MOSAiC news

[Glen Koehler]

    +1 Ditto, thanks for addressing the thickness question A-Team.

     I just looked at a bunch of August 18-19 images for thickness and concentration (HYCOMM, Bremen, Hamburg, NSIDC etc.).  Many of the images have a pole hole so not exact value for 90N, but the collection overall shows rather striking difference between what is suggested by the various concentration and thickness images and what is shown in the North Pole photo by MOSAIC and their description of ice conditions during the trip to 90N.  The NSIDC sea ice concentration map appears to most closely match the MOSAIC ground truthing observations.

     UCMiami's comments about new ASI conditions creating a need to recalibrate or reinterpret established ASI observation methods seems spot on.

  <snip>   I feel that the last fifteen years have truly changed the nature of arctic sea ice, but a lot of the systems and analysis was established as 'fact' before that change really manifested and to some degree it has yet to adjust.

<snip> "...measures (Piomas and others) are grounded in a 'solid pack' view of arctic ice and I believe struggle to deal with the 'real world' condition of the pack where 'thick ice' is actually a patchwork of loose flows held together by new and thin ice. Images from Polarstern seem to make this abundantly clear."

[miki]

     UCMiami's comments about new ASI conditions creating a need to recalibrate or reinterpret established ASI observation methods seems spot on.

UCMiami's post. Link:
https://forum.arctic-sea-ice.net/index.php/topic,3017.msg282116.html#msg282116

[jdallen]

    +1 Ditto, thanks for addressing the thickness question A-Team.
<snip>
     UCMiami's comments about new ASI conditions creating a need to recalibrate or reinterpret established ASI observation methods seems spot on.

  <snip>   I feel that the last fifteen years have truly changed the nature of arctic sea ice, but a lot of the systems and analysis was established as 'fact' before that change really manifested and to some degree it has yet to adjust.

<snip> "...measures (Piomas and others) are grounded in a 'solid pack' view of arctic ice and I believe struggle to deal with the 'real world' condition of the pack where 'thick ice' is actually a patchwork of loose flows held together by new and thin ice. Images from Polarstern seem to make this abundantly clear."

<wry look>  I tried to provoke a discussion along those lines in the Melt Season thread, but was pretty much told I didn't know what I was talking about.

I think it would be a very good idea, as a lot is happening which categorically *isn't* being captured in our traditional metrics.

[uniquorn]

bow radar on the way north, aug15-19. The last image was very blurred for some reason so it has been removed.
PS now headed along 100E

[A-Team]

tried a discussion in the Melt Season thread ...a lot is happening which categorically *isn't* being captured in our traditional metrics.

That thread has been hijacked by legacy extent trackers, 10-day wx disaster forecasters, and self-nominated Nobel laureates (for seasonal weather prediction, ice physics outcomes six months out, even 2022 melt season finale.)

a 'solid pack' view of arctic ice struggles with 'real world' conditions

Looking at Hycom ice thickness at the north pole from July 28th through the Aug 19th photo forward to the Aug 28th forecast, the ice is shown having a fairly narrow range of thickness, notably 0.75 m at the time the Polarstern was there. It continues to thin moderately over the forecast week but part of that is just ice pack motion bringing thinner ice in. Click to see daily thickness.

Piomas thickness could also be tracked down but it too has zero knowledge of melt pond specifics in its model.

Since no representative measured ice thickness data has been released, actual thickness remains unknown. It's a pity that Mosaic holds back data to 2023 instead of informing the current melt season.

-- What do we even want 'thickness' and 'volume' to mean when melt ponds of various depths and evolutionary stages occupy fully half the scene, extending out beyond the horizon?

-- Did ice physics models have any real awareness of current conditions at the scale needed? Now that we have data, has it been assimilated? No.

-- What is the effect of draining melt ponds on floe buoyancy and does that fool freeboard satellite altimeters?

-- Could 'averaging' somehow average away melt ponds and leads when they nevertheless dominant the scene at all scales? How is the remaining ice scored after melt ponds drain?

-- Who knew in advance what 88-90º would look like? Nobody. If the PS had not taken those photos, would we ever have known? No.

Meanwhile the PS has left the pole, already 150 km south along the 105th meridian looking for a suitable floe to moor onto. This will take them back into satellite range and back into the TransPolar Drift which is not drifting, but not quite back to their original October lat lon of 85.1  134.3.  Sentinel-2 can image melt ponds on clear days (lol) to 10 m resolution (not quite enough).

According to Ascat, this is FYI formed in the Laptev. It might have maxed out as 2m in late February but is assuredly thinner by now, sub 1m. It's questionable whether they will find a thick enough floe to work on safely -- freeze season is still six weeks off and they'll be leaving before that really sets in. The better CAB floes are off on higher different meridians, say the 180th meridian to Wrangel.

/*/*/*/*/*/*/*/*

Forum discussion of melt pond energetics -- a very big topic in Arctic academia -- went way off the rails. The sky and water are blue from Rayleigh elastic scattering, not adsorption of light and conversion to heat. Insolation energy at the surface is largely at visible wavelengths. Neither H₂0 nor ions like Na⁺ or C^l- have any ability to accept photons at these wavelengths.

Melt ponds are clear to the bottom regardless of salinity. The ice bottoms of melt ponds are reflective white, scattering incoming light in all directions including down. Melt pond albedo is thus less of a drag on fractional BOE than open water albedo. Indeed shallow melt ponds have an albedo of 0.4-0.5 per NSIDC. Yet ROV images show some sunlight manages to pass below everything above.

https://www.sciencedirect.com/science/article/abs/pii/S0165232X15003171

Adsorptive impurities abound even though the scene looks pristine: wildfire ash, miscellaneous industrial and agricultural pollutants and plastics, thermokarst shore erosion, river sediment, fallen aerosols, desert dust, volcanic ash, flakes of unburned bunker fuel, algae and copepods channeled into the ice bottom, and so on and on. Those do convert light to heat (molecular motion) and that heat equilibrates with the surrounding water.

In summary, a good portion of incident sunlight energy is scattered back into the atmosphere, melt ponds lessening that somewhat. The rest is absorbed somewhere in the 'water column', mostly in the upper few meters, the exact apportioning depending on many parameters that change locally over the melt season. The heat cannot readily equilibrate because of diffusive and conductive barriers but still manages to affect many processes.



The hycom image needs a click to display its thickness boxes. To make, I put a one pixel dot right at the pole and put the color picker too on it there so that it also highlighted the right box on the thickness palette, then cropped the stack of co-screenshots.

[UCMiami]

What Polarstern photographs bring home is how limited is our understanding of the state of arctic ice. Looking at Worldview the pole area looks like solid ice without the clear bluing of melt ponds we see during early season surface melt and yet 50% of this surface is actually melt pond. It is a good reminder to us all of the limitations of satellite monitoring and how imprecise 25 km sq or 3 km sq grids can be. It also clearly shows how hard getting accurate sea ice area measurements are when even photographs from a few hundred meters are hard to evaluate.

I would love to see a close photograph of one of those 'blue' sections of WV images from earlier in the year - I assume it must actually look like a huge lake of water on top of sea ice and not a 'pond'. Perhaps we need to refer to them as lakes in future.

[A-Team]

way insufficient knowledge of current ice conditions

I looked into true polar orbits for dedicated radar satellites with no pole holes: not gonna happen because the cost has to be spread out across many planetary interest groups that want sun-synchronous. Almost all Arctic data is derivative: older satellites on other missions that were later re-purposed for ice (eg smos, smap from ag soils). Potentially wider swath technology on a near-polar orbit makes more sense. Not happening any time soon enough.

It would be great if the PS shared a straight nadir overhead from their helicopters. They took those photos yesterday for sure but chose not to share. We could then overlay various grid scales to compare with the intrinsic scale needed to capture key melt pond parameters. The PS provides an in-scene scale of 117 m; the google earth type projection is easily rectified.

To me, the north pole photos sent a mixed message: wow the ice is really bad but worse the scientific Emperors Have No Clothes. Perhaps that is why the ship left the pole hole after only 9 hours on the scene, to change the subject. Yet these pole photos are likely all that will be remembered from the year-long cruise. There may never be a return visit by another R/V.
 
As you say, those bigfoot pixels are a serious mismatch with characteristic dimensions of the object under study. There do exist multispectral satellite methods for remotely determining melt pond depth (eg on Greenland) though again how low has to be flown to capture any detail.

Mosaic ticker talks about June 26th melt ponds visible in Sentinel-2 optical color, said to be 10 m. The initial image quality was mediocre but came out ok after post-processing and 4x enlargement. The problem is the Arctic Ocean is one of the cloudiest places on earth. It could take months to stitch up a picture of melt ponds over the whole ocean, meaning these would have wildly varying timestamps.

https://www.meereisportal.de/en/mosaic/sea-ice-ticker/

The two-slide gif below, after a click, shows the state of the art AMSR2_AWI at its native resolution and colors, plus re-paletted to show the situation the Polarstern is sailing into.

[glennbuck]

What Polarstern photographs bring home is how limited is our understanding of the state of arctic ice. Looking at Worldview the pole area looks like solid ice without the clear bluing of melt ponds we see during early season surface melt and yet 50% of this surface is actually melt pond. It is a good reminder to us all of the limitations of satellite monitoring and how imprecise 25 km sq or 3 km sq grids can be. It also clearly shows how hard getting accurate sea ice area measurements are when even photographs from a few hundred meters are hard to evaluate.

I would love to see a close photograph of one of those 'blue' sections of WV images from earlier in the year - I assume it must actually look like a huge lake of water on top of sea ice and not a 'pond'. Perhaps we need to refer to them as lakes in future.

Do they have drones on the Polarstern? I wonder if they filmed miles of the North Pole with drones.
why are satellites so rubbish over the Arctic in 2020?

<Removed conspirational part. O>

[Phil.]

<em>A-Team
Melt ponds are clear to the bottom regardless of salinity. The ice bottoms of melt ponds are reflective white, scattering incoming light in all directions including down. </em>

A-Team glad to see you back here. re the ice bottoms I would't expect them to be very reflective because the refractive index ratio between the water and ice will be very small compared to that between air and ice.
Seawater RI 1.34, ice 1.31
For normal incidence should be about 1% reflectivity as opposed to 4%, most of the light will be refracted down.

[binntho]

A-Team, you bring sanity to our discussion, as always.

Just one quick question - your discussion of water adsorbtion of energy from sunlight. I understand you to be implying that 100% pure water would be fully transparent, something that seems to be the case in nuclear power station cooling pools, see image.

So the albedo effect of water vs. ice is that water lets most of the visible light radiation through so it can reach impurities in the water column, while ice reflects most of this radiation, so only surface impurities catch any light.

[marcel_g]

A-Team, you bring sanity to our discussion, as always.

Just one quick question - your discussion of water adsorbtion of energy from sunlight. I understand you to be implying that 100% pure water would be fully transparent, something that seems to be the case in nuclear power station cooling pools, see image.

So the albedo effect of water vs. ice is that water lets most of the visible light radiation through so it can reach impurities in the water column, while ice reflects most of this radiation, so only surface impurities catch any light.

No binntho, you’re missing the fact that ice has a range of albedo values. It can range from clear (black ice) to light gray, depending on how fast it freezes, which affects how many air bubbles are in it.

So You seem to have it backwards, ice doesn’t have a high albedo like snow does and doesn’t reflect a lot of light, it normally has a low albedo, so it normally allows most of the light through, unless it has a lot of air bubbles in it. A-Team was patiently explaining That the light that’s reflected from the ice goes in all directions, not just back into space.

The end result is that ice at the bottom of a melt pond has a mid range albedo, reflecting more than open water, but still transferring a lot more solar energy to the water underneath than snow covered ice.

Arctic sea ice normally appears to be a medium gray, So a medium amount of albedo, so if large areas of ice look like that photo from the PolarStern, then even though there wasn’t a lot of open water in July, a much higher amount of solar energy was transferred to the water underneath the ice than normal snow covered ice with less melt ponds.

Ps - this discussion should be moved to a new thread if you really want to carry it further.

[binntho]

A-Team, you bring sanity to our discussion, as always.

Just one quick question - your discussion of water adsorbtion of energy from sunlight. I understand you to be implying that 100% pure water would be fully transparent, something that seems to be the case in nuclear power station cooling pools, see image.

So the albedo effect of water vs. ice is that water lets most of the visible light radiation through so it can reach impurities in the water column, while ice reflects most of this radiation, so only surface impurities catch any light.

No binntho, you’re missing the fact that ice has a range of albedo values. It can range from clear (black ice) to light gray, depending on how fast it freezes, which affects how many air bubbles are in it.

So You seem to have it backwards, ice doesn’t have a high albedo like snow does and doesn’t reflect a lot of light, it normally has a low albedo, so it normally allows most of the light through, unless it has a lot of air bubbles in it. A-Team was patiently explaining That the light that’s reflected from the ice goes in all directions, not just back into space.

The end result is that ice at the bottom of a melt pond has a mid range albedo, reflecting more than open water, but still transferring a lot more solar energy to the water underneath than snow covered ice.

Arctic sea ice normally appears to be a medium gray, So a medium amount of albedo, so if large areas of ice look like that photo from the PolarStern, then even though there wasn’t a lot of open water in July, a much higher amount of solar energy was transferred to the water underneath the ice than normal snow covered ice with less melt ponds.

Ps - this discussion should be moved to a new thread if you really want to carry it further.

I'm sorry but I don't agree at all with your explanation. And I am not "missing any facts" just because I don't cover all angles and expound all caveats.

And your "ice doesn't have a high albedo like snow does" seems to be "missing the fact" that snow is ... (tadah!)  ice! The rest is just rambling.

EDIT: This is off-topic, but I posted a very short question to A-Team in the hopes that he would answer, and we could all avoid the bessenwisserschaft that seems to fill too many of us to overflowing. So if anybody else wants to answer, please take it elsewhere.

[Phil.]

The end result is that ice at the bottom of a melt pond has a mid range albedo, reflecting more than open water, but still transferring a lot more solar energy to the water underneath than snow covered ice.

No, as I pointed out in a response to A-team submerged ice has an albedo of about one quarter of open water.

[A-Team]

The Polarstern has been at 87.7  104.2 for over nineteen hours, having settled on a floe for the rest of Mosaic. The idea was to land on their previous track to give a semblance of continuity. The new floe in reality will have a very different history and basically nothing in common with the previous Fortress floe, as can be seen by the precarious conditions setting up. Trackbacks in late summer are very difficult unless the floe has distinctive features.

The ship was last nearby (87.5  103.6) on 15 Jan 2020. They are just north of the area that can be imaged by Sentinel-1AB radar (ie in its pole hole) which is unfortunate as there won't be optical WorldView with weather so bad the helicopters can't fly.

There have been occasional Modis windows through the clouds though and a really clear view from space is shown below for Aug 10th (as enhanced by uniq).The resolution is just not there to capture details at the melt pond level.

Bow radar has not been turned back on. Today's 'Follow' photo though is quite instructive. It shows quite degenerate late season ice with a variety of novel platelets, melt ponds, sagging hummocks, open leads and so on, out to the horizon but of a different character than the north pole seen. I'll locate it in the next post on the most recent AMSR2_AWI.

[RoxTheGeologist]

The end result is that ice at the bottom of a melt pond has a mid range albedo, reflecting more than open water, but still transferring a lot more solar energy to the water underneath than snow covered ice.

No, as I pointed out in a response to A-team submerged ice has an albedo of about one quarter of open water.

I thought open water has an albedo around 0.1, so 90% absorption and submerged ice is around 50, so 50% absorption. Open water absorbs about twice as much energy as submerged ice.

[A-Team]

That's correct, Rox. Those are the numbers given at NSIDC that I posted above. The blue-green submerged ice is roundtrip incident light that makes a second reflective pass out through the water to reach the camera. Really really clear ocean water is actually not blue but purple (ROYGBV and all that.)

https://www.newscientist.com/article/dn12162-found-the-clearest-ocean-waters-on-earth/

Wiki has a strong article on the the adsorption spectrum of liquid water and ice. It has a faint but non-zero optical density in the visible that turns out to be vibrational quanta related to the hydrogen bonding, ditto for ice which has lattice adsorption as well.

https://en.wikipedia.org/wiki/Electromagnetic_absorption_by_water

I cannot emphasize enough the importance of skimming through the scientific literature on melt ponds. Nothing has changed with melt ponds over the decades so older papers are still good.

There are literally thousands of peer-reviewed melt pond papers listed at Google Scholar. Below are a few from 2020. No paywalls today, so no excuses if that's something people really want to delve into, rather than just make stuff up.

Maybe after assimilation, it would make sense to armchair intuit consequences for the north pole area for the current melt season. I've been wondering what happens if they never get around to draining and instead ice over for the next winter (into next melt season).

The salinity would vary depending on brine pockets that got taken up in the melt process but overall a lens results with quite different properties from the residual ambient ice surrounding the melt pond.

Many properties of the floe would be affected in the next melt season. The PS measured some of the melt ponds to a depth of 1.5m in ice of perhaps 2m. It is difficult to estimate the encased liquid volume fraction from the pole photos but it seems quite substantial. A whole field of melt ponds is a different scenario from 1-2 in isolation.

Meanwhile, back at new AMSR2_AWI and old AMSR2_UHH, what do they see at the new mooring location of the Polarstern? Not much, they do not have nearly the resolution to see small melt ponds and leads. The re-colored palette of ice under the red location star is entirely in the 99-100%.

Meanwhile the Polarstern is drifting away, currently 10-15 km off at 87.8  104.9  at 17:00 today. The AMSR2 are both from the previous day but that isn't a concern given the large consistent block. Hopefully the bow radar will be reactivated soon and maybe they will take some visuals to go with it (haven't since Nov 1st).

[Phil.]

The end result is that ice at the bottom of a melt pond has a mid range albedo, reflecting more than open water, but still transferring a lot more solar energy to the water underneath than snow covered ice.

No, as I pointed out in a response to A-team submerged ice has an albedo of about one quarter of open water.

I thought open water has an albedo around 0.1, so 90% absorption and submerged ice is around 50, so 50% absorption. Open water absorbs about twice as much energy as submerged ice.

Based on the physics the reflectivity of pure water is about 4% for vertical illumination whereas the reflectivity of a submerged ice surface under water is about 1%.
The absorption coefficient of red light is about 100 times that of blue light.
https://lh3.googleusercontent.com/proxy/yraspHmp2H0lx2EZJ7H05wt7rGUXmaxbh33l65iN_ycS84qGxJz8eptTrKb7WNoSYSKygcGjlvtztGIyNqyTxxVJ2zALuLpCndiLnUElRSU

[marcel_g]

Thanks for the level headed reply A-Team. Next time I will do some more research first, and I should probably just ignore posts by <another user> I find derailing.

<Edited said user name. O>

[gandul]

The Polarstern has been at 87.7  104.2 for over nineteen hours, having settled on a floe for the rest of Mosaic. The idea was to land on their previous track to give a semblance of continuity. The new floe in reality will have a very different history and basically nothing in common with the previous Fortress floe, as can be seen by the precarious conditions setting up. Trackbacks in late summer are very difficult unless the floe has distinctive features.

There is nothing precarious in that floe in terms of thickness though. That’s clearly much thicker than 2m. There is also the field of ice right behind that, which, while hollowed by melt ponds, is clearly pretty thick as well, and will make a decent base when everything freezes.

I’m concerned however by the strong winds forecasted for the next days. Things may then show its precarious state.

[binntho]

Thanks for the level headed reply A-Team. Next time I will do some more research first, and I should probably just ignore posts by <another user> I find derailing.

<Edited said user name. O>

Well, said user is of the opinion that if you condescend, then you must expect to be derailed.

[A-Team]

Not any further scientific news from the Polarstern today. The bow radar is still off. Their position continues to drift a bit under mild winds, currently 87.8  106.1  Aug 23 14:00 wind 6m/s from 200º. They are likely setting up a slightly Dispersed Camp near the ship to get away from its diesel and ramming impacts yet have reliable cables for data and electric. These caused endless grief during the main cruise.

They have not doubt determined ice thickness by drilling a few spots and weather permitting, helicopter and sled raster surveys. Hycom is showing the ice in their area to be about 0.75m thick on Aug 19 and thinning to perhaps 0.65 m by the end of the month forecast period. They are in FYI formed off the Laptev, not in older CAB ice.

Mosaic trip leaders prefer 1.0m or thicker floes for safety in severely decayed and sometimes deceptive ice but that seems not an option near their current position. Fog did not permit them to look around systematically for pressure ridges that might vaguely approximate the long-gone initial Fortress site.

So far, the PS has not set down its gangplank. It's not clear from the one photo we have where exactly it would go. The scientific equipment and snowmobiles would be winched down in any case. The underwater ROV could be launched right alongside the ship -- it does the underwater mapping of keels and freeboard that is needed along with on-ice surveys to map ice thickness in their area.

Snow? They've said it has snowed less than five times during the entire year. It never layered up on regular ice but was entirely blown in the lee of pressure ridges and the like. Rain, sun and warmth have probably obliterated any remnants of that.

Having already lost many weeks of research time in the south and with freezing season approaching, they settled in to the first workable site, focusing today on implausible self-rescue by ice pick drills for everyone going out on the very dangerous ice near the ship. The US Coast Guard has vastly higher safety standards for the Healy (which recently had two diving deaths).

The drowning death this week of veteran Greenland researcher Konrad Steffen in a well mapped, meltwater filled Greenland crevasse at Swiss Camp has surely heightened risk awareness.

[oren]

A-Team, to my untrained eye the ice where the poeple located and the ice in the far background seems flat and less than a meter, the far background seems very similar to the North Pole photo with all the melt ponds. However the ice in between appears taller out of the water and bumpy, and could be a pressure ridge remnant. I am also guessing they intentiinally picked a floe that was much thicker than the typical meager floes in that region. I wonder if you read this the same way.
(I refer to the wider image you provided in post #1013).

[uniquorn]

recent weather report from awi met

51 -- continuous light drizzle
28 -- fog
71 -- continuous light snow

water temp from sailwx still relatively high peaking at -1.4C. Probably lower salinity allowing that rise, probably due to bottom melt

[A-Team]

51 -- continuous light drizzle
28 -- fog
71 -- continuous light snow

You'd think a national weather service would have mouse-overs to the acronyms. Not rocket spreadsheet science to go through awiMet ship weather for the entire cruse (7561 entries) and pull out all the ones with weather comments (396 of them) sorted into class counts. So that's attached as a csv file below, ready for uniq(?) to process them further along the lines above, eg how many foggy days.

More challenging: go through the entire voyage on nullschool and do something with their 3hr precip claims (3HPA) for the whole Arctic Ocean. Not make a 5 hour mp4 but more of a summary. Rain is just devastating to snow and ice because of then phase change latent heat.

Oren: ice doesn't look all that thick overall but maybe locally

Very reasonable interpretation of the photo (which is all we have to go on).

Earlier in the year, 'Follow' did post a couple of raster thickness maps from above and below, in addition to near-horizontal mapping of topography. We have those above somewhere; they unfortunately forgot to include the numerical scale. Maybe this time they will!

The bow radar (above the upper bridge) just takes in surfaces reflections so maybe roughness and open water but not thickness. They unfortunately forgot to pair it with a camera or calibrate it with outings plus a Finnish student mistakenly cropped away the whole instrument parameter display for the entire cruise.

The ship sails with a electromagnetic doohickey suspended off the prow that measured ice thickness (ie distance to saltwater) along the entire way from Greenland north. That would not provide representative ice thickness since they were always looking for a route involving minimal icebreaking. Be the same for a time lapsed smart phone lashed to the mainmast.

To answer your question though, we are somewhat at cross-purposes with the Polarstern.

The forums are terribly concerned with the current state of the ice ocean-wide, learning about the mysterious ice conditions in the pole hole, mapping features like melt pond extent below satellite resolution, getting spot measurements of ice thickness to compare with piomas and hycom models, describing how the melt season will end up, quantitating heat that will continue to be available into the fall, and understanding how it all compares to the trends of previous years.

Mosaic has a different agenda: trying to comply with the terms and objectives described in their massive grant. These stress goals such measuring this and that on the ice over a full year to improve interpretation of satellite imagery that will monitor the ice in future years when no ship is out there.

Indeed those improvements may already surfaced in the new AMSR2_AWI as that PI from leg 3 explained in a recent ticker. That's why we've dropped Hamburg and Bremen, putting all the effort in understanding to the new product and perhaps adding inclusivity and presentation value.

The ship is just a moving point on millions of sq km of ice; efforts to widen the swath with remote deployments and buoy swarms have had mixed results. There's been little mention of larger scale  ongoing sonar tomography that uses the ship as passive receiving station. This would map out oceanographic detail for the whole Nansen Basin top to bottom.

The PIs have said repeatedly that the top cruise priority now is documenting the onset and mechanisms of early freeze season to complete the funding terms of the mission. Since they moored on 04 Oct 2019, that probably means going out a week at most to 12 Oct 2020 before heading back to Tromsø.

So in by way of providing us better understanding of the current melt season, it will take some time (2023) before all the data can be corralled into print. N-ICE2015 just put out another paper modeling AW as it was five years ago.

In fairness though, the 25 years of mooring studies through 2018 just published by Polyakov et al are extremely relevant to the current melt season as the slowly but inexorably shoaling Atlantic Waters is taking over the energy budget from surface weather.

Will the lessons learned by the Polarstern still be relevant to the ice of that day or fall more into paleo?

Given the pace of climate change, especially in the Arctic, chewing of cud seems an unaffordable luxury. Darwin could spend 40 years polishing his snail monograph but today we need something closer to near real time. New satellites sound great but take years of planning to implement. Thus air-dropped buoys, undersea gliders and transect-flying drones may be more the way to go.

[gerontocrat]

Given the pace of climate change, especially in the Arctic, chewing of cud seems an unaffordable luxury. Darwin could spend 40 years polishing his snail monograph but today we need something closer to near real time. New satellites sound great but take years of planning to implement. Thus air-dropped buoys, undersea gliders and transect-flying drones may be more the way to go.

Thus air-dropped buoys, undersea gliders and transect-flying drones may be more the way to go.

Many years ago, the US military allowed civilian access to their GPS satellties,
Many years ago the USAF allowed access to their DMSP satellites, and loa and behold, the NSIDC we know today was born.
HYCOM is a US Navy operational product, with civilian access.

On the thread "Robots and AI: Our Immortality or Extinction" you can see how the military have developed the use of AI + drones and undersea ROVs for military objectives with disturbingly hgh rates of success. In today's political climate, can we hope for the application of these AI systems and hardware for peaceful purposes, e.g. looking at and measuring the cryosphere?

[A-Team]

Civilian air-dropped buoys that anyone can purchase and deploy unregulated for any reason were tabulated here earlier. Some 33 of them have been put out on the Arctic Ocean in recent years. They report over iridium to open-source IADP at UW.

University science gliders are out there already too in the Beaufort, doing many thousands of km of surveys. This is a sawtooth path technology that makes clever use of very slight buoyancy differences that take very slight amounts of battery power to implement. They can surface from time to time to send data so are an improvement in that sense over fixed moorings. These are not high tech and any hobbyist could build one at home and launch it.

As Rod noted earlier, the PS likely has small drones that spare routine helicopter tasks and their limited fuel. Such drones have been widely deployed at glacier fronts in Greenland to make orthographic time series. Again, something anyone could buy at ebay, launch themselves, and process and post the data. In the US this is somewhat regulated and not allowed say at Grand Canyon or over police and fire emergencies.

Now the larger drones capable of flying back and forth across the Arctic Ocean with a lot of camera technology, that could be more of a problem in terms of acquisition options, expense, transport, massive files and approvals. The Coast Guard might very well want to partner with scientists on this for marine safety reasons and given the short shelf life, what would be the reason for not distributing it?

There is a lot of commercial air traffic over the Arctic now in normal times but way above the cloud deck (though nobody seems to set a minimal altitude on that).

Even the big ones are easy to fly over the internet and indeed all the US drones in the Middle East are run down the road from me by young enlisted video gamers. It might be simpler just to charter a small plane from Western Aeronautics and have them shoot large format through the floor. Who is going to stop you?

Here we spent $150m euros to get 9 hours at the North Pole during which we learned we didn't know anything at all about ice conditions at high latitudes from existing satellites, buoys or one-off cruises, ice tourism or adventuring.

[uniquorn]

So that's attached as a csv file below, ready for uniq(?) to process them further along the lines above, eg how many foggy days.

101 -- continuous light hearted laughter

actual weather this morning (24th) for PS has been mist and fog

[A-Team]

No information at all today from Mosaic, get over it. Bow radar still not working ... does that mean they need it for operations like looking around more for a good floe? Mist and fog mean they cannot leave the ship because of the polar bear visibility issue.

At 15:00 though, the wx staffer put visibility at 96 which is 4-10 km but fully overcast at 8/8 and cloud deck 1 which is low, 50-100m overhead. Conditions are only noted 4x a day, not hourly like position, wind, and temperature. There is practically no wind, 1m/s (despite the storm that's been raging on the melt forum). Wind is measured from a mast 38m above sea level.

FoMo is back to hugs and cherished memories from yet another well-paid creature-comforted job-related cruise; Mosaic's main expedition page always makes me feel I've stumbled onto a private TikTok party link rather than public-funded climate change research site. In past decades, the Polarstern has always posted a detailed weekly ship log.

We do learn however that the Polarstern skipped town without retrieving all the expensive equipment left on dispersed Fram floes. However the chartered Russian icebreaker Akademik Tryoshnikov did manage to pick some of it up. The rest will sink to the bottom off Greenland's national park.

From wiki: Alexey Fyodorovich Tryoshnikov (Алексе́й Фёдорович Трёшников) was a Soviet polar explorer, the leader of the 2nd Soviet Antarctic Expedition and the 13th Soviet Antarctic Expedition and later a member of the Soviet Academy of Sciences. He was involved in defending the Northern Sea Route during World War II and participated in the 1948 Soviet expedition to the North Pole. Between 1954 and 1955, he was the leader of the North Pole-3 ice station in the Arctic Ocean.

[uniquorn]

weather update:
06:00    41 -- patches of fog
03:00    48 -- fog, depositing rime, sky visible

I suppose with poor visibility and treacherous ice there are few opportunities for buoy deployment.

[A-Team]

New buoys ... hopefully a lot of them. They are just laying out snowmobile routes now for the semi-dispersed camps. The question is, would buoy-like instruments report over fiber cables or a local LAN rather than Iridium in which case the data would not show up online at IABP or meereis gallery.

Mosaic has various rules around wind speed, precipitation and visibility that constrain helicopter flights. Right now wwWW is saying 0354 which is 'clouds developing, drizzle, fog' with height saying very low clouds 50 to 100 m at 20-08-26 09:00 UTC. These are probably no-go for launches.

The 'Follow' photo for today shows two people filling in a melt pond with shovel and digging bar against a background of heavy fog. Frankly it is not possible to mount a bear guard under these conditions but maybe a double electrified fence line has been set up.

Both have unzipped the top half of their safety suits despite 15 minutes maximal survival time in -1.8ºC water. They could not possibly self-rescue with all the weight of water immediately filling their suits. http://www.ussartf.org/index.html

The two are not shoveling snow since there has hardly been any the entire year. It looks more like quite rotten ice -- it would not be possible to make a dent in solid ice with these tools. In any event, they need a level haul track to get heavy gear and power lines out to the new off-ship observing center. Again, we do not know the true date of this photo as others have been 3-4 days behind despite a current time stamp.

Interesting minor project: take the awiMet weather lines, do the lookup of all the codes, and replace them with the applicable weather text. MS Word has that glossary feature built in.

It's not clear that new dispersed buoys would drift away from the Polarstern in the short term. OsiSaf is showing a lot of purple arrows and grid cell zero-movement dots, having given up entirely on the light blue periphery.

The bow radar has been turned off for a full week now (or at least not posted). The last blurry scene was at 09:00 on the 19th. The ice is not likely to be very mobile right now; it's not clear whether melt ponds can be distinctively resolved from other ice features with this radar.

The field of view of the bow radar is about 500 pixels per 5000m so about the same as Sentinel-2 which we haven't used much over the summer because of the inconvenience of the portal and now clouds and diminishing sunlight. The new AMSR2_AWI would have only 4 pixels to cover the bow radar scene, if you could find them.

Note nh_20200826_21_SIC_72h.tiff has a long diagonal swath line almost like a WorldView.

[Latent]

Thank you A-Team.  Your clear explanations and perspectives are very welcome.  Especially the point about the purpose and funding drivers of the Mosaic expedition.

[UCMiami]

That last picture posted is certainly a bit strange. Could they be attempting to discover the depth of the porous ice above whatever is left of the solid ice? It appears the top of whatever they are standing on is about a meter above the melt pond level. Temperature must be pretty warm for one to be bare armed (or they are working very hard.)

[uniquorn]

from follow mosaic

[marcel_g]

A-Team, to me it looks like the 2 individuals might be shoveling some heavy snow that’s gone through a few cycles of thaw/wetness and refreezing. Looks similar to the snow here in the spring.

[A-Team]

Plausible. Also they do have various snow plows on board and this material may have been worked over already. They are also reporting fog and steady drizzle which would affect snow appearance and properties. In today's staged photo, the sled is not pushing any snow ahead of its horizontal member.

The only place experiencing massive Arctic snowfall this year was once again the main forum. The Polarstern itself only reported 4-5 measurable snows. Those were quickly redistributed into drifts in the lee of pressure ridges. The Polarstern visited this spot back in mid-January but has not disclosed or compared conditions.

We need to re-think snow cover impacts on the albedo and melt pond scenarios in the face of observation. Also missing: rainfall data and air heat capacity (fog).

Amazing that they were able to recover and re-build this flux sled. It really got smashed up by floe motion a few months back. M Shupe posted a good story about its rescue in the Fram:

https://blogs.agu.org/thefield/2020/08/17/postcards-from-a-formerly-frozen-icebreaker-part-55/

They still are not posting the bow radar so it is difficult to assess local ice motion today. The unofficial arrow points to a possible melt-through of the melt channel. It is still quite foggy. That backpack and other items should lashed to a small sled, not worn.

The Arctic Ocean is currently just that: over half is open water. The simplified scene below is comprised (by pixel count) of 19% land, 42% open water and 39% ice. Not counting land, the ratio of open water to ice of any kind is 52:48.

OsiSaf today is coloring 37% of the ice pixels in light blue. This is the ice most at risk from winds and bottom melt over the next month. Note the patch of low concentration ice just below the pole. Persistence over multiple days needs to be established; it is too cloudy for WorldView VIIRS to help.

[FishOutofWater]

The pace and style of old academia, as done by the Polarstern scientists is failing us. By the time they get their results published those results will be out of date. Yes, the basic science has not changed, but the ice conditions have deteriorated over the past 3 years.

We are mostly flying blind on this forum, speculating or measuring insignificant changes to infinitesimally diminishing significance.

I plead guilty to speculating about what's happening below the ice between Greenland and the pole. The key factor was probably the unprecedented clear weather and direct sunlight, but we need buoys down there recording what's happening and we need real time analysis by knowledgeable scientists like A-Team. I agree with him about the deployment of surface and below surface instruments. Satellite measurements have limited capabilities as we can see with our own eyes from the photos at the pole.

The Arctic is melting rapidly while the traditional scientists play their old academic games. And the atmospheric circulation is shifting all the way up to the upper stratosphere while they dither over which ice floe to moor on. (The rapid jump in the QBO this summer is very unsettling.)
https://www.geo.fu-berlin.de/en/met/ag/strat/produkte/qbo/index.html

Thanks A-Team for your analysis.

[uniquorn]

Some new buoy data.

[uniquorn]

T78, nearest to PS, has been deployed on ice ~1.5m thick (~therm30-105). Nice to see some below zero air temps there.

[A-Team]

A few days of moderate surface winds are forecast at GFS, including easterlies along the Beaufort, a large but moderate anti-cyclone on the Laptev, and a commonly seen tight cyclone around Svalbard.

As a reminder, the bottom line is friction between local surface winds and ice roughness (respectively water) is what moves today's ice (makes waves). What the MSLP might be doing in the Azores over the next ten days, how the jet stream pattern could change far above as a consequence of fractional BOE, and whether ENSO will be trending in the eastern Pacific don't move ice. They might ultimately drive big-picture wind but predicting it below 2m isn't even an aspirational goal.

That friction is likely higher in the Beaufort because of exposed freeboard edges of dispersed floes. Thus a few days of persistent easterly wind may push much of what's left of the ice below the threshold of radar satellite visibility towards the warmer Chukchi. Waves are always a factor in late season but here long reach swells will not be a player.

The image below places these winds relative to the unfiltered 21:00 UTC of AMSR2_AWI (v101/nh_20200827_21_SIC_12h.tiff) by deleting the five highest palette ice concentrations (less likely to be affected by the winds) to let the 21:00 UTC of nullschool winds show through.

By changing the underlay, that could be visualized 2-3 days out but the problem is that the AMSR2 concentration overlay has no forward map. The difference between 24 hour 21:00's is already quite striking so the wind overlay on a fixed 21:00 of Aug 27 would rapidly become inexact.

Hycom does offer a forward map so AMSR2 could somehow be morphed along to follow its outline. More easily, the wind prediction could be shown behind deleted hycom open water. As interstitial noted, hycom's edge is way too far out. However those colors can be picked and filled with ocean gray.

The scale on nullschool was set to 2800 in its stereographic projection so rescaling by 141% matches it to AMSR2_AWI.

Note this new AMSR2 is unique among products used on the forum in that it is correctly constructed as an 8-bit color geo-located tiff, rather than 24-bit png, jpg, or ever-changing 8-bit gif89. Tiff specs are defined on page 23 of the 1992 defining document:

"Palette-color images are similar to grayscale images. They still have one component per pixel, but the component value is used as an index into a full RGB-lookup table." .

The particular tiff ColorMap used at AWI was laid out somewhere within the bundled file package as an ordered 3-column RGB list; a Baseline tiff reader such as ImageJ's then displays the color map tiles via the 'Edit Lut' command.

An ordinary full color image such as png can be changed to 8-bit color and saved as tiff but it won't have an orderly ColorMap as it is simply made from the highest frequency color usages. However as a csv file it can be sorted to order and loaded as the working LUT.

https://www.adobe.io/content/dam/udp/en/open/standards/tiff/TIFF6.pdf

[UCMiami]

A-Team - gorgeous image. Should this not be posted on the melt thread? Removing the condensed central pack from the image clearly focuses the eye on not just the areas which will be most affected by the winds, but also those areas most likely to be most involved in extent and area changes.

I understand what you say about the Beaufort not developing long reach swells. My question would be is the Laptev open water and a fairly prolong wind pattern exposing that front to more of those long reach swells?

[A-Team]

UC: Laptev more exposed to swells in coming days? cross-post melt forum?

Right, ice in the western Laptev ice appears more at risk from swells than that of the Beaufort though the latter is currently farther gone. The reach of the wind there is potentially 800 km or more, the speed is gale force at 15 m/s in places, the forecast calls for a few days of persistence, and the putative swells would hit head on.

That combination seems more than enough for a significant or even dramatic swell damage. While no one is out there to provide confirmation, if something significant does come to pass on satellite, swells should certainly be considered as candidate explanation. We don't have any real data on how common major swell events are; reports are entirely serendipitous.

People are welcome to cross-post to other forums or anywhere else on the internet -- and lay on their own interpretation. I recall that forum software will re-display images and animations just from its url, as wrapped with a 2nd row button above. The image does not need to be downloaded and re-uploaded.

There are two advantages to this: the image does not count towards the posting limit of 4, it can be reside in the middle of a text sandwich where it fits better rather than sit context-free on the bottom.

Below is a spectacular north pole ice photo I came across on a news site. It could be placed in any number of forums. Actually I am just trusting the caption; the photo is undated, has no lat lon nor  EXIF data. The Polarstern was at the pole on August 19th but Folke Mehrtens may not be the actual photographer as stated, she has been at the AWI's communications press office since 2008.

The second photo is on AWI media twitter so presumably is at the north pole as stated. It's very unfortunate that AWI does not put together a permanent gallery of properly labelled, dated and credited North Pole photos -- these will never be published or analyzed in a scientific journal but nonetheless are an important part of the record.

So many generic stock photos are shown, including by AWI, that the public never knows for sure when they are looking at the real thing.

[uniquorn]

This is a bit late, posted on aug17. Collecting buoys etc in the Fram strait
edit: written on aug4 and aug6. Photos dated aug12

Postcards from a (formerly) frozen icebreaker: Part 55

First we found a SIMB3 buoy bobbing amid little ice floes. Easy to just pluck this one up out of the water. Then the ocean flux buoy; this was much harder as it was still embedded in a small chunk of ice. The ship’s crew pumped ocean water on it to melt out around the base and then lifted it up and out of the water, followed by a tether below and additional instrument packages. I’m glad we recovered this system as we believe the next crew of scientists will be able to redeploy it during leg 5 to help get ocean heat flux measurements at this crucial time of year when the ice melt slows and eventually stops, giving way to another freeze up. On this day we did not get the ice tethered ocean profiling buoy, as it was nowhere to be found, but we have to call this a highly successful day nonetheless.

edit:fixed link

[D-Penguin]

As a reminder, the bottom line is friction between local surface winds and ice roughness (respectively water) is what moves today's ice (makes waves). What the MSLP might be doing in the Azores over the next ten days, how the jet stream pattern could change far above as a consequence of fractional BOE, and whether ENSO will be trending in the eastern Pacific don't move ice. They might ultimately drive big-picture wind but predicting it below 2m isn't even an aspirational goal.

A question - Are you saying that the condition of the Jet Stream, at a point in time, plays no significant part influencing and is unrelated to the Arctic weather system forecasts?

[A-Team]

A bit late, posted on aug17. Collecting buoys etc in the Fram strait

So once again, a bogus caption on a photo? They were not anywhere near the Pole on the 17th. Nor were they in the Fram but rather 86.8 12.3. Even the cheapest flip phone records time of photo on the photo.

AWI does provide a photo database for media but access requires a login that does not recognize the account you just created.  Sound familiar? https://multimedia.awi.de/medien/earlylogin.jspx

Jet stream important influence?

Of course in the big picture it's all tied together, though to date we appear no closer to reliable 5-day much less seasonal predictions than ever. Knowledge of the jet stream does not lead useful prediction of the surface winds that actually physically couple with the ice,.Especially this August, that would be a big leap given the jet stream's complex and semi-chaotic structure.

Rain, fog, melt pond and snow distribution: terribly important to the ice but where in the jet stream graphic (or any satellite product 42 years in) do I find reliable data for them at the scale we need? We have to get some better sensors out there; the Polarstern is not the answer.

Consequently, I would rather glance at nullschool 1000 hPa winds twice a day rather than read ungrounded weather speculation on the melt forum. And that is true too where I live, despite a full-time gov't Wx staff of seven, they cannot get a half day ahead on weather specifics such as rainfall.

So yes, maybe we can usefully do 2-3 days forward but really after-the-fact (yesterday) is the best option, if it is not a broad-brush climate forum. But frankly, given the importance of Arctic sea ice to the global climate system, the complete ignorance demonstrated on Aug 19th of the state of the north pole regional ice makes me wonder what they've been putting in the models all these years in place of observational fact. Having half the ice pack in melt pond doesn't matter?

But even with weather facts in hand, the explanation can elude us. Search 'arctic amplification' at google scholar. The top hit is a 2006 paper by MC Serreze and JA Francis called "The Arctic Amplification Debate". They could write a second paper with that same exact title today. The phenomenon has been under debate since 1969.

https://en.wikipedia.org/wiki/Polar_amplification#History

[D-Penguin]

The pace and style of old academia, as done by the Polarstern scientists is failing us. By the time they get their results published those results will be out of date. Yes, the basic science has not changed, but the ice conditions have deteriorated over the past 3 years.

We are mostly flying blind on this forum, speculating or measuring insignificant changes to infinitesimally diminishing significance.

I plead guilty to speculating about what's happening below the ice between Greenland and the pole. The key factor was probably the unprecedented clear weather and direct sunlight, but we need buoys down there recording what's happening and we need real time analysis by knowledgeable scientists like A-Team. I agree with him about the deployment of surface and below surface instruments. Satellite measurements have limited capabilities as we can see with our own eyes from the photos at the pole.

The Arctic is melting rapidly while the traditional scientists play their old academic games. And the atmospheric circulation is shifting all the way up to the upper stratosphere while they dither over which ice floe to moor on. (The rapid jump in the QBO this summer is very unsettling.)
https://www.geo.fu-berlin.de/en/met/ag/strat/produkte/qbo/index.html

Thanks A-Team for your analysis.

"By the time they get their results published those results will be out of date"

So very, very true!!!

In my more recent postings I have directly stated or inferred that the quantitative analysis of the science is inadequate; that scientists and policy makers are failing to relate to the reality and provide the required leadership.

"We are mostly flying blind on this forum"

Again, so very, very true!!!

There is much 'belly button gazing' going on. There is not much by way of idea exchanges. If an attempt is made to open a debate on the 2020 Melting Thread and its significance to future events, it is immediately swamped by postings commenting about weather forecasts, images from space and reference to the events of previous years; the comments are of interest but limited 'use'. Are we just posting to move up to the next grade of membership or gain 'I like' points?

Surely, the purpose of this forum and indeed the science is to 'inform' but this now requires a new perspective based on qualitative assessment of the reality but relative to the basic science. We now have to believe what we can literally SEE is happening, not blindly rely on imprecise satellite information and projections based on historical data.

We need to present reasoned opinions about when the Arctic ocean will become ice free and think creatively to stimulate new avenues of exploring events. Or do we passively report and comment on current events? Do we agree with the latest science indicating that we could see an Arctic ocean free of summer ice by 2035? Is it possible for the Arctic ocean to be ice free in summer and winter by 2035? Will the progress to ice free events be transitional or abrupt? Will Atlantification prevent ice from reforming after a consecutive summer/winter BOE? Are we not now at the time when informed opinion is of greater value than the science to inform policy makers?

IMO we have run out of time for the science; science is still trying to catch up explaining the history of events.

As I have said before, the science will still be projecting  that the Arctic will be free of ice 10 years after the event at worse or confirming the Arctic is ice free on the date of the event at best. We do not need the science for either of these two 'scientific' outcomes.

Not sure to which thread this rant should be posted.

+1 just for telling it as it is

[D-Penguin]

Jet stream important influence?

Your detailed reply to my question is very much appreciated. Thank you for the link to the paper on Atlantification.

IMO if we have to rely on the science in preparation for the future catastrophe...we are knackered!

+1 just for the insight and honesty you bring to the Forum

[uniquorn]

The 55 postcards are interesting.
https://blogs.agu.org/thefield/category/polarstern/

[A-Team]

Shupe posts a whole lot better information than AWI, also posts lectures

Right. Cires, AGU, Helmholtz, Ticker and scattered Twitter have provided occasional blog information of much better quality than the upbeat censored drivel coming out of AWI's Follow. It is most odd that these outside resources are scarcely mentioned or promoted on the Mosaic site.

Of the 600 scientists involved no more than 5-6 have made any effort whatsoever to communicate what really goes on in Polarstern research.

Mosaic is not interested in the current melt season or characterizing the overall state of the ice. The focus initially was directed entirely to a highly atypical pressure ridge, since disintegrated, that they settled on in desperation last October. That's been replaced by a one-off pressure ridge in a sea of melt ponds that cannot possibly provide year-long continuity.

The ship, in effect a gigantic polluted buoy, made good efforts to widen its swath of observation but even close-in, unexpected ice mobility made it all but impossible to keep equipment running and reporting. We have a good idea of overall drift from buoys but the bow radar picture of floes jostling and leads opening has not returned. This radar, essential to ship operations, is in good repair, just not being shared.

Another camper was killed by a three year old male polar bear in Svalbard. His mother and more recent cubs had recently been airlifted out of town. The victim was actually the manager of the barren Dutch facility and in his second season there. Six other campers witnessed the pre-dawn attack and were said hospitalized for shock.

The presence of two bears and persistent fog at the current location raise a number of issues about how researchers are being protected away from the ship; in the early months of the cruise, safety was taken very seriously.

[Jim Hunt]

If an attempt is made to open a debate on the 2020 Melting Thread and its significance to future events, it is immediately swamped by postings commenting about weather forecasts, images from space and reference to the events of previous years.

Whilst I broadly agree with the points you raise I must also plead guilty to posting "images from space" in the 2020 Melting Season that reference "the events of previous years", not to mention "the ice between Greenland and the pole".

What's not to like?

By way of example, here's Kap Morris Jesup and thereabouts in March 2013

[A-Team]

Another good source of information is the preprint page at The Cryosphere, conveniently reverse chronological.  A third Mosaic paper recently appeared there along with yet another from the predecessor cruise N-ICE2015. Overall though, a majority of Arctic sea ice papers are not based on either expedition.

https://tc.copernicus.org/preprints/preprints.html

Surface-Based Ku- and Ka-band Polarimetric Radar for Sea Ice Studies
J Stroeve et al

Experimental evidence for a universal threshold characterizing wave-induced sea ice break-up
J Voermans et al
https://tc.copernicus.org/preprints/tc-2020-201/tc-2020-201.pdf

Waves can drastically transform a sea ice cover by inducing break-up over vast distances in the course of a few hours. However, relatively few detailed studies have described this phenomenon in a quantitative manner, and the process of sea ice break-up by waves needs to be further parameterized and verified before it can be reliably included in forecasting models.

In the present work, we discuss sea ice break-up parameterization and demonstrate the existence of an observational threshold separating breaking and non-breaking cases. This threshold is based on information from two recent field campaigns, supplemented with existing observations of sea ice break-up. The data used cover a wide range of scales, from laboratory-grown sea ice to polar field observations. 

Trends and spatial variation in rain-on-snow events over the Arctic Ocean during the early melt season
T Dou et al
https://tc.copernicus.org/preprints/tc-2020-214/

Rain-on-snow (ROS) events can accelerate the surface ablation of sea ice, thus greatly influencing the ice-albedo feedback. However, the variability of ROS events over the Arctic Ocean is poorly understood due to limited historical station data in this region. In this study early melt season ROS events were investigated based on four widely-used reanalysis products (ERA-Interim, JRA-55, MERRA2 and ERA5) in conjunction with available observations at Arctic coastal stations.

Our results show that ERA-Interim better represents the onset date of ROS events in spring and ERA5 better represents the phase change of precipitation associated with ROS events. All reanalyses indicate that ROS event timing has shifted to earlier dates in recent decades (with maximum trends up to -4 to -6 days/decade in some regions in ERA-Interim), and that sea ice melt onset in the Pacific sector and most of the Eurasian marginal seas is correlated with this shift.

There has been a clear transition from solid to liquid precipitation, leading to more ROS events in spring, although large discrepancies were found between different reanalysis products. In ERA5, the shift from solid to liquid precipitation phase during the early melt season has directly contributed to a reduction in spring snow depth on sea ice with the largest contribution in the Kara-Barents Seas and Canadian Arctic Archipelago.