua!"
"And that's all I'm going to say about that". Forrest Gump
"Damn, I wanted to see what happened next" (Epitaph)

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #260 on: November 16, 2019, 05:07:00 PM »
There's a static array with the buoy names in the code that be quickly edited in case you want to remove some odd ones or add new, I tested with P & S Type buoys for csv format handling, both do drift speed I think.
Thanks macid. I download all the meereis buoys separately and copy the ones I want to see into the work file. My biggest problem is making a ~80buoy iabp animation work. The files and dates are a different format. I think/hope I'm doing ok tinkering with the meereis csv's.

As we have some significant movement at the moment here are the  meereis Pbuoys from nov14-16 04:30, though not all buoy data may be collected up to that time.
The animation has 54 frames, which should be 1/hr though looking at the 'brownian motion' I suspect some averaging over time for the buoys that report every 10mins or lack of pixels for an accurate representation. Running a larger scale animation may determine which it is.
I've left the red PS dot in from yesterday's position for fun ;) at 86.11N 118.52
« Last Edit: November 17, 2019, 10:42:16 PM by uniquorn »

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #261 on: November 16, 2019, 08:51:28 PM »
Pbouys at 0600 yesterday overlaid onto polarvew jpg at 0554.
Click for full res

blumenkraft

  • Nilas ice
  • Posts: 1586
  • Fans of Hans Club - circa 2018
    • View Profile
  • Liked: 707
  • Likes Given: 982
Re: MOSAiC news
« Reply #262 on: November 17, 2019, 06:16:15 AM »
Not quite the right direction.
Refugees welcome

blumenkraft

  • Nilas ice
  • Posts: 1586
  • Fans of Hans Club - circa 2018
    • View Profile
  • Liked: 707
  • Likes Given: 982
Re: MOSAiC news
« Reply #263 on: November 17, 2019, 03:31:22 PM »
Looks like a crack is developing right where the Polarstern is.
Refugees welcome

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #264 on: November 17, 2019, 03:49:28 PM »
The Mosaic floe experienced only moderate ice dynamics during the record winds of Nov 16th (going by the only data of the 15th and 17th); we don't know what effect if any this had on equipment set up at the Ice Camp. Another WHOI buoy is gone; again we lack a causal attribution. Tomorrow will bring further challenging conditions with strong divergent winds.

One surprise initially was the thinness of the ice; neither satellites nor models prepared them for the fragility and thinness of the ice at the selected floe. Next, the mechanical strength of the ice (which cannot be measured remotely) was less than what they needed to plow logistics runways with the Piston Bullys -- that seems unthinkable even now.

Reporters onboard the AF noted a layer of dirty ice sloughing off the bottom of overturned floes; the trip leader spoke yesterday of porous immature ice. How should these be reported at Piomas, how do they count towards Cryosat2 freeboard or SMOS ice thinness? It doesn't work very well to idealize ice from far away because it has become quite heterogenous and very complex close up.

The third surprise has been atypical winds. There is no indication of when trans- or circumpolar winds will set up -- indeed the Fram has been running in reverse. That could change, but at this rate the Polarstern won't get very far on its expected drift. Even if it resumed, the ship won't get near the pole because an enormous impenetrable block of thick MYI ice has occupied that region for close to a year. What would move that out of the way?

It's not really clear how to compare ice dynamics year on year but leads, ridging and deformation may be worse than the expedition had anticipated, again because of seemingly deteriorating ice capacity for resisting wind stress.

I wonder how the guy wires are engineered on the two tall meteo towers. Presumably with very stout springs to take up relative ice motion, yet elasticity would leave the tower swaying in the strongest winds of the trip, 21m/s (75.6 km/hr or 47 mph) on Nov 16th. Recall early on the long crack with legs of the first tower a couple of meters away from doing the splits.

  86.1  122.2 19-11-16 20:00   16  300     -5.2
  86.1  122.1 19-11-16 19:00   16  290     -5.0   994.4
  86.1  122.0 19-11-16 18:00   19  260     -5.2   994.8
  86.1  121.8 19-11-16 17:00   19  260     -5.9   995.7
  86.2  121.6 19-11-16 16:00   20  260     -6.4   997.3
  86.2  121.5 19-11-16 15:00   21  250     -6.6   998.7
  86.2  121.3 19-11-16 14:00   20  250     -6.5  1000.2
  86.2  120.9 19-11-16 12:00   19  250     -6.7  1002.9
  86.2  120.8 19-11-16 11:00   18  250     -6.5  1003.6
  86.2  120.6 19-11-16 10:00   17  260     -6.3  1004.6
  86.2  120.4 19-11-16 09:00   17  260     -6.0  1005.1

The first animation below shows 2.6 km of en bloc motion mostly to the south in the 112 minutes between S1AB timestamps. That works out to 33.9 km/day.
« Last Edit: November 17, 2019, 08:11:46 PM by A-Team »

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #265 on: November 17, 2019, 07:03:53 PM »
update on drift speed of the meereis Pbuoys from nov14-16 ~11:30

see below
« Last Edit: November 18, 2019, 06:37:23 PM by uniquorn »

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #266 on: November 17, 2019, 08:48:15 PM »
Quote
drift speed nov14-16 ~11:30
A keeper! These are fascinating to watch. Below the whole trajectory of each buoy is shown (gimp 'stack average' + level contrast tools; often used to set minimal crop boundaries that don't miss anything).

The Polarstern will be in a very curious situation tomorrow at 15:00 utc: low wind locally but situated in an 84º turn in much higher wind stress which will result (?) in a giant lead opening along the blue band.
« Last Edit: November 18, 2019, 03:22:20 AM by A-Team »

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #267 on: November 17, 2019, 11:16:44 PM »
Update on the 4 SIMB buoys measuring ice thickness. Temperatures confirming the current warm spell.
https://www.cryosphereinnovation.com/382860

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #268 on: November 18, 2019, 06:38:46 PM »
update on drift speed of the meereis Pbuoys from nov14-18 ~03:30

Posting this for interest while I work out which is the wayward buoy in the middle.

blumenkraft

  • Nilas ice
  • Posts: 1586
  • Fans of Hans Club - circa 2018
    • View Profile
  • Liked: 707
  • Likes Given: 982
Re: MOSAiC news
« Reply #269 on: November 18, 2019, 06:49:13 PM »
That looks like a slowly failing GPS, no? First, you see a delay in updating and then a failure, pinpointing it to a location in the past.
Refugees welcome

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #270 on: November 18, 2019, 08:13:21 PM »
Possibly. Badbuoy is 2019p101_300234067705700_proc, file attached


macid

  • New ice
  • Posts: 26
    • View Profile
  • Liked: 13
  • Likes Given: 16
Re: MOSAiC news
« Reply #271 on: November 19, 2019, 03:58:54 AM »
P101 went bad a month ago, 2019-10-19 19:24:00, accelerometer variance reports a whopping 7076 mystery units for the last regular report in. Accelerometers are supposed to report x y and z values but I'll imagine they're combining the outputs into a sum of accelerations on all axes, indicating a tumbling of sorts when spiking.

Polar bear ran off with it? Fell into a widening lead? Latter seems more likely, graph attached. curious wave action going on before as well. heavy winds, massive drift may have played a role in it's demise.

After that quiet for 8 days, since then intermittent reports of ~half days, indicating poor satellite reception, which would also account for the erratic movement. GPS accuracy drops quickly when you get down to just a few sats in view.


Hefaistos

  • Frazil ice
  • Posts: 450
    • View Profile
  • Liked: 31
  • Likes Given: 131
Re: MOSAiC news
« Reply #272 on: November 19, 2019, 10:47:23 AM »
"How not to get stuck in the Arctic sea ice"
Akademik Fedorov, a Russian supply vessel to Polarstern is part of Mosaic. Now playing a difficult game in the high Arctic: trying to stay in the thickening winter sea ice without getting stuck.
Long read.

https://www.bbc.com/future/article/20191114-how-mosaic-expedition-avoids-getting-stuck-in-arctic-sea-ice

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #273 on: November 19, 2019, 01:16:49 PM »
Quote
looks like a slowly failing GPS, no? First, you see a delay in updating and then a failure, pinpointing it to a location in the past. GPS accuracy drops quickly when you get down to just a few sats in view
The cadillacs among buoys, the four Mosaic-deployed SIMB3 from Dartmouth, have columns reporting the number of GPS satellites in range, how good an Iridium signal they are getting and how often the buoy calls again if it doesn't get through at first. (They also have pitch, roll and heading but those are not currently activated.)

On average, there are 8.7 GPS satellites within range and these buoys are reporting to phenomenal accuracy, eg 85.739648, 121.329032, which is rounded off to one less decimal by IABP in tracking the 100 buoys of Mosaic.

On land, GPS accuracy greatly benefits from a local tower that supplements satellites that might not have the greatest positions. The Polarstern is moving around too much (drift, jostle, internal wave recoil) for its mast to serve as a fixed local base (see motion below for Nov 18th and 19th).

I don't believe any of the buoys carry radar reflectors or emitters that would allow them to be positioned directly on Sentinel-1AB imagery which is about 41m at PolarView at 85-86º. So there is no way of independently checking reported accuracy even at that level.

source: SIMB3_387850.csv
#gpsSats   iridium_signal   iridium_retries
8.7   4.4   0 averaged over 259 entries

They did have a battery failure on 08 Nov 191/8/19 on the L1 at the remote atmospheric flux station. This turned out to be a defective fuel cell that was no longer charging the batteries. They could not get out there on the helicopter to replace it because of low clouds. This would not be applicable to buoy battery failure: no fuel cells. Ditto polar bears chewing on deployed cables: no cables. Ditto cracks propagating under the foot of the 'swingset' supporting downward looking radiation instruments: no supporting platform for buoys.

Part of their current polar bear problem may be attributable to Mosaic passengers on the Akademik Fedorov feeding them scraps of meat to take selfies.

The Polarstern, contrary to what we have been repeatedly told, has had phenomenal internet service (100 Mbps) the entire way including at its farthest north location, thanks to a new pair of satellites. The PS does not use much Iridium, too expensive for limited bandwidth.

Quote
Kepler provides 120 Mbps uplink to Arctic via satellite broadband
by Bevin Fletcher | Nov 8, 2019 11:32am
https://www.fiercewireless.com/tech/kepler-delivers-120-mbps-uplink-to-arctic-via-satellite-broadband

Toronto-based Kepler Communications said that it demonstrated data rates of 38 Mbps downlink and 120 Mbps uplink aboard the German icebreaker Polarstern to a 2.4m Ku-band Very Small Aperture Terminal (VSAT). The ship is home to the international MOSAiC expedition that involves hundreds of scientists and operations crews examining the effects of climate change on the Arctic.

Kepler is providing high-bandwidth satellite connectivity from low-Earth orbit (LEO) to help the team share large data files between the vessel and shore. Kepler’s satellites are placed around 575 km from Earth, so the closer location helps significantly reduce latency compared to geostationary (GEO) satellites that are 35,000 km from Earth.

“Our Global Data Service provides a cost-effective means to transfer large data volumes that will be gathered over the course of MOSAiC,” said Mina Mitry, CEO at Kepler, in statement. “Rather than only storing data locally and analyzing once physical storage can be sent back with supply vessels, we are giving scientists the ability to continuously transfer test and housekeeping data sets. When its two polar-orbiting satellites pass over the vessel, Kepler is providing 100-times higher data speeds than would otherwise be possible”

Markus Rex said “Kepler's new Global Data Service now enables us to send back bulk data, including key data files for monitoring the status of instruments together with experts at home. This will contribute to the success of MOSAiC."

Quote
How many users 100 Mbps internet line can support?  - Quora

I have 100 Mbps internet line at my work. this supports 130 wifi access points, 400 desktops, 5000 wifi users.  My network becomes limited when more are using. Is 100 Mbps line enough?
 
Is an Internet speed of 100Mbps considered fast? - Quora

Yes, 100 Mbps is considered fast. If you have this connection, it will take only 8 seconds to download a 100MB file at 100 Mbps. This means streaming Netflix, YouTube TV, Hulu, and any other streaming services.
« Last Edit: November 19, 2019, 05:23:32 PM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #274 on: November 19, 2019, 05:41:41 PM »
The image below locates the four SIMB3 surface mass balance buoys relative to the Polarstern on Nov 19th. This worked out because S1B took an image at 07:00 at which time all four buoys reported with an exactly matching timestamp. Not all buoys report all the time when they should (see attached databases of quadruples and triples).

The angles and lengths of the position arrows were provided by the side-angle-side tool (https://www.triangle-calculator.com/?what=sas) given that 86º latitude is 10843.4 pixels from the pole and the 120º longitude is at a 75º angle on Sentinel images served by PolarView.

There is plenty of opportunity for confusion here but when it is all sorted out and their locations and properties plotted over time with variable values such as ice thickness interpolated within the triangles by kriging, it could be of interest. The Polarstern is measuring many of the same things and sending the data to shore with their high bandwidth internet but is not sharing it until 2023.

Ice movement as predicted above from the GFS-envisioned cyclone resulted in cracks, leads and ridging -- some of them repeats -- and considerable agitation of the whole Mosaic floe Ice Camp as seen on Sentinel despite pushing up against its resolution.
« Last Edit: November 19, 2019, 08:38:08 PM by A-Team »

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #275 on: November 19, 2019, 08:41:00 PM »
update on drift speed of the meereis Pbuoys from nov14-19 ~04:00

I took P101 out though it's still sending data sporadically.
(see animation below)
« Last Edit: November 20, 2019, 11:58:54 AM by uniquorn »

blumenkraft

  • Nilas ice
  • Posts: 1586
  • Fans of Hans Club - circa 2018
    • View Profile
  • Liked: 707
  • Likes Given: 982
Re: MOSAiC news
« Reply #276 on: November 19, 2019, 09:11:18 PM »
Those are super cool, Uniquorn. Although they look like sperms. (One of those 'once seen, never unseen' kind of things i guess. :P )
Refugees welcome

blumenkraft

  • Nilas ice
  • Posts: 1586
  • Fans of Hans Club - circa 2018
    • View Profile
  • Liked: 707
  • Likes Given: 982
Re: MOSAiC news
« Reply #277 on: November 19, 2019, 09:13:07 PM »
not really seeing ice thickening at PS location on Cryo2Smos

That's disturbing.  :-\
Refugees welcome

Tor Bejnar

  • Young ice
  • Posts: 3115
    • View Profile
  • Liked: 409
  • Likes Given: 200
Re: MOSAiC news
« Reply #278 on: November 19, 2019, 09:28:14 PM »
It's not climate change, it's all the heat from the scientists running around burning calories!  [/sarc]   ::) :o :-X
Arctic ice is healthy for children and other living things.

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #279 on: November 20, 2019, 12:02:57 AM »
There are quite a few impacts from the Polarstern having to run its generator engines all the time ... twitter scrapes from reporters on the AF and PS and a Science commentary piece 0https://tinyurl.com/yjey6av4).

The bottom line is that it is hard to do large scale science safely and sustainably at such a remote pristine location without observer impacts. But what is the alternative: crappy models, more satellites lacking sufficient resolution, more buoys that break down in the first month, more imagery that cannot be interpreted, no real understanding of ocean-ice-atmosphere couplings or the physics behind Arctic amplification?

How do the Polarstern's impacts stack up relative to commercial shipping, say of anthracite coal being dug out of a Siberian national park today and shipped to China and India?
 
"The next day they told us not to throw meat over the side again to the polar bears ... The data cables would later become chew toys for the polar bears."

"I've wondered sometimes whether the #Polarstern and #AkademikFedorov could have been 1K cooler, a little less brightly lit? It felt like a lot of the heat was coming from the engines or the incinerator. But it was definitely super warm on board. Don't know about the details of the heating system."

"The Polarstern can't be shut down completely, because it must serve as the expedition's power plant, command center, and hotel—which some might call a five-star. There is a dining room that serves mouthwatering cakes every afternoon at tea time, a sauna, a swimming pool, a gym, a bar, and even small shops where you can buy sweets, tobacco, and alcohol. Although the ship is passively drifting with the ice, it will guzzle 15 tons of diesel fuel per day."

"Aboard the Polarstern at MOSAiC's start, I looked forward to seeing the North Star twinkle close to the top of the sky. But the ship floods the ice with light to enable the researchers to work safely and efficiently, creating a powerful glow that blots out the sky. I did not see a single star during my 2 weeks aboard."

The solid layer of blue ice in between snow on top and mushy rotten ice below, has rarely made it past the 50 cm gradation on the measuring stick. The new layer at the top is the firm ice that can support weight. The older rotten ice below is unreliable, although there is a question about whether a thicker layer of it helps or hinders refreezing during winter.

"For AWI biologist Allison Fong, the glow is a problem. She wants to better understand how microbial plant life survives in the Arctic winter, when there is no sunlight for months on end, but she can't study samples that have been exposed to artificial light. So, once a week she plans to drive a snowmobile to what she calls the "dark zone" hidden behind a fortress of small ridges of snow and ice some 500 meters from the ship. There, researchers will don red headlamps (red light doesn't affect most organisms) and collect ice cores that they will analyze back at the ship."

"Oceanographers, meanwhile, are dealing with the heat, turbulence, and wastewater produced by the Polarstern. "At the main site around the ship, there's sort of a dead zone" for collecting data, says Tim Stanton, an oceanographer at the Naval Postgraduate School in Monterey, California. To avoid the issue, he and others have deployed a handful of autonomous instruments, which sit on the ice or float in holes, far from the ship."

"The expedition's own noises, meanwhile, have forced atmospheric scientists to build special defenses. At a site called "Met City" a half-kilometer from the ship, a barrier nearly as tall as a person shelters an instrument that uses pulses of sound to measure the density of air, and thus gain insight into the structure of the atmosphere. The barrier is designed to deaden any artificial noise that might rumble across the seascape, such as blasts of the ship's horn and the whine of snowmobiles and nearby electronics"

"Then there is the Polarstern's sooty exhaust—a huge problem for the many MOSAiC scientists studying the Arctic's unusually clean air. The researchers want to better understand the sources and fates of aerosols—tiny particles suspended in the atmosphere—that contribute to the formation of small ice crystals within clouds. The number of crystals can determine whether a cloud acts as a blanket to warm Earth or an umbrella to shield it from solar radiation. But scientists aren't sure exactly where Arctic aerosols originate: Are they swept north from land, or are they formed by ocean organisms and sent skyward by breaking waves?"

"To gain some insight, Lauriane Quéléver, a chemist at the University of Helsinki and her colleagues will spend the next year pulling air samples through tubes mounted on a refashioned shipping container that serves as her lab. Instruments will parse the samples, analyzing and counting particles.

"Quéléver was astonished to find that the number of particles soared to 28,000 per cubic centimeter—a few thousand times greater than the typical Arctic level—when a second icebreaker arrived to swap instruments, personnel, and fuel. "I really hope that I'm going to get some very natural and pristine data," she says. "Otherwise I can just go home."

"A snowstorm or strong wind could help her and other scientists by clearing the air, but that rarely happened during my stay. When I visited CU atmospheric scientist Byron Blomquist, he was trying to collect air samples at a tower hanging off the ship's bow. But his instruments weren't sampling anything; they had automatically shut down after an alarm signaled that exhaust from the stern of the ship was enveloping the tower. Inside his container, pumps roared like vacuum cleaners as they pushed clean air out through his instruments, to prevent contamination by the smog."

"Less than 24 hours after the Polarstern settled in, the ship's emissions were smudging the ice. I noticed a black flake as large as my fingernail within the icy floe, removed a mitten and picked it up, rubbing the flake between my fingers. It turned to an oily smear. At the spot where the Polarstern had just spent the night. There hung a thick yellow and brown cloud. "Just think," one researcher said, "if we open the Arctic to shipping it's all going to look like that."
« Last Edit: November 20, 2019, 02:27:56 AM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #280 on: November 20, 2019, 10:59:02 AM »
Quote
That's disturbing: not really seeing ice thickening at PS location on Cryo2Smos
Here the time series are using 26 rolling weekly averages according to the Cryosat2 freeboard  / Smos thinness merger of observational ice thickness. There is some thickening over this period in the 0.8 to 1.2 m range. Oddly the ice is thicker to the southeast of the Polarstern's position.

It's hard to say whether that is notably fast or slow relative to past years because there's no long term consistent data set using these satellites and the refined merger algorithm.

Mosaic, in their floe planning, expected to find 1.2m ice here back on October 5th. That is 17 days before the first data frame here and 43 days before the final frame (which has finally attained that thickness). So it's probably fair to say that 1.2 m thickness is quite late.

20191111_20191117
...
20191016_20191022

Technical note: after setting the rectangle bounding the Polarstern's drift over these dates (using locations in the S1AB database from sailwx), set the color picker in Gimp to radius 5 which will fill the bounding box and take the average. Then fill a small overlaid rectangle with this color and pick the whole image with the non-contiguous color selector, capturing the local selection on the palette bar after adding 0.05 m subdivisions from the grid tool. The desired product (ice thickness growth in the Polarstern's vicinity) is then the animation of the palette bar. This could be done on the original png file without the -45º rotation to 'Greenland down' or better still by utilizing the nicely done netCDF file in Panoply which includes error estimates etc and has the option of exporting and analyzing numeric values of the equal area projection pixels. However this won't change bottom-line talking points obtained from much faster image analysis.
« Last Edit: November 20, 2019, 11:09:02 AM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #281 on: November 20, 2019, 11:44:36 AM »
Hefaistos calls our attention to another excellent BBC piece on the Akademic Fedorov trying to get home by Martha Henriques. It is best read in its entirety but a few highlights are being posted below:

How not to get stuck in the Arctic sea ice
https://www.bbc.com/future/article/20191114-how-mosaic-expedition-avoids-getting-stuck-in-arctic-sea-ice

Quote
News reaches us that boxes of kit and the Polarstern’s large gangway fell into a widening stretch of water.

Odd, no mention of this :( event at 'Follow_Mosaic' :). Possible allusion to it in Shupe's AGU blog. I struggled with 'boxes of kit'. Did the reporter mean boxes of clothes, unloaded scientific equipment or just stuff/things in general? Did they sink to the bottom of the sea or were they retrieved? Ditto the large gangway, loss of that leaves them stranded on or off ship, other than the mummy box chair unloaded by crane.

The Polarstern  is looking at a week of total blah weather ahead according to GFS. I don't expect any significant net drift; there's no indication of when or if the transpolar drift wind pattern will set up. Ice deformation can act from afar but I don't expect that this week because the millions of sq km of weak ice acts as a sponge that limits long range teleconnection of compression and extension.

However the Mosaic floe is looking very beat up after the storms and wind divergence on the two S1AB available for Nov 20th:

S1B   2019 11 20 T 06:02
S1B   2019 11 20 T 04:24

Quote
The AK rams the same piece of ice back and forth, maybe 70 times. The drift of the ice meanwhile carries the ship north. During those hours of ramming, the ship drifted so far   that it got back almost to its starting latitude to the north before we finally broke through.

Then the ship gets stuck a second time a few hours later and begins a second period of ramming. That’s when the captain decides to turn on the ship’s reserve engine. The Fedorov has four engines, but rarely uses all at once because this ramps up the ship’s fuel consumption. The fourth engine does the trick, helping the Fedorov break free again.
The captain really really did not want to get stuck for the winter in the ice. The ship is sitting unusually high in the water after the crew shifted scientific cargo onto the Polarstern plus 700 tons of diesel. The loss of weight left the Fedorov sitting so high in the water that the hull reinforcement area is now floating above the ice pack. The Fedorov is not a true icebreaker to begin with.

If the Fedorov were to get stuck, we would have two options: help from the Polarstern or hailing a nuclear-powered icebreaker such as the Yamal. "They are all busy and you don’t know when you will get one. And of course they are very, very expensive.”

The photo below shows how the 25 m wake sealed up almost immediately. The ship was in a compressive regime so ridging was almost immediate.
« Last Edit: November 20, 2019, 12:34:26 PM by A-Team »

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #282 on: November 20, 2019, 11:47:32 AM »
Quote
Here the time series are using 26 rolling weekly averages according to the Cryosat2 freeboard  / Smos thinness merger of observational ice thickness
Nice analysis
Quote
there's no long term consistent data set using these satellites and the refined merger algorithm.
The merged cryosat2-smos v2.02 data appears to go back to nov2010 at AWI (someone's been busy :) ) so a mid term comparison could perhaps be made.

It's been a long journey through google, guesswork, commas and brackets to buoy labels. Today I got the lucky click. So, an update on drift speed of the meereis Pbuoys closest to Polarstern from nov14-20, latest data at ~04:00, with labels. (temporarily without tails)

edit: Polarstern is currently roughly half way between p204 p201 and p207. The buoy on the edge of the floe 'fortress' STIPAKAF ;) is not shown here.
somewhat thicker ice previously also known as fortress

Added an overlay of Pbuoys with labels near Polarstern, circled in red, on today's S1B 04:24 image. The buoy colours just happen to be inverse of drift speed of a much shorter animation. It was an easy way to make them more visible. Click for full resolution.
« Last Edit: November 20, 2019, 05:28:45 PM by uniquorn »

Archimid

  • Nilas ice
  • Posts: 2410
    • View Profile
  • Liked: 310
  • Likes Given: 184
Re: MOSAiC news
« Reply #283 on: November 20, 2019, 02:04:48 PM »
I suggest the term "fortress" is permanently banned to describe supposedly highly dependable thick ice. It may be jinxing things.

A-Team thanks for the amazing series of posts.

 A seasoned Arctic sailor/explorer saying the Polar stern is like a "5 star hotel"  is not likely the same judgement a much less field seasoned scientist might reach. I think that given the environment they face,  a warm ship is imperative. Good food has "magical powers" in stressful situations. Entertainment, good rest and privacy are all required for optimal operation of long missions,  under -20C temperatures, in absolute darkness, over a cracking surface, surrounded by hungry polar bears.

 The Polarstern and its crew of live subjects are not there just to do experiments in the Arctic, they are part of the experiment. They cannot remove themselves from it. They can only minimize their impact and keep their biases in check.  That is not easy because their presence there is like fire on ice.

 This is already amazing research, and this is only the beginning.
I am an energy reservoir seemingly intent on lowering entropy for self preservation.

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #284 on: November 20, 2019, 09:01:12 PM »
Well put, Archi.

I say lay on the amenities, whatever it takes for creature comforts. They have been worried about sustaining Leg 2 morale since the get-go. They'll be beyond the reach of planes, helicopters and even support icebreakers over that time frame.

Mosaic has not released a ship's roster so we don't even know whom to thank. There's crew of 44 and capacity for 80 scientific and technical staff at any given time according to wiki. Most of the people are there to maintain complex instruments at remote sites (no easy job) though scientific team leaders are alongside working too. The breakthrough on internet bandwidth has made a big difference on who needs to be present.

Esther Horvath, photographer and communications manager of the MOSAiC expedition "wants to document the unknown faces on the front lines of climate science. "Who are the scientists? This is what I'm extremely interested in; to show climate change stories through the eyes of scientists, showing how they live in such an extreme remote location."

Chief scientist:
Markus Rex PS122/1: 20 Sep 2019 – 15 Dec 2019
Christian Haas PS122/2: 15 Dec 2019 – 15 Feb 2020
Torsten Kanzow PS122/3: 15 Feb 2020 – 15 Apr 2020
Markus Rex PS122/4: 15 Apr 2020 – 15 Jun 2020
Dieter Wolf-Gladrow PS122/5: 15 Jun 2020 – 15 Aug 2020
Markus Rex PS122/6: 15 Aug 2020 – 14 Oct 2020

The photo today shows some of the damage from recent ice dynamics. We've seen this already on Sentinel but don't know the specifics. However Mosaic has been very cagey about providing a revised map of Ice Camp as-built (and re-built and re-built again) or shareing measured ice elevations above and below.

For example, the open lead necessitating the bridge has itself been cropped out of the photo. We have no idea where the Ocean City and Remote Sensing Site are located today nor how serious the fracture is nor how to follow it on remote sensing.

I don't think this secrecy serves them. The damage control and data hoarding is pure paranoia. No ship radar with 100 Mbps internet? Dumbing down GPS position -- do they fear a cruise missile? What is someone going to do with bearing data beyond a scarcely publishable fourier analysis?

What it boils down to is somebody on the grant committee -- Mosaic dipped deeply into a pot shared with Antarctica and Greenland -- said 'this is never going to work plus it's been done before by N-ICE2015 and SHEBA'. And if they had access to minute by minute floe motion, would now be saying 'I told you so, no way time evolution of a single floe can be followed for a whole year'.

So it seems to be about embarrassment. They are making zero headway on the drift, have no prospects of drifting over the Pole, and the floe is rearranging itself daily if not disintegrating, undercutting the  key talking point about research novelty. This is why Shupe chose the name 'Fortress' for a crummy ice jumble and put it on Mosaic maps.

Today saw the usual turmoil on the Mosaic floe and also some puzzling (since winds have abated) regional torque with the ice no longer moving coherently over the short time frame between two S1AB. Divergent wind stress from yesterday cannot be the attribution because there's not even short term conservation of linear or angular ice momentum because of way too much frictional dissipation.

The two Sentinels are co-registered on the Polarstern (red dot); the faux interferometry that shows complex regional motion in this coordinate frame used the S1AB grayscale average for the 3rd luminance channel of the HSL 'RGB'.
« Last Edit: November 20, 2019, 09:24:59 PM by A-Team »

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #285 on: November 20, 2019, 10:56:00 PM »
Well, if nothing else, it's a pleasure to have almost daily multiple sentinel images of the same area and a wide range of buoys to determine, eventually, how to interpret those images. Sometimes the devil is in the detail.

At risk of overkill, while setting up for a longer term buoy overlay onto S1AB animation I came across this. nov17-18 with 50% intermediate frame, click to run
« Last Edit: November 20, 2019, 11:10:48 PM by uniquorn »

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #286 on: November 21, 2019, 11:12:34 AM »
A lot of movement overnight looking at polarview. New leads on both sides of PS.

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #287 on: November 21, 2019, 04:27:57 PM »
Quote
A lot of movement overnight
This was the worst ice dynamics yet ... and just after they got done laying a bridge over a recurrent lead with scrap palettes and a wooden sled. It is unfortunate that the remote experimental stations have to be tethered to the ship with power, data and fuel lines because these are subject to breakage and burial.

The massive rift chain to the Polarstern's southeast (fresh leads are black on S1AB) shows the response of the ice to to divergent wind stress: extensional faulting resembling NW trending Nevada basin and ranges as the Pacific Plate scrapes past the N American -- and about on the same distance scale. Both play out slowly in time; here the ice failed a couple days after the "84º" wind divergence event pictured in the GFS post above.

A parallel fault zone runs above, below and so far around the Mosaic floe. So far, it has mostly been around because the Fortress is not the path of least resistance for fractures (being thicker). The two Sentinels for the day are in and 'follow-mosaic' :) seems to run a couple days behind events, so unless skies are infrared-clear we won't know until 03:30 utc on Friday what is currently going on.

No question though that the Mosaic floe has experienced some strong ice dynamics even though the wind has been and will be mild short-term. The shifting ice we can see between Sentinel orbits only 3.25 hrs apart may or may not translate to significant damage to remote towers and lines.

The Polarstern itself, forming a barrier much higher than ice, seems to attract ridging, which on the stern side would seem to require daily resetting of slack lines mooring the ship to ice anchors. This recurrent lead/ridge may have originated from the icebreakers maneuvering early on.

Between onboard pollution from the ship's stacks and daily repairs to deployed sensors, scientists may not be getting the steady data streams expected. It would be no better on another floe. Trip leaders said early on that this was probably the last possible year for such an expedition. Maybe it was last year.

It's always been clear that Sentinel radar images, like Ascat scatterometery, arrives at sub-optimal visual contrast but it's never been clear how to best fix a given image, much less make a matching stack for a time series given variations between different orbital angles and nadirs.

Sentinel currently provides 82 images back to mooring on Oct 5th with only five days of missing coverage. The attached S1AB csv database has been merged with sailwx, AwiMet and nullschool GFS to bring in location, wind speed, wind direction, air temperature and links.

The easiest thing to do in terms of small file sizes, convenient format, quick turn-around and pioneering a pipeline for the the ultimate unattended objective fix of the jp2 is use the jpg for now and its intrusive graticule, followed by the Level tool in gimp, as shown in the mini-tutorial attached.

Sentinel images on ice are inherently speckled and while 'there's a app for that'  a slight unsharp mask can bring further small benefits. One size does not fit all however: improving a 100x100 km scene is not optimal for smaller sub-scenes like the Mosaic floe. There we would like to maximally improve the resolution or clarity of the image right around the Polarstern and the Mosaic floe.
« Last Edit: November 21, 2019, 05:36:34 PM by A-Team »

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #288 on: November 21, 2019, 07:16:32 PM »
Awright! AWI finally posted an updated but already outdated map of the ice camp and its drifting pieces. The bad news is that the map is badly out of focus because designers did not anticipate what 'social media' would do resizing it for phones. This topic will never be trending with millennials; few could even locate Norway on a map.

I have not been able to locate the actual map. Actually it is very affordable to have an ordinary web site and simply post stuff in html 1.0 and upload intact big figures with ftp. In fact they could just use my site for free and get their incompetent graphics fixed to boot. They really shot themselves in the foot with this crazy complex javascript and cascading style sheet eye candy. It doesn't work.

Next we read about how they cannot track locations of their remote equipment. Apparently they don't co-install a buoy, tracker beacon, two-dollar flasher or radar reflector on expensive remote equipment (OPM: other peoples money). They could have just emailed uniq on their 100 Mbps internet to get the current drift.

The fuel cell failure mode is exactly what they would get in Minnesota on a cold day. The rising number of mundane equipment failures makes me wonder about pre-trip testing standards. This is an austere lunar landscape, not a mid-latitude campus.

Reading to end of this blog, it appears one co-leader has a very loose grip on what drives TransPolar Drift (in recent years, not much and very erratically) and on what is going down right now: one cyclone chasing another when TPD anti-cyclones are way overdue. They seem not to look at GFS even 2-3 days out for a drift forecast.

The way things are going -- and given what is coming -- perhaps it is time to consider a staged withdrawal:

-- Forget about fancy remote area deployments, pull back to the ship, put out more autonomous buoys, do helicopter forays to benefit satellite photo interpretation, carry out excursions to measure ice thickness, let the ship loose and motor slowly into the wind to keep the bow air clean. The ship moored to solid ice makes for a poor platform not fit for purpose because unavoidable emissions severely impact atmospheric and other monitoring goals.

-- Wind down Leg 1 in an orderly manner and consider downsizing, repurposing or even eliminating Leg 2. The Mosaic floe as originally envisioned is dead in the water. It is being torn apart and rearranged almost daily; no purpose is served by continued monitoring the core ice jumble-ridge which is atypical and unrepresentative of the Arctic ice pack.

-- Don't admit to planning mistakes but don't continue with them either. Give the redesign a new wrapper called responsive resilient adaptive management (RRAM) or similar.

Quote
11/9/19 Another Failed Mission
Postcards from a frozen icebreaker – Part 12 19 Nov 2019
By Matthew Shupe

L1 radiometer flight as high on the agenda today… but kept getting pushed back. Including by a bear visit, which closed down most operations. But finally in the afternoon we were able to hit the air. Loaded our gear on the helicopter….. sled, ladder, shovel, feather duster (great for instrument cleaning), tool box, numerous safety bags, rifle for the bear guard and a spare, personal kits (need those spare gloves), and most importantly our replacement power systems.

It is about 280 lbs…. two big batteries and a fuel cell, all set up in a box. We fired it up for about an hour before the flight to get it all toasty warm inside. Then shut it down, closed the box, and loaded it on the helicopter.

On site the plan is to quickly get it over to the flux station, plug it into a fuel cartridge, and fire it up again, to stay warm while we prepare the station for the power system transplant. Such a great plan…… but unfortunately the remote controlled helicopter landing lights at the remote station did not work, so we flew around for 1.5 hour looking for it.

So tough to find as everything is white (and snow covered) but it is totally dark out. Part of the challenge is that GPS coordinates quickly become outdated due to the ice drift. We get a reading from a buoy from ½ hour before but the whole pack has moved. This unsuccessful search had me really bummed and wondering if/when we would find this site.

It all went off without a hitch or at least without a big hitch on 11/10/19. Well planned in advance, we had the right tools and our specially concocted power system in a box. Beautiful clear skies and cold…. Below -30C for the first time at MOSAiC. The helicopter dropped us off at the L1 site, some 15 km away from Polarstern. Our new power system is getting colder. It won’t start if it gets below the freezing point.

We started up the spare system right away so it could keep itself warm, then opened up the old system, dark and cold, to remove the frozen system. The exhaust system was the culprit, a huge frozen beard extending down from where the water exhaust drips out. Ice extending up the exhaust tube
« Last Edit: November 21, 2019, 11:37:17 PM by A-Team »

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #289 on: November 21, 2019, 08:43:40 PM »
Quote
They could have just emailed uniq
One day I'll post a picture of our satellite dish strapped to a tree :) .

Made a mistake after playing with the buoy animation but left it as is since it shows the full path. It may be the sudden change in direction that opened up all the leads. A bit late, I was catching up with the spacex thread.

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #290 on: November 21, 2019, 11:33:41 PM »
Quote
that sudden change in direction opened up all the leads
Or whiplash. When the eye of a hurricane/cyclone passes overhead as it kinda did here, the wind reverses direction. Extensional leads rather than compressional ridges because ice on the edges moves centrifugally outwards. But where exactly and what good would advance knowledge do for the Polarstern?

In terms of replacing the white base under your buoy animations with appropriately dated S1AB or nullschool GFS winds (to get at the correlation pictorially), the S1AB go about 20 hours between 3 close coverages and the GFS 3 hours whereas the buoys commonly report hourly. So there is somewhat of a mismatch in time resolution to buoy GPS.

The Simb3 buoys report every four hours. However most buoys can be programmed to call in at whatever interval, from ten minutes on up. Some may be able to take downlink re-programming calls.

Windy gives update intervals and the time of update release (as of 2016 unfortunately) for all the common weather satellites and has stopped providing Arctic ECMWF.

https://community.windy.com/topic/3340/when-and-how-often-are-the-weather-data-updated

On the four Simb3, call them abcd, retrieving their record of 924 dial-ups back to Oct 6th deployment revealed quite a mix of buoy position reporting combinations that complicate matters for making two delaunay triangles or just one  (or just an inter-buoy line) consistently.

The attached csv database is tricked out with extra sort columns to enable extraction of any desired sub-collection. The Simb3 are key to local ice thickness development, in combination with whole-ocean 25x25 km cryo2smos.

636 abcd   (all buoys reporting)
086 bcd    (buoy a not reporting)
041 cd     (buoys ab not reporting)
012 abd    (buoy c not reporting)
012 acd    (buoy b not reporting)
011 c      (buoys abd not reporting)
006 abc    (buoy d not reporting)
002 ac     (buoys bd not reporting)
002 bd     (buoys ac not reporting)


Uniq and macID have built a better mousetrap. That is, one-hour ice movement is much more refined than two-day OsiSaf as a better match to the intrinsic time scale of winter ice movement.

That is, the PS lat lon changed, in 1026 available hour increments at awiMet as poorly measured by one dp data are distributed as below. Note 0.1 in lon is a lot less distance travelled than 0.1 in lat this far north; the distribution estimate could be refined by using 4 hr windows. The Polerstern is currently moving very slowly south and east.

667 no change lat, no change lon
151 no change lat, westward 0.1 lon
117 no change lat, eastward 0.1 lon
010 0.1 south in lat, no change in lon
004  0.1 north in lat, no change in lon
002 no change lat, 0.3 westward in lon
006 no change lat, 0.2 westward in lon
008 no change lat, 0.2 eastward in lon 
002 0.1 north in lat, 0.1 west in lon
003 0.1 north in lat, 0.1 east in lon
010 miscellaneous

While ice motion so described is limited to inside the buoy envelope, it could give advance notice to the Polarstern of leads and ridges outside their immediate location in advance of the next S1AB. The buoy pattern is in effect an imaging device, resembling the discretized ommatidia of a small invertebrate eye.

OsiSaf uses pattern recognition but those point features (virtual buoys) change from day to day and can be quite sparse. OsiSaf is not currently pointed at S1AB but could be, giving ultra-refined ice motion between timestamps in conjunction with the buoy product.

Clearly uniq could dump the gridded underlying data into netCDF time series format to distribute the underlying numerology though trailing tails (but not colors) would be lost.

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #291 on: November 22, 2019, 11:47:17 AM »
The situation has gotten increasingly dire in the immediate vicinity of the Mosaic ice camp as extensional forces have opened multiple km-wide bands. A second Sentinel image of 05:46 has just come in that provides a snapshot of ongoing motion -- the situation has worsened considerably in the last three hours. PolarView has just now posted a third intermediate time, the 04:08, but we won't be able to follow events further until tomorrow morning other than hourly weather and drift which are unremarkable (there's no storm) since the last S1AB:

   Lat  Long  YY-MM-DD  UTC     Wind       T(C)  hPa
  85.7  120.7 19-11-22 10:00    5   50    -16.3 1016.0
  85.7  120.8 19-11-22 09:00    4   40    -17.0 1015.5
  85.7  120.8 19-11-22 08:00    5   10    -16.1  1015.0
  85.7  120.8 19-11-22 07:00    4  340    -16.1  1014.5

This event will be extremely disruptive to all deployed experimental equipment and effectively destroys the floe for purposes of studying its evolution over a year. Let's hope the Polarstern and all the people aboard stay safe as the event continues to unfold today.

This is a very unfortunate situation ultimately attributable to climate change, its amplification in the Arctic Ocean and the long-term trend to thinner, more mobile ice that cannot resist wind stresses.

The upper two images follow Mosaic convention with north up, the first needs a click to display full size; the lower animation has S1AB orientation of the Pole down the 120º meridian with inverted insets in coordinates attached to the ship at 2x enlargement.  The 4x enlargement is also north down after contrast leveling and light unsharp mask.

Some of the 87 surviving autonomous buoys out of the 100 deployed in the Mosaic project will likely succumb to the same ice pack deformations affecting the Polarstern. So far, they all reported on the 22nd or 23rd with the exceptions below. All but the last two are Chinese buoys that have previously had reporting delays.

300234068118580   SVP    PRIC   MOSAiC   15 11 19   86.19   118.51   
300234010077180   GPS    TUT    MOSAiC   19 11 19   86.08   117.27   
300234065091830   GPS    TUT    MOSAiC   19 11 19   86.01   123.00   
300234065090840   GPS    TUT    MOSAiC   21 11 19   85.51   120.54   
300234065980590   GPS    TUT    MOSAiC   21 11 19   85.73   123.98   
300234065981590   GPS    TUT    MOSAiC   21 11 19   85.61   118.57   
300234065984580   GPS    TUT    MOSAiC   21 11 19   85.95   121.98   
300234065985570   GPS    TUT    MOSAiC   21 11 19   85.92   121.06   
300234067705700   IT      AWI    MOSAiC   21 11 19   85.75   121.63   
.................90094   ITP   WHOI   MOSAiC   21 11 19   85.68   121.98   
« Last Edit: November 22, 2019, 04:33:00 PM by A-Team »

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #292 on: November 22, 2019, 04:35:35 PM »
Why so many big leads during relatively calm weather? I reread the two posts above :)

Update on drift speed, this time looking more closely at the polarstern area from nov19-22 ~03:30. PS is roughly half way between p201 and p207. The diagonal is 120E, the line top right is 85.5N. I'll attempt to add more graticule with labels and a rough time as we go along.
I wonder if they can feel any of that movement, in particular the recent jolt to the north shown by the more easterly buoys.

Quote
Clearly uniq could dump the gridded underlying data into netCDF time series format to distribute the underlying numerology though trailing tails (but not colors) would be lost.
@macid, have you ever tried ncdf4
« Last Edit: November 23, 2019, 01:45:09 PM by uniquorn »

blumenkraft

  • Nilas ice
  • Posts: 1586
  • Fans of Hans Club - circa 2018
    • View Profile
  • Liked: 707
  • Likes Given: 982
Re: MOSAiC news
« Reply #293 on: November 22, 2019, 04:38:59 PM »
Wow, thanks for pointing that out, Uniquorn.

But what's the chicken and what the egg here? ;)
Refugees welcome

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #294 on: November 22, 2019, 05:07:44 PM »
Quote
Why so many big leads during relatively calm weather?
Even though wind speeds have been mild, wind velocity has been very confused in the 200x200 km sector centered to the SE and including the Polarstern. The weak ice of today experiences mechanical failure upon being pulled in different directions even under moderate stress.

The alternative, a delayed reaction to stresses from storm last weekend, doesn't hold water: too many days have elapsed. However that storm was likely a major contributing factor in establishing the initial ice fractures that later became big leads, opening easily under mild divergent winds.

The next cyclone rolling in will brush the Polarstern on Nov 24th before wandering off to Severnaya Zemlya. Winds will be fairly strong (peaking ~32 km/hr) for a day but GFS does not foresee surface winds anywhere near the 51 km/hr of the last storm. Wind stress goes roughly as the cube, so the previous event was 4.3x worse.

Note Uniq's buoy trajectories are in effect a non-gridded replacement for OsiSaf ice motion with a fantastically improved time resolution, an hour vs 48 hours. Given that buoys report location to 4-5 decimal points, whereas OsiSaf is applying AI to low resolution all-Arctic imagery, the accuracy is much higher with the buoy mesh.

The tangent vectors to these curves at the times the S1AB are taken (not marked or paused yet) are then the arrows that should be drawn over the high resolution imagery instead of what we see at mosaic_multisensor. These arrows will be very accurate both in magnitude and direction.

Even if F not quite = ma under the circumstances, it is close enough for citizenry work. Here we want to get at the force acting on the floe in which the buoy is emplaced or rather its rigid body extension. Here you would want go back to an old indian trick called Newtonian Differences rather than differentiate fitted continuous curves.

Set up a column of positions, let the adjacent column subtract t from t-1 and fill down, let the third column do the same for the second, recalling change of position with time is velocity and change in velocity with time is acceleration which with mass fixed amounts to force acting (geometrically the tangent to the velocity curve). These new columns are to be colored differently in supplemental animations though the time steps and positions remain the same.

OsiSaf ice motion is severely biased by the big moves during the 48 hour block; if for example, these might be in the first two hours with the latter 46 providing little displacement, then the image served will not be at all current. This explains the apparent incomparability with GFS 3-hr winds.
« Last Edit: November 22, 2019, 06:34:54 PM by A-Team »

gerontocrat

  • First-year ice
  • Posts: 6824
    • View Profile
  • Liked: 1728
  • Likes Given: 22
Re: MOSAiC news
« Reply #295 on: November 22, 2019, 05:34:46 PM »
Just when I was going to post a little gif A-team makes it redundant.

Here it is any way -
- calm tomorrow,
- 24 Nov       South wind 30 kmh
- 25-26 Nov  East wind 20 kmh
- 27 Nov much calmer but still FJL next stop.

Given the state of the floe and the surrounding area it all looks a bit dodgy, but just as well those big winds in the Beaufort / Chukchi / Bering Strait are a long way away.

As they say in every episode of all the best Police series "Take care out there".

gif plays 3 times & stops
"Para a Causa do Povo a Luta Continua!"
"And that's all I'm going to say about that". Forrest Gump
"Damn, I wanted to see what happened next" (Epitaph)

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #296 on: November 22, 2019, 06:57:26 PM »
Let's get this right because this is a very serious ongoing event.

Supposing this weekend storm had a causative role in the present troubles at the Polarstern, it might be worth overlaying the 3-hr nullschools over the changing S1AB, using the PS positions in the computed nullschool url as provided for everyone's convenience in an earlier cvs attachment.

The synthetic problem though is the resolution of nullschool maxes out at its '3000' scale whereas the S1AB ROI for Mosaic is just a postage stamp then under the GFS, with vice versa no better. However S1AB still conveys the lead situation zoomed out so there might be a workable compromise that doesn't overshoot the 700 pixel forum width too badly.

Technical note: as above, take whole window screenshots and layer up in Gimp. To avoid jerking the nullschool image with inadvertant mouse-overs, use keyboard commands lik j and k to advance and back the time by 3 hours. Three days is 24 layers; that won't result in an overly large final file. Duplicate the file twice and crop to the url and data box for wind speed. Crop the main file to final size and enlarge the canvas on the duplicates to match. Then tile up as three layers, add the S1AB layers duplicate to match, add transparency to the GFS, flatten over the radar, then slice the tile over in ImageJ, save as animated gif, open and re-save in gimp with good pauses where appropriate. YES YOU CAN.

85   S1B   2019 11 22 T 05:46   22 11 19   05:46   85.7   120.8    4   320   -14.8
84   S1B   2019 11 22 T 04:08   22 11 19   04:08   85.7   120.8    4   320   -15.0
83   S1B   2019 11 22 T 02:30   22 11 19   02:30   85.7   120.8    4   350   -14.9
82   S1B   2019 11 21 T 06:43   21 11 19   06:43   85.7   121.0    4   360   -16.2
81   S1B   2019 11 21 T 03:27   21 11 19   03:27   85.7   121.0    6   30   -16.0
80   S1B   2019 11 20 T 06:02   20 11 19   06:02   85.8   120.5    6   230   -13.8
79   S1B   2019 11 20 T 04:24   20 11 19   04:24   85.8   120.5    8   240   -15.3
78   S1B   2019 11 19 T 07:00   19 11 19   07:00   85.8   120.6    6   240   -11.2
77   S1B   2019 11 19 T 05:21   19 11 19   05:21   85.8   120.5    7   240   -15.4
76   S1B   2019 11 19 T 03:43   19 11 19   03:43   85.8   120.5    8   240   -15.3
75   S1B   2019 11 18 T 06:19   18 11 19   06:19   85.9   120.8   14   20   -10.6
74   S1A   2019 11 18 T 05:30   18 11 19   05:30   85.9   120.9   16   10   -10.2
73   S1B   2019 11 18 T 04:41   18 11 19   04:41   85.9   121.0   16   10   -10.5
72   S1B   2019 11 17 T 05:38   17 11 19   05:38   86.0   122.5   13   320   -15.7
71   S1A   2019 11 17 T 04:49   17 11 19   04:49   86.1   122.5    4   320   -16.2
70   S1B   2019 11 17 T 04:00   17 11 19   04:00   86.1   122.4   14   310   -17.7
69   S1B   2019 11 15 T 05:54   15 11 19   05:54   86.2   118.4   11   170   -26.1
68   S1B   2019 11 15 T 04:16   15 11 19   04:16   86.2   118.3    2   260   -15
67   S1A   2019 11 15 T 03:27   15 11 19   03:27   86.2   118.3    2   260   -14.9
66   S1B   2019 11 14 T 03:35   14 11 19   03:35   86.2   118.2    9   190   -14.5
65   S1A   2019 11 13 T 03:43   13 11 19   03:43   86.1   117.9    9   200   -13.8

AWI communications have been strangely unscientific. The daily 'follow' blog is misdated, a day or more behind events depicted; photos have no timestamps, exifs are stripped; ship time is updated weekly but not provided, we have no idea what UTC the news refers to; the twitter feed is hardly used, no one on board is sharing despite huge bandwidth, the Helmholtz blog has gone dead, ship GPS is obfuscated and so on. It is not a good idea to apply spin templates from political campaigns to scientific matters. Honesty is the best policy.
« Last Edit: November 23, 2019, 10:39:48 AM by A-Team »

macid

  • New ice
  • Posts: 26
    • View Profile
  • Liked: 13
  • Likes Given: 16
Re: MOSAiC news
« Reply #297 on: November 22, 2019, 09:27:22 PM »

Quote
Clearly uniq could dump the gridded underlying data into netCDF time series format to distribute the underlying numerology though trailing tails (but not colors) would be lost.
@macid, have you ever tried ncdf4
can read netcdf and still some room left for learning, what should go in the netcdf, just a conversion of the bouy csv with dimensions for lat, lon, time and values for the working buoy data?

I don't understand what you want to do with it A-Team tbh but happy to look into

uniquorn

  • Nilas ice
  • Posts: 1636
    • View Profile
  • Liked: 790
  • Likes Given: 164
Re: MOSAiC news
« Reply #298 on: November 22, 2019, 10:21:14 PM »
He could load it into panoply and look at stuff without R skills.

Here is a buoy overlay onto yesterdays 03:27 S1B. The buoy locations are taken from frame 3 of a 23 frame animation of yesterday's buoy data (I haven't worked out how to limit hours yet). macid may correct me but I think R animate makes frames from an average of available data, letting you choose how many frames make up the animation. Buoys that send data every 10m would probably be averaged. Perhaps it would be interesting to select only those buoys and run a more detailed ani, depending on their locations. Many of the pbuoy's data ends at 2300 when I download it s we have to wait. 3 or 4 hours is a lot of movement these days.
The file is big but it's click to run  It's only Neven's storage bill slowly creeping up ;) Optimising reduced file size by 40%. An animation of this poses some challenges.

Polarstern shows up clearly between p201 and p207.
should have reversed the order
« Last Edit: November 22, 2019, 10:38:38 PM by uniquorn »

A-Team

  • Nilas ice
  • Posts: 2453
    • View Profile
  • Liked: 305
  • Likes Given: 22
Re: MOSAiC news
« Reply #299 on: November 22, 2019, 10:38:08 PM »
Quote
why netcdf?
It is first about mainstreaming the buoy project. It's such a good one that resistance is futile.

For better or worse, nc and hdf have become the meeting ground for sharing climate data, excel/mysql wrapped in steroids. System integrators -- builders of the immense robotic pipelines of climate science -- do not want to crawl a hundred sites, massage a hundred formats, write a hundred unresponsive sysadmins, nor tweak other people's ad hoc code.

In fact the whole idea is not to look at the raw data at all but assume it is schema-compliant so a remote climate modeler or weather prediction provider can ingest it over and over without further intervention.

It is second about getting out of banks of numbers into bottom-line pictures: working on a more limited scale within Panoply, an open and active low learning curve netCDF data viewer, data type combiner and map-maker that allows the average joe/sephine to put out a decent and original product, for example changing projection to equal area or re-scaling to overlay properly on satellite products, or here correlating ice mobility with say surface wind speed or vertical ice temperature profile.

In other words, it is better to crunch pictures than numbers. Or rather multiple correlated stacks of pictures. Why? Because you can see what you are doing, because there are ten thousand intuitive photoshop-type algos, because the data is inherently indexed by location and timestamp.

It's all about rapid prototyping. Work with the pictures until you know there will be a result, then go back to the numbers and do it right.

Is this a spec for implementing buoy position and sensor data as netCDF? I would say no, it's little more than a vague vision statement looking down the road to better sharing.

Is the priority so high that the weekend should be spent immersed in human-unreadable binary? For me, I would say no to that.
« Last Edit: November 23, 2019, 09:26:26 AM by A-Team »