MOSAiC news

[uniquorn]

Any volunteers to keep an eye on a new snow buoy, 2020S98, close the the pole?

Also 2020S106, 2020S107 and Also 2020S108
The ever improving meereisportal makes it visually simple.

update, a large range in snow depth, increasing by 20-60cm since deployment. (click)

[A-Team]

The Smos and Smos-Smap ice thinness retrievals show the advance of their cut-off (ice thicker than 0.5m, shown as tan color) over the freeze season. For a slow re-freeze like this year, some regions can lose quite a few weeks of ice growth leaving them poorly conditioned for the melt season next spring.

The first graphic below compares 2020 with 2010-2019 for Nov 9th. The red color shows >0.5m ice in an earlier year but not this year; blue indicates the opposite. There is an excess of red in each of the comparison years available -- overall 4x as many red pixels (191,886 to 53,412) -- even though the Beaufort has had a normal or even rapid re-freeze this year.

The second figure shows successive nesting of half meter ice at ten day intervals from Sept 15th to Nov 09th. The idea here was to establish ice classes based on how long ice growth has continued since newly attaining 0.5m. While there has not been enough ice movement yet to distort the nesting, neither 5- nor 10-day intervals proved long enough to provide clear categories.

It may be necessary to go to 21- or even 30-day intervals to establish clear classes; not enough time has elapsed so far this fall for this. The open water still seen on Nov 12th may not even freeze over at all until late December for the Chukchi and Svalbard-FJL areas, with more weeks needed to attain a half meter thickness.

It's also easy to track boundaries of thicker ice using CryoSat2 in combination with SMOS. The perimeter of >1.0m ice is shown at the bottom for 13 days ending Nov 9th. It's hardly advancing.

ftp://ftp.awi.de/sea_ice/product/cryosat2_smos/v203/nh/2020/10/

The Laptev-ESS are forming a measurable ice skin but an elevated SSTfnd in the underlying mixed layer water column could delay attainment of half meter or more thickness. Although ice can thicken rapidly at first given cold enough air temperatures above, when several weeks of freeze season are essentially lost, not enough time remains for brine exclusion maturation..

Three drifting buoys are still reporting hourly subsurface seawater temperatures from the shrinking open water pockets of the Laptev. Uniq attached that data as a txt file above; it opens readily in excel etc if .csv is appended giving fileName.txt.csv The third file shows the graph of buoy #761 up to day number 316 (Nov 11th).

A lot more analyses could be done on these three buoys and others set in the ice, for example comparison of observed SST and posted Ospo-GHRSST. Uniq has requested 'more hands on deck' several times but so far forum members have not stepped forward to help with charting.

Ironically this is one of the few areas where Mosaic data is openly available and well-organized at meereisportal. The Polarstern installed many dozen buoys in the ice and 74 of these are still actively reporting data that cannot be obtained from satellite.

https://tinyurl.com/yx9yle3o meereis buoy data portal
https://iabp.apl.uw.edu/maps_daily_table.html additional international buoys in Arctic

[Technical note: The daily png ice thickness maps provided by Cryo2Smos are badly dithered, as is the scale bar. That is, each day uses many thousands of colors rather sticking to a palette of 256. This causes successive days to use similar but not identical colors for ice thickness. There is no upside to this, no visual benefit.

The downside is it is not possible to make a time series in gif format because the choice of 256 colors has to be fixed. Out of gamut colors then have to be brought to the 'nearest' element of the palette, causing major distortions and inconsistencies in data representation.

The fix: tile up the daily images in gimp, then change from RGB to indexed color. This forces a global choice of 256 best colors. Upon slicing, each of the frames uses the same palette so upon reassembly into layers, saving out as gif gives a satisfactory animation.

The netCDF file here is properly made so the png problem may have arisen from improper resizing somewhere along the production pipeline]

[A-Team]

The image below shows an accurate view of one day gains and losses in open water, Nov 11th relative to Nov 10th. This is done in AMSR2_AWI because its clean palette allows for every pixel in the image to be unambiguously assigned to land, ice at 1-100% concentration or open water. Upon 'grain merge' in gimp, only three colors result: black for no change in classification, red for open water loss, and blue for open water gain. (Land has been restored to provide context.)

Click to see at the full original resolution.

Regions of open water loss exceeded those of open water gain by 2.76:1. Overall though, very little changed: only 1.8% of the ocean pixels shown. It appears that the Laptev, ESS will be completely frozen over in 3-4 days; the Kara may be delayed by a passing weather system. However the Chukchi and Sv-FJL are not moving towards freeze-over any time soon.

Unlike during the Mosaic year, no Transpolar Drift has set up yet, meaning no significant movement yet of ice formed in the Laptev heading for export out the Fram. There is a fair amount of variability historically so not too much can be read into this for a mid-November date.

The Polarstern helped acoustic thermometry moorings as a moving ship receiver. The Healy had set three more out in the Beaufort but an engine room fire left the US without an icebreaker to retrieve the data; however a Norwegian ship crossing the Northeast Passage in mid-October picked them up the week of  Nov 9th before their batteries went out.

The CAATEX project moorings measure temperature and heat content under the ice across the whole Arctic Basin based on sound propagating faster in warm water than in cold. This requires atomic clock accuracy but that device in the mooring is a huge drain on the battery.

So far we have seen sound spectra but not Radon tranforms that invert integrated signals -- but it could be very important in replacing models and spot sampling with broad observational time series.

https://www.nersc.no/project/caatex
https://tinyurl.com/yyhjxxkv Acoustics Today excellent introductory article
https://ui.adsabs.harvard.edu/abs/2020EGUGA..2220347S/abstract
https://www.nersc.no/project/caatex

[uniquorn]

A quick catch up with the mosaic floe2 cluster before the low pressure passes nearby.
Roughly between 88N and 89N and 10E-33E

[A-Team]

Seasonal summary for the Siberian side: 182 days from earliest open water on May 15th to end of season on Nov 12th. (A small area of open water, 1350 pixels worth in AWI's determination, has persisted through Nov 13-14th.)

Peak open water occurs significantly offset from peak isolation though considerable overlap occurs by mid-July; the match will become better as Arctic Amplification, asymmetrically warming 2m air on the Siberian side, and Laptev atlantification proceed. This will increase adsorption of solar energy by low albedo open water, assuming similar cloud conditions.

The third image shows the average occupancy of open water over the 182 days (using 'royal' to color a grayscale mean). It's not clear whether hot spots are one-off locations or regular features favored by currents and bathymetry.

The fourth image is a snapshot showing the entire timespan in miniature; the slow ascent but steep decline show up very clearly. The former is driven by melt pond momentum and early sun; the latter is due to diminution of vertical mixing and very cold air temperatures arriving in the late fall.

It's wrong to think a date lid exists for the Siberian side; however freeze-over will likely remain earlier than the Pacific-influenced Chukchi for years to come. Eventually, as fractional BOE advances, there weill be more open water both earlier and later, with freeze season restricted to mid-winter months.

Open water on the Siberian side is affected by quite different processes from that in the Chukchi or along the Svalbard-FJL-SZ line. These have to be considered separately lest the unprecedented 2020 statistics on the Siberian side be diluted. Here the 110-180 wedge effectively isolates the
Siberian side but with a slight amount of ESS included. No other year in centuries remotely approaches the open water extent and persistence into mid-November this year of the Siberian side.

The Polarstern, despite the vast data collected, did not deploy any instruments in this area during the relevant time period. Mosaic had no way of anticipating the extreme TransPolar Drift much less the anomalous open water off Siberia. Options for adaptive re-scheduling are very limited because of complex ship logistics.

[Technical note: ImageJ has two very useful commands for opening up files of a large time series. One is buried in buried among the file opening commands, Import ... Image Sequences...'stack'; the other in Image --> Stack --> Tools --> Stack Sorter that can assure the set of stack files are in correct chronological order. Another very useful item in the stack menu namely 'measure stack...' thatcalculates a specified variety of quantities for each frame in turn, putting the outcome in a small text database as well as graphing it forward or reversed. It would otherwise not be feasible to go through a 182 layer stack taking manual histogram readings.]

[uniquorn]

A look at relative drift in the mosaic buoy cluster. This presentation centres on one of the buoys in the cluster by subtracting its coordinates from all of the others, allowing us to study their relative drift.
oct11-nov13
edit: made it a bit smaller
the next few days should be interesting

[oren]

The CAATEX project

Cross posting an article about this project and the recent retrieval, easily readable in google translate.
https://www.nrk.no/urix/unike-klimadata-fra-polhavet-redda-etter-kamp-mot-klokka-1.15231671

[A-Team]

Thanks! Good new information there:

"We have measured across the Arctic Ocean for a whole year and down more than 1000 meters where there is little data from before, says M Dzieciuch and H Sagen. The six moorings retrieved have a total of 375 instruments."

Open water in the northern Laptev hangs on as a shrinking polynya despite a 962 hPa cyclone  moving the ice pack around significantly (fig.3). The time series through Nov 15th (fig.1) has had its AMSR2_AWI recolored by concentration class to bring out open and near-open water.

The Laptev water column SST to 10m is still a degrees or two above the freezing point of sea water according to OSPO-GHRSST, with the Bering Sea, Chukchi and Sv-FJL-SZ and Barents so warm they have to be shown with a much broader temperature scale extending 7ºC above -1.8ºC (fig.2).

The distribution of open water in the Laptev Sea has narrowed to a very restricted region (hot spot) consistently showing elevated temperatures for a couple weeks with respect to the rest of the sea.

This special spot might be a result of turbulence bring warm AW up near the surface where the flow diverges at the junction of the Lomonosov Ridge and Laptev Shelf (fig.4), abnormal geothermal flows associated with the nearby Gakkel Ridge spreading zone, momentum changes induced past the shelf curvature, or just coincidence. Geology and bathymetry at LR/LS junction:

https://academic.oup.com/gji/article/163/2/698/634963
https://tinyurl.com/yy4sbyml
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2014TC003590

The Polarstern measured currents, turbulence, tides  etc. throughout the drift, though not close by this region of the Laptev. No data has been released so far:

85.1  133.2 19-10-04 starting lat lon
85.0  136.2 19-10-08 farthest east.

The hot spot is somewhat east of the mooring data acquired by Polyakov et al 2020 at the top of this curvature but the wind and tide mechanisms identified there may still be applicable:

"Intensification of Near‐Surface Currents and Shear in the Eastern Arctic Ocean
IV Polyakov et al. Aug 2020
https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2020GL089469 free full

Measurements of currents from a 15‐year duration mooring record in the eastern European Basin  of the Arctic Ocean demonstrate that the previously identified weakening of stratification in the halocline has been accompanied by increased upper‐ocean current speeds and associated current shear.

Most of this increased energy and shear is in the semidiurnal band, which includes baroclinic tides and wind‐driven inertial oscillations, with little change of mean along‐slope water transport. The increased shear together with weakening stratification identified earlier indicate a greater potential for shear‐driven turbulent mixing, consistent with the recent transition in sea ice and upper ocean state to conditions previously unique to the western Nansen Basin [[ie atlantification north of the Sv_FJL-SZ line]].

We hypothesize that this increased coupling between AW heat and the sea ice may lead to a positive feedback between reduced sea ice and higher mixing rates as the longer periods and increased areal extent of open water facilitate more energetic wind‐driven inertial oscillations, less damping of baroclinic tidal currents, and associated upper‐ocean shear coinciding with weakening halocline stratification.

As sea ice declines, a new Arctic state is emerging which due to the positive feedback mechanism outlined above may be pushing the system toward a tipping point."

[uniquorn]

https://cmems.lobelia.earth/data?view=viewer looking at Laptev salinity. The charts are useful and interactive showing data from a range of years, location and depth. Here choosing quite a tight salinity range to highlight small changes noticeable since 2018.

[A-Team]

no Beaufort Gyre yet?

Classical wind patterns have not set up so far, no consistent Fram export either. Polar vortex issues?

That CMEMS-Lobelia is quite an astonishing portal for global climate change, it's not limited to the Arctic. Below, that same salinity but run through the 'isophote' contourer provided at the equally amazing G'Mic portal under the gimp filters menu.

For inter-year comparison, it might be better to use the grayscale palette so the screenshots can later be layered in gimp for 'grain extract' and subsequent recoloring with the divergent red-blue 'union jack' lookup table in ImageJ.

It's not so easy to keep the image size steady when mousing around. It seems like 5-6 other years may have this salinity product though the source itself, Mercator Ocean, doesn't archive much. The next available depth was 100m, maybe too deep.

Note that new blue arrow bottom left will hide some of the extraneous overlays as does the old arrow upper left. It doesn't have the built-in tinyUrl of Worldview which lets readers jump right into a setup too complicated to describe.

There does not seem to be a salinity anomaly corresponding to the Laptev temperature hot spots. The salinity is very banded -- contours can be followed from the Yermak Plateau all the way to the New Siberian Islands.

Atlantic Waters form a topographically steered boundary current controlled by relative buoyant density. For that reason, the fine details of the new IBCAO bathymetry at the partial turn near the junction of Laptev shelf and the Lomonosov Ridge are worth a look, best done with a glasbey palette from the DEM because 'adjacent' grays are given maximally distinct colors.

The Polarstern made hundreds of CTD casts and set out profiling buoys. No salinity data has been released despite almost 14 months going by. However one of the research goals was better understanding of the return route of Atlantic Waters. It initially froze in just west of the Lomonosov Ridge but well north of the NSI at 85º and drifted west from there.

Meanwhile, other research ships have transited key areas on the salinity charts -- the whole Northeast Passage is undergoing intensive study. However scheduling has to be done well in advance so it is unlikely that this year's unique opportunity to study ice-free Laptev waters could generate a responsive outing.

We don't know to what extent any of this recent data has been assimilated into CMEMS and related products.

[S.Pansa]

I hope it hasn't been posted here before.

Yesterday Germans first national broadcaster ARD aired a 89min documentary about the Polarstern/MOSAiC expedition.

Well worth watching imho ... what a fascinating place. They even caught some Atlantic cod at their camp close to the New Siberian Islands (?).

Unfortunately though it is in German and the content is  geo-restricted; with a vpn & a German proxy it should work however.

[uniquorn]

Hopefully that will become viewable worldwide in time. 2023 maybe 

There does not seem to be a salinity anomaly corresponding to the Laptev temperature hot spots. The salinity is very banded -- contours can be followed from the Yermak Plateau all the way to the New Siberian Islands.

It's likely that a model would struggle with such local detail but as noted above, the 80N 140E polynya could just be coincidence.
gimp value invert used as a quick fix on the bathymetry map to improve visualisation of ice concentration over shallow water

[A-Team]

Looking at AMSR2_AWI for Nov 16th, a small patch of Laptev open water remains in addition to several 'lee polynyas' which occur downwind of the De Long islands Bennet and Zhokhov as well as the main NSI. Quite a bit of ice is still unconsolidated (below 100% concentration).

Ice thickness in the Laptev and ESS is almost entirely below 0.5m on Nov 15th. It replaces open water as seasonal tracking priority -- several weeks of thickening have been lost already in the Laptev. The SST 10m water column still has some warmth left even under surface ice that may continue too slow its thickening though wind mixing is gone and the Siberian air warmth anomaly is likely much more important.

Lee polynyas are seen all winter at Wrangel and other islands depending on wind direction; they are important to marine life but not indicative of freeze-up conditions. They are too small in extent to affect open water statistics though ideally should be subtracted off. AMSR2_UHH_large has more detail on the even smaller De Long islands (at 3x below); we don't know swath timing on either product, they could agree given it's from the same satellite.

[uniquorn]

A reminder about the four mosaic thermistor buoys on floe2

full 2020T78 drift path since aug8. A little burst of acceleration recently.
Temperature profiles of all 4 buoys since oct1
Static chart of T78 showing surface temp (~-0.2°C on nov12, -31C nov14) and a selection of ice core temps.

[A-Team]

The ice cover has finally extended over the entire Laptev so the monitoring focus must shift away from the open water anomaly to sea ice thickening (and lack thereof due to a very late start). The Chukchi-Bering Sea remains largely open water under Pacific Ocean influences as does the strip north of Sv-FJL-SZ affected by surfacing Gulf Stream waters.

For the Laptev, this means the daily netCDF of sea surface temperatures SSTfnd no longer has applicability. The best remaining tool is probably SMOS-SMAP ice thinness. That drops out when ice thickness exceeds 0.5m, at which point the hybrid Cryo2SMOS can step in.

SMOS-SMAP is posted with a two-day lag but archived back to 2015. The current season is shown below, restricted to a wedge needed for year-on-year comparisons without the overwhelming background variation from non-Laptev regions. Hardly any thick ice has formed yet south of 80ºN in stark contrast to Nov 17th in the five previous years.

In terms of quantitation, the tan thick ice can be isolated with a colorpicker in a tile-up, set to white with the inverse set to black. ImageJ then provides a ready graph of the growing fraction of >0.5m ice that can then be compared across years. (The same was done earlier for open water.)

Ice thickening in the Laptev is 2-3 weeks delayed; subsequent tracking can determine whether it will catch up or fall further behind. If the anomalous persistence and extent of Laptev open water in the fall of 2020 establishes as a trend, then an asymmetric ‘Siberian side’ scenario of fractional blue ocean event may be developing, in part attributable to increasingly dominant Atlantification (Polyakov 2020) which does not have the seasonal variability of Siberian weather.

Overall the NorthEast Passage seems to be opening earlier and closing later -- the new nuclear icebreakers may end up mainly serving tourism. The fabled NorthWest Passage remains an transiting dud.

[gandul]

If the anomalous persistence and extent of Laptev open water in the fall of 2020 establishes as a trend, then an asymmetric  ‘Siberian side’ scenario of fractional blue ocean event may be developing.

Maybe, although if one recalls weather patterns that vary year to year and decade to decade, with the only common distinguishable trend of warmer atmospheric temperatures, I don’t believe we are going to see this “asymmetry” in many years, like we have not seen the 2007 persistent Pacific dipole since 2007, the incredible 2012 Greenland blocking since 2012, and who knows when we’ll see the 2020 Siberian Warmth and GAAC like what we have seen in 2020, which undoubtedly have contributed to the Laptev refreeze delay.

[oren]

Looking at the SMOS animation (thanks A-Team) it appears 2019 had thinner ice than the previous years, so the trend is already two years back to back. I have no doubt the poor 2019 volume  growth contributed to what happened in 2020. At the end of the 2019/20 winter the Laptev volume (black line) was at an all time low, long before the Siberian heat wave and the GAAC.
Of course, the extremely late refreeze and delayed thickening doesn't necessarily mean the trend will continue, but it's certainly stacking the odds.

Click to enlarge.

[A-Team]

 

At the end of the 2019/20 winter the Laptev volume (black line) was at an all time low, long before the Siberian heat wave and the summer cyclone.

Right. Don't forget the 800kg oceanographic gorilla swimming in the Laptev below the meteorology! We have not seen much progress in getting people to actually read Polyakov 2020 despite it representing 15 years of in situ observational data describing the increasingly dominant atlantification influences at the top of the Laptev bathymetric bight.

It's free full text, the latest in a decades-long series rarely considered here, posted maybe a dozen times on various forums but to no effect. Atlantification pulses, being oceanography, have a certain inertial inevitability about them unlike wait-and-see seasonal variability of Siberian weather (that overlies the strong autumn Arctic Amplification warming trend).
 
Intensification of Near‐Surface Currents and Shear in the Eastern Arctic Ocean
IV Polyakov et al. Aug 2020
https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2020GL089469

'Measurements of currents from a 15‐year mooring record in the eastern European Basin demonstrate that the previously identified weakening of stratification in the halocline (Polyakov 2017, 2018) has been accompanied by increased upper‐ocean current speeds and associated current shear. Most of this increased energy and shear is in the semidiurnal band which includes baroclinic tides and wind‐driven inertial oscillations, with little change of mean along‐slope water transport (Pnyushkov 2018).

Increased shear together with weakening stratification indicate a greater potential for shear‐driven turbulent mixing, consistent with the recent transition in sea ice and upper ocean state to conditions previously unique to the western Nansen Basin [[Sv-FJL-SZ line]].

This increased coupling between AW heat and the sea ice may lead to a positive feedback between reduced sea ice and higher mixing rates as the longer periods and increased areal extent of open water facilitate more energetic wind‐driven inertial oscillations and associated upper‐ocean shear coinciding with weakening halocline stratification. As sea ice declines, a new Arctic state is emerging which, due to the positive feedback mechanism outlined above, may be pushing the system toward a tipping point.'

This article was submitted back in the spring of 2020 before any of this unprecedented season unfolded. It's aging well so far!

[A-Team]

So is there a fall Siberian Side or Laptev wedge trend, what do we even mean by trend, how many years of data does it take, what does normal variation mean in a rapidly changing Arctic climate, what can be considered statistical proof of a trend, is there a plausible physical basis for the trend, is the Laptev wedge a robust or biased regional choice, and how accurate is the guidance for future (ie next) years?

Oren raises an interesting complexity in #1266. On a daily coin flip, what happened in 2018 is irrelevant to 2019 which in turn is irrelevant to 2020: sampling with replacement from a fixed and familiar probability distribution. Coin flip ensembles average 50-50 but individual runs can drift off never to reach 50-50, the counterpart to natural variation.

Now suppose the ‘head’ of the coin wears down slightly during 2018 flipping process, biasing the flip towards ‘tails’ in 2019, with even worse wear-down biasing 2020 still more.  A slow runaway situation of positive feedback with earlier years becoming increasingly unrepresentative.

That may be the case for fall Laptev ice trending if each year strongly preconditions the ice — and indeed the weather — to the next. Trends are then largely self-induced (with ‘rebound' years perhaps anti-induced) rather than driven by generic global climate change, advancing atlantification, ocean-albedo effect and so on.

The figures below greatly compress 65 days the Siberian Side (upper half the Arctic Ocean in Greenland down orientation) and the Laptev-only wedge to make visualization patterns. There do indeed to be two trends: more open water earlier and a two week later shut-down. (Numerical treatments are given in earlier posts.)

Here if each year were cut into an unlabelled strip and a viewer asked to logically order them, there being 6!=720 top to bottom arrangements, it is not plausibly coincidental that the viewer’s order would agree with observed chronological order (2017 an outlier) unless the latter had a noticeable trend.

Smos-Smap provides six years of daily data back to 2015. Smos by itself goes back to Oct 2010 at U Bremen but is archived at an sub-satisfactory resolution, 202 vs 584 pixel width for the Laptev wedge. The figure below shows 2012 to the right of 2020 in the 6-year joint  mp4 (2020 2016; 2019 2018; 2017 2016 smos-smap; 2015 omitted).

The Smos-Smap combination is better, especially nearer to the 0.5m reporting limit (however it's not used with the CryoSat2 hybrid product!). Smos-Smap is examined below with Smos held back for later (hypothetical) testing and validation in earlier years like 2012 of presumed declining applicability.

A Consistent Combination of Brightness Temperatures from SMOS and SMAP over Polar Oceans for Sea Ice Applications
AU Schmitt L Kaleschke  30 March 2018  6 cites
https://www.mdpi.com/2072-4292/10/4/553/htm free full

Combined SMAP–SMOS thin sea ice thickness retrieval
C Patilea G Heygster M Huntemann G Spreen 2019 Feb 2019 13 cites
https://tc.copernicus.org/articles/13/675/2019/ free full

Assessment with Controlled In-Situ Data of the Dependence of L-Band Radiometry on Sea-Ice Thickness
P Sánchez-Gámez C Gabarro A Turiel M  Portabella  13 Feb 2020 0 cites yet
https://www.mdpi.com/2072-4292/12/4/650/htm  free full CFDD Upward Looking Sonar tests

https://seaice.uni-bremen.de/data/smos/png/
https://seaice.uni-bremen.de/data/smos_smap/png/north/

[uniquorn]

Going back a bit further. NOAA sea ice concentration, Barents on nov21, 1981-2020 compressed and animated.
82 and 85 had no data
Things look a bit better for nov22

[A-Team]

Ice is finally beginning to thicken in the Laptev wedge in late November — blocks of tan ice thicker than a half meter are beginning to appear on both Smos and Smos-Smap (black stars). These are one day and two days behind the current date, the latter because it combines data from two satellites.

The ice thicker than >0.5m will probably occupy the entire Laptev by the end of a week though in most years, thin pockets and lee polynyas persist for several weeks longer. (Note the Kara is still only half iced over and half of that is still thin.)

The trend in the Laptev wedge is shown in the attached figures. The 11 years shown are 2010-2020. The 65 days this fall are shown greatly compressed but this is still enough to see trends and outlier years. Quantitation can be done in ImageJ using 'plot Z-stack profile' command.

This year was exceptional for its open water extent, its persistence, and slowness to thicken. Given the positive feedbacks in play, it would come as no surprise to see the Siberian side show even more extreme autumns in the near future.

A fractional blue ocean event will likely be preceded by a one meter event (BOE 1.0m) for which a large part of the Arctic Ocean never freezes to a thickness greater than 1.0m per Cryo2Smos or the half-meter limit of the soil moisture satellites alone (BOE 0.5m), with ice not plausibly surviving in summer. On 01 Mar 2020, a fair amount of ice was only a meter thick but very little was under a half meter.

[A-Team]

Here is some additional analysis of the Laptev region. The first figure shows an all-in-one analysis of open water for the first 65 days past the Sep 15th minimum for 11 years back to 2010, the second looks at growth of ice to a half meter or more in Smos-Smap which has been very slow in 2020 with much of the Laptev still only 20-30 cm thick on Nov 23rd.

The third figure looks at ice less than one meter thick during late winter of 2020, with some still present at the start of melt season.
\
The final figure removes the very highest concentrations of ice in AMSR2_AWI  on 21 Nov 2020 to let the ice thickness below from Smos-Smap show through. Note that the tan color at lower right is likely a weather artifact in AMSR2; indeed selecting this color in Smos_Smap is a way of removing it.

[SimonF92]

Excellent visualisation. Just so I have my eye in here properly, the figures read from 2020 on left and each peak further to the right is 2020-1year?

[A-Team]

left to right, 2020 to 2010?

That's correct. These figures need to be redone when the Laptev settles down for the season; better labelling and quantitation can be done at that time. Right now, the open water part could be finalized but not the ice growth part. There is also cross-talk in some years from Chukchi open water extending into the East Siberian Sea west of Wrangel Island that doesn't really belong in the Laptev narrative.

These all-in-one charts are made from tile-ups of year x date (here 11 x 65), followed by setting open water to white / everything else to black, followed by de-tiling to 715 single frames in imageJ, applying the plot tool, coloring and quantitating with the color picker in gimp. It looks like the Laptev will need another two weeks of freeze-up to become comparable to previous years so 11x80 which is still quite doable at the lesser resolution 185 x 175 of Smos.

The imageJ graph has a fixed width which limits horizontal resolution but the accompanying  numerical table scores each frame individually and displays immediately in excel where zero cells could be inserted as separators between years and no limits restrict graph size.

This method gives a fabulous level of compression with excellent retention of key information and has many area / sensor applications (eg uniq's analysis above of the Barents). Comparing so many concatenated individual images visually is unworkable.

The Polarstern was east of the 110º meridian (in the Laptev wedge) from 25 Sep 2019 to 12 Jan 2020 measuring sea ice thickness and thickening along with many meteorological parameters. However this time was spent almost entirely at higher latitudes than the near-shelf emphasized here. When that data is eventually released, our understanding of fall 2019 -- and indeed Smos-Smap data interpretation -- may be considerably improved.

[josh-j]

I know this is a scientific discussion and I have no science to add so I'll be brief, but A-Team those "plot Z-stack profile" images, particularly the laptev wedge one, are worthy of display in a gallery somewhere.

An absolutely brilliant visualisation. Like a thinning forest of leaves. It brings the data to life!

[SimonF92]

I know this is a scientific discussion and I have no science to add so I'll be brief, but A-Team those "plot Z-stack profile" images, particularly the laptev wedge one, are worthy of display in a gallery somewhere.

An absolutely brilliant visualisation. Like a thinning forest of leaves. It brings the data to life!

Exactly, very very nice stuff. If only you had a bunch of funding like Zach Labe to host this stuff online. It is very good science.

I got some money from Amazon ($500) to host the MOSAiC Buoys stuff we were doing last year- but the costs ended up adding up and I had to purge the server. I would love to see your and uniquorns stuff hosted somewhere.

[Sepp]

Just out of curiosity, what are the actual requirements of such hosting regarding traffic, cpu power and bandwidth?

[SimonF92]

I hosted on AWS server which was a linux micro server with 1Gb RAM. The speed was ok but the killer was moving data around, depending on which kind of storage you use they can really hit you quite hard.

I aware of someone who used elastic storage (to an SSD) when they shouldnt have and they got hit with a $8000 bill.

Some of these images and datasets are probably quite big and would be updated regularly so it definitely would not be cheap.

There are also redirect costs to your domain which are cheap on a single basis but can increase exponentially depending on traffic- unless you want to advertise a dodgy looking ip address as your website .

If you want to try it out you can get a free 750 compute hours a month with Amazons basic account- this is actually enough to host a few images and datasets provided not too many people visit.

[uniquorn]

Very grateful to ASIF for hosting all our stuff.

cmems lobelia now create mp4 or gif. Tutorial here

This animation shows ice velocity (northward and eastward + and - ) and sea foundation temperature from oct1-nov24. The original gif size was 2090x1365px so the legend text and dates are a bit small.

Interesting that the full arctic ani shows movement in the mclure strait. Example from nov 17 https://go.nasa.gov/3m77FLL

[A-Team]

nice graphic!

These graphics are in the public domain -- anyone can propagate them anywhere! Mostly they just show up at google image and video searches so the file names are chosen to facilitate that.

so the legend text and dates are a bit small

Duplicate the image, paste the old legend over the too-small legend, resize layer, make new from visible?

The Laptev story can have higher resolution if the 'plot' numbers are put in Excel; the wind rose is a compact variation of that.

Looking now at the Kara on day 70 past the Sept 15th 2020 minimum, thick ice there is quite delayed but not extraordinarily so for the 11 years available for Smos-Smap ice thinness. It fits though with a hypothetical Siberian Side trend (Chukchi ESS Laptev Kara) for this year, perhaps largely attributable to anomalous warm air temperatures.

It's a lot of compression: 11 years * 108 days = 1188 frames = 538,933,824 pixels to  231,840 which is 2325:1. This becomes essential with multi-year time series to make the information in them assimilatable.

[A-Team]

The mp4 shows the Kara sea from Sep 15th to Dec 31st (108 days) for the years 2020-2010 (L to R, top to bottom, 2020 repeated with days 71-108 grayed as not arrived). It is quite interesting to see how the Kara Sea fills in, mainly from the land side (and by hundred of islands) but supplemented by growing the Arctic Ocean sea ice, with ice often forming just south of Severnaya Zemlya (a cold spot which had late relic ice this August).

Tracking open days of the NorthEast Passage shows it has been opening earlier and closing later over the last forty years.

SSTfnd 10m water column temperatures are just a degree or two above sea water freezing according to data from Nov 24th.

Water circulation patterns in the Kara Sea are complex. The Kara Sea tends to be sea ice covered between September and May,[7] and between May and August heavily influenced by freshwater run-off (roughly 1200 km3/yr from the Russian rivers (Ob, Yenisei, Pyasina, Pur, and Taz). The Kara Sea is also affected by the water inflow from the Barents Sea, which brings 0.6 Sv in August and 2.6 Sv in December. The advected water originates from the Atlantic, but it was cooled and mixed with freshwater in the Barents Sea before it reaches the Kara Sea. Simulations with the Hamburg shelf ocean model suggest that no typical water current pattern consists in the Kara Sea throughout the year. Depending on the freshwater run-off, the dominant wind patterns, and the sea ice formation, the water currents change.

Surface area   926,000 km2 (358,000 sq mi)
Average depth   131 m (430 ft)
Water volume   121,000 km3 (98×109 acre⋅ft)
Frozen   Practically all year round [[1999 wikipedia]]

[A-Team]

The Chukchi is, as usual, the last remaining large area of open water in the Arctic Ocean. Small areas remain open into late December, depending on Bering Sea influxes, wind mixing and Pacific weather system influences.

The first graphic shows sea surface temperatures SSTfnd up to 4ºC above the freezing point (which is slightly higher than -1.8º because of lower salinity). The mp4 provides the Smos-Smap ice thinness history for Sep 15th to Nov 24th.

Another use for Smos is a more nuanced definition of extent that avoids misleading influences from areas of overly thin ice. That's shown below in a still taken from @seaice.de.

A new paper also addresses some of the questions raised above as to what exactly drives Laptev freeze-up: 

Analyzing links between simulated Laptev Sea sea ice and atmospheric conditions over adjoining landmasses using causal-effect networks
Z Rehder et al   09 Oct 2020
https://tc.copernicus.org/articles/14/4201/2020/ free full

"We investigate how sea ice interacts with the atmosphere over adjacent landmasses in the Laptev Sea region as a step towards a better understanding of the connection between sea ice and permafrost. We identify physical mechanisms as well as local and large-scale drivers of sea-ice cover with a focus on one region with highly variable sea-ice cover and high sea-ice productivity: the Laptev Sea region. We analyze the output of a coupled ocean–sea-ice–atmosphere–hydrological-discharge model with two statistical methods. With the recently developed causal-effect networks we identify temporal links between different variables, while we use composites of high- and low-sea-ice-cover years to reveal spatial patterns and mean changes in variables.

We find that in the model local sea-ice cover is a driven rather than a driving variable. Springtime melt of sea ice in the Laptev Sea is mainly controlled by atmospheric large-scale circulation, mediated through meridional wind speed and ice export. During refreeze in fall thermodynamic variables and feedback mechanisms are important – sea-ice cover is interconnected with air temperature, thermal radiation and specific humidity. Though low sea-ice cover leads to an enhanced southward transport of heat and moisture throughout summer, links from sea-ice cover to the atmosphere over land are weak, and both sea ice in the Laptev Sea and the atmospheric conditions over the adjacent landmasses are mainly controlled by common external drivers."

[A-Team]

The mp4 below shows the 73 days of refreeze in the Laptev wedge since the Sep 15th minimum partitioned between disappearing open water (blue), ice of intermediate thinness (Smos-Smap <0.5m multi-hued palette) and encroaching ice thicker than a half meter (tan). Conceptually this is reminiscent of NSIDC’s sea age graphic in which the older ice classes are getting pinched out by first and second year ice. Here open water has been pinched out by newly formed ice which in turn has thickened to the point where it is being pinched out by the 0.5m thick ice category.

Going by the last twelve days and a linear regression fit, it appears that a complete thick ice cover in the Laptev is still a few weeks off. The initial thin ice cover insulates bottom water, where growth occurs, from the cold atmosphere; wind mixing has ceased but turbulent mixing can still occur, with SSTfnd data still showing the 10m water column to be a degree or two above the -1.8ºC freezing point.

The freeze in the Laptev is considerably slower than in recent years. The offset ranges roughly  from 19 to 35 days based on selecting the most similar distribution of >0.5 m ice in previous years. (These years had different patterns so the comparison is inexact.)

[A-Team]

Lobelia has added flexible gif and mp4 time series generation downloads to the CMEMS ocean page linked to by a 'social media' icon next to 'add layers' that also allows iframe embedding (not supported here), tinyUrls and static pngs.

Uniq has already posted excellent examples on the main and test forums. Lobelia-CMEMS is a very important new resource for the Arctic Ocean but over the last six weeks only 2-3 of the 1795 site registrants have used it. As is said about the lottery, you can't win if you don't play.

Below, grayscale palette gif output from Lobelia has been run through ImageJ, Gimp and CloudConvert to add some graphic refinements, notably for freedom to change palette to an arbitrary LUT, move the date to a better place, recolor ocean blue, crop, rotate to Greenland down, overlay the Polarstern drift path and so on.

The mp4 shows 84 days of ice thickness as modeled by nextSim from Sept 15th out to a predicted Dec 7th. The scale has been set to range over 0-3m which has the effect of not displaying thinner ice optimally; that would require a scale setting of 0-0.5m which would not display the main ice pack at all.

It's not completely clear what the 'floes' and 'leads' actually represent down on the ice as these features are below the resolution of any of the satellite tools used by nextSim. Whatever, they do seem to allow an accurate depiction of ice movement, notably the pick-up in Fram export in mid-November.

The second mp4 shows the full range of 111 weeks contained in the archive. It is a little jumpy but the daily scale would make for quite a large file and the hourly would be way out of bounds.

The third mp4 shows ice thickness for the full year of the Mosaic expedition restricted to the Svalbard area to emphasize the vigorous TransPolar Drift and Fram export compared to the virtual lack so far this season. The file size is still quite small meaning 2x the dimensions would be about 8 MB.

These mp4 were initially made as gifs where resolution is better (because mp4 involves lossy compression). After loading and study of most effective frame rate in ImageJ, 'hourly' can easily be changed to 'six hourly' or daily' can  to 'every other day' etc. This avoids choice limitations in the Lobelia panel. However, ordering more frames slows down product delivery which is a lot slower now than Nasa's WorldView.

"The Arctic Sea Ice Analysis and Forecast system uses the neXtSIM stand-alone sea ice model running the Maxwell-Elasto-Brittle sea ice rheology on an adaptive triangular mesh of 10 km average cell length. The model is available back to 01 Nov 2018. 

neXtSIM uses surface atmosphere forcings from the ECMWF (European Centre for Medium-Range Weather Forecasts) and ocean forcings from TOPAZ4, the ARC MFC PHY NRT system (002_001a). neXtSIM runs daily, assimilating OSI-SAF sea ice concentrations (both SSMI and AMSR2) from the SI TAC and providing 7-day forecasts.

The output variables are the ice concentrations, ice thickness, ice drift velocity and snow depths, provided at hourly frequency. The adaptive Lagrangian mesh is interpolated for convenience on a 3 km resolution regular grid in a Polar Stereographic projection. The projection is identical to other ARC MFC products."

https://tinyurl.com/yxfwhpkb

A Maxwell elasto-brittle rheology for sea ice modelling
V Dansereau et al 01 Jul 2016
https://tc.copernicus.org/articles/10/1339/2016/tc-10-1339-2016.pdf free full text
https://en.wikipedia.org/wiki/Dashpot Maxwell 1867 key feature of viscoelastic model

A new rheological model is developed that builds on an elasto-brittle (EB) framework used for sea ice and rock mechanics, with the intent of representing both the small elastic deformations associated with fracturing processes and the larger deformations occurring along the faults/leads once the material is highly damaged and fragmented. A viscous-like relaxation term is added to the linear-elastic constitutive law together with an effective viscosity that evolves according to the local level of damage of the material, like its elastic modulus.

The coupling between the level of damage and both mechanical parameters is such that within an undamaged ice cover the viscosity is infinitely large and deformations are strictly elastic, while along highly damaged zones the elastic modulus vanishes and most of the stress is dissipated through permanent deformations. A healing mechanism is also introduced, counterbalancing the effects of damaging over large timescales.

In this new model, named Maxwell-EB after the Maxwell rheology, the irreversible and reversible deformations are solved for simultaneously; hence drift velocities are defined naturally. First idealized simulations without advection show that the model reproduces the main characteristics of sea ice mechanics and deformation: strain localization, anisotropy, intermittency and associated scaling laws.

The availability of ice buoy and satellite data has allowed three all-important characteristics of the deformation of sea ice to be revealed: its strong localization in space (heterogeneity), its localization in time (intermittency) and its anisotropy.

The anisotropic nature of sea ice deformation is made evident by the analysis of satellite-imagery derived ice motion products which shows that high strain rates concentrate along oriented, linear-like faults, or leads, often termed “linear kinematic features” (Kwok, 2001). The signature of the strong heterogeneity and intermittency of sea ice deformation is the emergence of spatial and temporal scalings in the deformation fields over a wide range of scales.

[uniquorn]

Nice animations and we have a good ice model there that looks a lot like ascat without the weather interference. Must try an overlay sometime to see how they match up.
They also plan to use buoy data.
from https://tc.copernicus.org/articles/10/1339/2016/tc-10-1339-2016.pdf

Further validation of the Maxwell-EB framework and the determination of the range of model parameters values suitable for sea ice call for a thorough comparison of the scaling properties of the simulated deformation rates with that estimated from the available ice buoy and RGPS data. Such analysis necessitates carrying out numerical experiments over periods of several days to months and over realistic domains of  regional  to  global  scales.  At  these  spatial  and  temporal  scales,  deformations  within  the  sea  ice  cover  become large;  hence  advective  processes  cannot  be  neglected.

Mosaic buoy drift update using iabp data, nov15-dec1

[gerontocrat]

The last mp4 from A-team showimg the process of ice export to the Greenland sea.
Gives a real insight into the process. Wow

Being greedy, wouldn't it be great to have it extended south to show the growth of the Greenland sea ice in winter followed by the death of the sea ice in the following spring and summer?

______________________________________________________
ps: The colours make make the ice look like some evil blob heaving and threshing about in its death throes.

[Glen Koehler]

<snip>
  "It's not completely clear what the 'floes' and 'leads' actually represent down on the ice as these features are below the resolution of any of the satellite tools used by nextSim. Whatever, they do seem to allow an accurate depiction of ice movement, notably the pick-up in Fram export in mid-November."

and <snip>
    "The anisotropic nature of sea ice deformation is made evident by the analysis of satellite-imagery derived ice motion products which shows that high strain rates concentrate along oriented, linear-like faults, or leads, often termed “linear kinematic features” (Kwok, 2001). The signature of the strong heterogeneity and intermittency of sea ice deformation is the emergence of spatial and temporal scalings in the deformation fields over a wide range of scales."

    Nice work A-Team.  The fracturing of the ice pack depicted in the B&W mp4 is stunning.  How close to reality is that imagery, or are the dark lines and pockets exaggerated by model processing of the input data?

   The Sept 19, 2020 image is bad enough, but the Oct 3 image is even more striking.  While there is some consolidation just north of the CAA and Greenland between Sept. and Oct., a large section of the Atlantic - western Siberian front (accounting for ~10-20% of the total remaining ice, and extending almost to the North Pole) in October looks even more broken up and disbursed than in September, and even closer to being obliterated.  That makes me wonder if there had been one more push over the edge, 2020 would have quickly been at or below the 2012 record low minimum Extent, Area and Volume.

   Either way, that imagery highlights the frailty of the ASI condition in Sept. and Oct. more dramatically than the numerical charts and even the charts showing loss of MYI. 

[A-Team]

extend animation south to show the growth of the Greenland sea ice in winter followed by loss in the following spring and summer?

It's probably better to use AMSR2_UHH_large for this (been done but where) as it is observational vs a model of a difficult area to model accurately. We don't want to be producing eye candy per se.

How close to reality is that imagery, or are the dark lines and pockets exaggerated by model processing of the input data?

It is not what you would see out an airplane window but nextSIM does seem to be producing scenes that are descriptive of ongoing ice physics. Uniq notes above that correspondence to satellite infrared brightness leads is underwhelming. Enhanced Ascat does support certain aspects of the nextSim animations above, notably the surge in Fram export but leads and floes don't match very persuasively (fig.3).

The 96 hours of animated contoured GFS surface temperatures supports the notion of a 'cold pole' on the Canadian side. The common error here is to think the North Pole is the cold pole. It is not.

This could be alternatively phrased, since warm/cold are relative, as the Siberian side being consistently warmer this season, resulting in delayed freeze-up of the Laptev, ESS and Kara. This asymmetry can often be seen on anomaly graphics; Arctic Amplification is not affecting the Arctic Ocean uniformly.

The NP is a fixed surface point defined by the intersection of the earth's rotational axis with the surface. There's not an analogous geophysical concept defining the 'cold pole'; it is regional, shifting from day to day though consistently associated with the CAA. The term has not been used to any extent in journal articles.

https://www.tandfonline.com/doi/abs/10.1080/00385417.1960.10769814?    1958

[uniquorn]

First impression of nextSIM is that new ice looks a lot thinner than the representation by both ascat and amsr2. It all depends on the scale though.

Two first hand accounts of the mosaic expedition. Basic, but includes a lot of photos not seen on this thread before.

Antonia Immerz of the Alfred Wegener Institute is a data scientist for the MOSAiC expedition. Join her for a live video call on December 2nd where she'll talk about her experience working in the Arctic and what's going to happen to all of the data collected during MOSAiC.

Jackson Osborn is a scientist at the Cooperative Institute for Research in Environmental Sciences at the University of Colorado Boulder who served as an engineer for Team Atmosphere during MOSAiC. Join Jackson for a live video call to learn what it's like to troubleshoot scientific instruments in the Arctic.

[uniquorn]

Well we have another chance to develop our thickness estimate algorithm soon. Core ice temperature is dropping to levels where we thought we had some reasonable estimates last year. I have doubts though.
Looking at T78 data the deployment report (1) states the ice was 1.52m thick with 2cm of fresh snow on aug23. Thermistor buoy sensors are 2cm apart so we can't really detect snow at that point but we are also given the sensor id of the air/snow interface as 33. Great. So the bottom of the ice on that day was at sensor 110.
The temperature profile then was almost flat by winter standards but a more detailed look shows the ice thickness quite well. (2)
Thermistors 33 and 110 have been marked roughly at the beginning of the animation(3) and again on nov22 when the temperature profile shows them clearly (with a short pause both times). There is not much additional thickening since then.
So my doubt is this. Did the ice melt quickly during end of august/september or has it been 1.5m thick and possibly porous all the time but took 3 months to cool down again?

added T78 drift path(4)

[oren]

Well we have another chance to develop our thickness estimate algorithm soon. Core ice temperature is dropping to levels where we thought we had some reasonable estimates last year. I have doubts though.
Looking at T78 data the deployment report (1) states the ice was 1.52m thick with 2cm of fresh snow on aug23. Thermistor buoy sensors are 2cm apart so we can't really detect snow at that point but we are also given the sensor id of the air/snow interface as 33. Great. So the bottom of the ice on that day was at sensor 110.
The temperature profile then was almost flat by winter standards but a more detailed look shows the ice thickness quite well. (2)
Thermistors 33 and 110 have been marked roughly at the beginning of the animation(3) and again on nov22 when the temperature profile shows them clearly (with a short pause both times). There is not much additional thickening since then.
So my doubt is this. Did the ice melt quickly during end of august/september or has it been 1.5m thick and possibly porous all the time but took 3 months to cool down again?

added T78 drift path(4)

It's nice to have the actual thermistor air-snow and ice-water numbers at deployment.
Looking at the animation, it appears like there was massive bottom melt, and then re-thickening.
This is what I would expect in general with the given temperature profile.
At first, ice top is warmer than at the bottom. This surely brings about bottom melt, as heat trickles from above and the salt water eats away from below, and I think would also depend on the rate of drag of the ice over the water.
The image from Aug 23 clearly shows ice-water interface at thermistor 110.
Next, the temp profile is flat. My understanding is that the bottom would still be melting, albeit at a much slower pace.
The image from Sep 13 seems to be the last day of top melt, as well as the ice having a cold core again (cold gradient from the middle to the bottom).
Eventually, the ice top is colder than the ice bottom, with a gradient appearing through the ice. From this point on the ice starts bottom freezing and it's easier to spot where the ice ends.
The image from Sep 21 in my opinion shows the ice-water interface at around thermistor 55.
The image from Oct 15 seems to show the ice-water interface at thermistor 70.
The end of the animation again shows the ice to exceed thermistor 110.

Admittedly these are just generalizations, since I can't put any quantitative expectations as to the rates of melting and freezing. The rates do depend on the steepness of the warm or cold gradient though. Therefore I would not expect a whole lot of bottom melting, also given the rather late date of deployment. However, eyeing the animation, it would appear as if the bottom melted all the way to thermistor 55. Some possible explanations:
* Bottom was higher than thermistor 110 to begin with, with 110 being some protruding edge. However, the animation data does support this initial placement for the ice-water interface.
* Melting did not actually reach thermistor 55, though I can't see any other explanation for the cold gradient ending where it ends on Sep 23, except that it was the ice-water edge.
* The ice was extremely mobile which enhanced bottom melt way beyond what the temp profile would suggest. This can be checked by looking at T78 drift.
* The ice has not cooled enough for the gradient to reach the ice-water interface.
* My intuition for rates of bottom melt is disconnected from reality (very probable).

All in all, a very interesting mystery, with wider implications. If indeed ice can bottom-melt so quickly at the end of the season, we are not so far away from blue ocean as the extent extrapolations seem to suggest.

[uniquorn]

T84 also has additional deployment data but unfortunately a photo is not available. Thermistor 26 at surface, ice thickness 1.14m, Snow depth 0.03 m, so bottom of the ice at thermistor84. Similar analysis below with a closer look at temperatures near to -1.8C.

It's possible that drilling the deployment hole affects the results, allowing faster local melt that refreezes over time.

[SimonF92]

Slowly but surely getting the algorithm working to be more accurate for thin ice. Hopefully this reflects uniquorn's and oren's interpretation of the R analysis

[uniquorn]

My own intuition agrees with Oren and the algorithm that roughly -1.8C marks the ocean/ice boundary but the data from T85 suggests otherwise. Here we have 4m thick ice (or thicker) where thermistors 1 to 10 are clearly above surface so all or nearly all of the remaining sensors are in ice.
It's possible that thermistors 210-240 are in the ocean but look at the temps. Flat line from thermistor 140 onwards. The thick ice is still cooling.

T85 may be on centre right on the photo. It looks like there is a camera on the battery unit.

[SimonF92]

My own intuition agrees with Oren and the algorithm that roughly -1.8C marks the ocean/ice boundary but the data from T85 suggests otherwise. Here we have 4m thick ice (or thicker) where thermistors 1 to 10 are clearly above surface so all or nearly all of the remaining sensors are in ice.
It's possible that thermistors 210-240 are in the ocean but look at the temps. Flat line from thermistor 140 onwards. The thick ice is still cooling.

T85 may be on centre right on the photo. It looks like there is a camera on the battery unit.

For the time being, instead of having a set value of -1.8 I just set ice bottom as the thermistor below which the temp is 0.2degC less than whatever the ocean temp is at that measurement time.

''
ice_thermistors=ice_or_water_thermistors.where(ice_or_water_thermistors<mean_ocean_heat-0.2)
''

We can talk about this somewhere else if need be so as not to detract from the good posts on this thread.

Also, 4metre thick ice? Didnt think there was any of that left.

[uniquorn]

4m thick ice?

T85 was "deployed in ridge"

I couldn't find any ridge thickness references for floe2 but there is this one for floe1
https://www.carbonbrief.org/inside-mosaic-how-a-year-long-arctic-expedition-is-helping-climate-science

Three days into the search, the researchers strike lucky. The team from Polarstern have found a floe with an unusually thick centrepiece, says Rex:
“The floe we’ve selected has a core of very compressed and thick ice. Thickness in that area of the floe – which we’ve decided to call ‘the fortress’ – varies between four to five metres in many places and 1.5m in other places.”

Better late than never here is a map for floe1 on july27. Guessing that the whiter parts are ridges since, as ever, there is no scale. Click for full res and ALS scan, jun30, without map.

[A-Team]

why have a Mosaic forum the Polarstern expedition is over

Best not to conflate the Polarstern (platform, no forum) with Mosaic (Arctic science, active forum). To date we have seen only 3 of an expected 60 journal articles (5%). The ship may be in port but its deployed buoys are still actively reporting on the the 2020 freeze season as are better-interpretable satellite products.

One rationale for model and forecasting papers is predicting better routes for research icebreakers and increasing commercial traffic. The last thing a 2m ice-capable icebreaker (or support ship) wants to do is break its way through 2m ice. It's jarring, fuel consuming and even icebreakers get stuck.

The Polarstern had 5k of bow radar and two helicopters to scout leads plus proprietary satellite radar, SIDFEx (Sea Ice Drift Forecast Experiment https://tinyurl.com/y3fg7bgk) and various flawed ice thickness and concentration products. However the ice encountered was a lot thinner/weaker than expected in reaching the Oct 2019 Laptev starting floe and later en route to the North Pole from the Fram. neXtSIM did not engage with MOSAiC planners via SIDFEx (which includes TOPAZ4) even though it dates to 2016.

Forecast tools such as SIDFEx and neXtSim-F (above CMEMS graphics) assert ship traffic safety is a paramount justification for funding. However Arctic Ocean accidents to date have had nothing to do with sea ice. The underpowered NorthGuider had two trawl nets out despite high winds, strong tides and complete darkness in late December and could not control its near-shore position in the ice-free Hinlopen Strait. The cruise ship groundings have all been attributed to inadequate charts, not collision with floes.

It's really questionable that an increasingly ice-free Arctic Ocean will create -- given the connection with collapse of civilization -- a glorious bonanza of resource extraction and low prices for EU consumers of Chinese goods. (Twelve thousand shoe boxes fit in a 20' container costing twelve hundred dollars to ship so 50¢ on a hundred dollar product with perhaps a 5¢ saving that won't be passed on with a trans-polar route.)

The forecast tools are very limited by their need for daily assimilation of weather forecasts, notably the wind stress field over a vast un-instrumented ocean. How plausible are 7-10 day outlooks when 1-3 days is already a stretch?

Over seasonal time scales, on matters such as TransPolar Drift, there appears to be a complete lack of forecast skill. That greatly affected Mosaic because the selected floe drifted far too fast; had they started farther east and north, the plan might have worked out.

As an illustration, three mp4 below (AMSR2_UHH, Smos-Smap, Ascat) compares TransPolar Drift visualizations over the 62 days since the Polarstern left high latitude (79.0/9.6) on 04 Oct 20 with the same dates from 2019 (shown reflected). What will the next six months of TPD look like? Even given all the initialization and historic data, no one has the slightest idea.

=/=/=/=

Where does the dramatic CMEMS imagery above, generated by neXtSIM, actually come from, what all is assimilated from daily satellite and weather products, how do they get to 3km resolution from much worse imagery, can we independently verify at least some of the very detailed ice surface features with say infrared lead imagery from WorldView, do leads and thickness even matter if it accurately shows ice movement?

Note CMEMS presents a stack of neXtSim, including ice velocity components (not displayed as vector arrows) and surface snow thickness, in addition to the main ridging/lead imagery.

The place to start is with a paper prematurely submitted to The Cryosphere on June 25th of 2019. Two helpful and detailed reviews were posted on August 23rd and 30th of that year.  However the review is still in limbo.

A greatly improved manuscript is discussed in the responses to peer review but isn’t available even on researchgate pages; the June 2019 mss is still posted. No editorial decision has been made about publication. It seems that CMEMS (aka Mercator Ocean) prematurely released the product -- with extravagant claims implications -- in Oct 2020 ahead of adequate documentation.

-1- A new sea ice rheology (Maxwell Elastic Brittle) respecting the statistical properties of sea ice deformations namely localization and intermittency.
-2- A Lagrangian adaptive mesh reduces numerical diffusion and enables sharp deformation features smaller than the 10km mean mesh length, allowing 3km resolution.
-3- A more advanced sea ice thermodynamic model following Winton 2010.

The following drawbacks should be noted:

-4 -The new product does not provides ocean current, temperature or salinity.
-5- The Canadian Archipelago, Baffin Bay and Hudson Bay are not modeled.
-6- Sea ice thickness and sea ice drift observations are not assimilated (as with Hycom) leading to understated summer Fram Strait export.
-7- There is no reanalysis product for 1993–2019 despite coverage by this product
[-8- Wave interaction with the MIZ is not included though written up separate preprint]

https://tc.copernicus.org/preprints/tc-2019-154/

[uniquorn]

Informative comparisons.

1. Here with 2020T60 we have a heroic deployment in 6.1m ice on a ridge crest on jan8 2020 in temperatures of -26C or lower. In this case using a 10m sensor chain with the 251 sensors spaced 4cm apart. As shown in the photo(2), thermistor1 is 1m above the ice with 10cm of snow so the ice bottom is roughly thermistor177.

3. The static chart shows temperatures from jan8-11. I think they pour water down the deployment hole and it looks like it takes a few days for that to cool to surrounding ice temperature. Maybe the water only gets as far as ~thermistor45 before it freezes.

4. In the animation, top and bottom of the ice have been marked as before. Once again the ice is at -1.8C well before the ice/ocean interface. I think there was a surface flooding event on mar19 followed by a more catastrophic ridging event at depth on apr19

 

[jdallen]

<lots of snippage>

<snippage>

The following drawbacks should be noted:

-4 -The new product does not provides ocean current, temperature or salinity.
-5- The Canadian Archipelago, Baffin Bay and Hudson Bay are not modeled.
-6- Sea ice thickness and sea ice drift observations are not assimilated (as with Hycom) leading to understated summer Fram Strait export.
-7- There is no reanalysis product for 1993–2019 despite coverage by this product
[-8- Wave interaction with the MIZ is not included though written up separate preprint]

https://tc.copernicus.org/preprints/tc-2019-154/

That's a lot of limitations.  I hope they can do some reanalysis with the '93-'19 data.  That would be pretty critical to understanding the limitations of the model, and help assess its accuracy and utility.

[A-Team]

”This revision took quite some time as problems with the sea ice model surfaced
resulting in 2 major updates to the rheology, and in addition we needed to provide two
releases of the forecast to CMEMS. However, we have now been able to finish it — T Williams et al 7 Nov 2020”

So will the original paper be withdrawn? The (unavailable) revised manuscript would have to be peer-reviewed from scratch, despite many initial review suggestions adopted (to be reviewed, next post). This leaves CMEMS hosting imagery documented by an unreleased preprint. Without the daily dramatic imagery provided by CMEMS/Lobelia, this model would just be another tree falling silently in a forest of forgotten sea ice models. Why now, Maxwell Elastic Brittle after all these decades of viscous plastic?

https://link.springer.com/article/10.1007/s40641-020-00162-y/tables/2 long table of model acronyms with links

From the perspective of our forums, there has to be quality assurance at the source. If unreliable data is hosted for one variable, is it a one-off situation or pervasive site-wide? We’ve seen too much ice thickness and snowpack eye candy over the years with no (or no demonstrated) underlying reality. It is just not realistic to expect 1800 members to individually evaluate dozens of technical continuum mechanics and adaptive mesh numerical analysis papers. More likely, they will be taken in by fancy graphics and animation, assume here CMEMS has done its job, and possibly misinform themselves and others from there.

https://psl.noaa.gov/forecasts/seaice/ CAFS RASM POP2 CICE5.1 CLM4.5 CESM GEFS AMSR2
https://www7320.nrlssc.navy.mil/GLBhycomcice1-12/navo/arcticictn_nowcast_anim30d.gif Hycom
https://tinyurl.com/y3zcaqpq CMEMS hosting stack of 5 neXtSim products: snow, thickness, x, y velocities

Should Sea-Ice Modeling Tools Designed for Climate Research Be Used for Short-Term Forecasting?
E Hunke et al  26 September 2020
https://link.springer.com/article/10.1007/s40641-020-00162-y very readable review

In regards to the five neXtsim products (https://tinyurl.com/yxm847ye), Lobelia has not yet added vector display as requested @lavergneth; separate display of x and y components of modelled sea ice velocity by color is utterly ineffectual. OsiSaf has long offered two-day arrows and is currently looking into swath-to-swath orbitals which provide many more AI matches and a hope for shorter term ice motion display as well as auto arrow-artifact removal. (However it is mostly pitched at a 2028 satellite launch; the long lead time is incompatible with rapidity of Arctic sea ice change.) Neither CMEMS nor OsiSaf provide Panoply-compliant netCDFs (in which x,y arrows can be drawn).

Towards a swath-to-swath sea-ice drift product for the Copernicus IMR mission
T Lavergne et al
https://tc.copernicus.org/preprints/tc-2020-332/tc-2020-332.pdf

The snow product is fascinating but where does it come from, how well does it agree with ice buoys especially designed to measure it? Here it is necessary to text-search the full succession of neXtSim papers to see if snow depth is observation-based altimetry or just some funky pre-reanalysis weather precip forecast. The attached graphic shows its Dec 6th thickness semi-transparently over the puzzling sea ice thickness leads. [Note Lobelia does not dim the palette nor show % transparency.]

Snow is not a well-defined term: on sea ice, it evolves through many states according to its initial properties and subsequent history. Mosaic could only document five snowfalls over 12 months -- and that snow blew around, mostly piling up on the lee side of pressure ridges.

Like all CMEMS products, the neXtSIM daily releases are shown in EPSG 3408 [EPSG 6931 intended?], the NSIDC EASE-Grid 2.0 GeoTiff-compatible WGS84 ellipsoidal earth Lambert azimuthal equal area projection (EA), using the 0º meridian rather than the conformal polar stereographic EPSG 3413 projection used by WorldView and almost all satellite imagery in ’Greenland down’ orientation of the -45º meridian. (A rotation would degrade image quality.)

This is unfortunate because these two projections cannot be readily rescaled to match, meaning satellite imagery cannot be layered onto CMEMS products without passing through netCDF (not offered) and Panoply. While it is possible to invert a satellite image in PS coordinates back to how it looked on the curved earth before projection and then then re-project that back down to the equal area plane, that would considerably degrade image quality. It’s further unclear whether the net radial distortion is already implemented in Gimp by the ‘lens refractive index’ setting in the Filter —> Distort menu.

However the meridian in CMEMS could easily be changed and an area scale added to facilitate both approximate overlay rescaling and accurate conversion from pixel counts to accurate  surface area, one of the main reasons for using an equal-area projection. Smap is apparently served in EASE 2.0 though that is nowhere stated and pngs are provided rather than GeoTiffs (unlike the more modern AMSR2_AWI). There does not seem to be an open source online tool to interconvert PS and EA.

https://www.mdpi.com/2220-9964/1/1/32/htm

neXtSIM papers (reverse chronological order):

Presentation and evaluation of the Arctic sea ice forecasting system neXtSIM-F
T Williams A Korosov P Rampal E Olason    25 Jun 2019 submission
https://tc.copernicus.org/preprints/tc-2019-154/
https://ui.adsabs.harvard.edu/abs/2020EGUGA..22.9136W/abstract

“We assimilate OSISAF SSMI and AMSR2 sea ice concentration products and the SMOS sea ice thickness product by modifying the initial conditions daily and adding a compensating heat flux to prevent removed ice growing back too quickly. We present an evaluation of the platform over the period from November 2018 to present, looking at sea ice drift and concentration and extent, and thin ice thickness.”

Probabilistic forecasts of sea ice trajectories in the Arctic: impact of uncertainties in surface wind and ice cohesion
S Cheng, A Aydoğdu, P Rampal A Carrassi L Bertino   17 Nov 2020
https://arxiv.org/pdf/2009.04881.pdf

“We suggest that in order to get enough uncertainties in a sea ice model with brittle-like rheologies to predict sea ice drift and
trajectories, one should consider using ensemble-based simulations where at least wind forcing and sea ice cohesion are perturbed.”

On the statistical properties of sea ice lead fraction and heat fluxes in the Arctic
E Olason P Rampal V Dansereau January 2020 preprint
https://tc.copernicus.org/preprints/tc-2020-13/tc-2020-13.pdf

On the multi-fractal scaling properties of sea ice deformation
P Rampall V Dansereau et al 2019
https://tc.copernicus.org/articles/13/2457/2019/

Impact of rheology on probabilistic forecasts of sea ice trajectories: application for search and rescue operations in the Arctic
M Rabatel P Rampal et al   2018
https://tc.copernicus.org/articles/12/935/2018/

Parallel implementation of a Lagrangian-based model on an adaptive mesh in C++: Application to sea-ice
A Samaké P Rampal S Bouillon E Olason 2017
https://www.sciencedirect.com/science/article/pii/S0021999117306368

Wave–ice interactions in the neXtSIM sea-ice model
T Williams P Rampal S Bouillon   2017
https://d-nb.info/1142799980/34

Probabilistic forecast using a Lagrangian sea ice model: application for search and rescue operations
M Rabatel P Rampal et al 2017
https://pdfs.semanticscholar.org/f7fc/b5551b8a352e271999ff736c5674b2d3afa9.pdf

Ice bridges and ridges in the Maxwell-EB sea ice rheology
V Dansereau J Weiss P Saramito et al   2017
https://tc.copernicus.org/articles/11/2033/2017/tc-11-2033-2017.pdf

neXtSIM: a new Lagrangian sea ice model
P Rampal S Bouillon E Olason M Morlighem  2016
https://tc.copernicus.org/articles/10/1055/2016/tc-10-1055-2016.pdf
https://tinyurl.com/yyz2xgwj large meeting poster

Sea ice diffusion in the Arctic ice pack: a comparison between observed buoy trajectories and the neXtSIM and TOPAZ-CICE sea ice models
P Rampal S Bouillon J Bergh E Olason 2016
https://tinyurl.com/y3vx5bmg

A Maxwell elasto-brittle rheology for sea ice modelling
V Dansereau J Weiss et al 2016
https://tc.copernicus.org/articles/10/1339/2016/tc-10-1339-2016.pdf

Presentation of the dynamical core of neXtSIM, a new sea ice model
S Bouillon P Rampal 2015
https://www.sciencedirect.com/science/article/abs/pii/S1463500315000694

Scaling properties of sea ice deformation from buoy dispersion analysis
P Rampal J Weiss et al    2008
https://doi.org/10.1029/2007JC004143