The 2018 melting season

[Shared Humanity]

Sorry Jim. My fault.

[BenB]

This is still a bit far out to be reliable, but if it's close to true, it should get melt-pond May going strongly, with temperatures above zero across large swathes of the Arctic:




A 3.6 degree anomaly would be very unusual for this time of year - perhaps unprecedented? Either way, temperatures are forecast to be high over the coming few days, with the anomaly growing day by day, so we should see some acceleration in melting over the coming week.

[jdallen]

This is still a bit far out to be reliable, but if it's close to true, it should get melt-pond May going strongly, with temperatures above zero across large swathes ....

Revisit this in two days and see how it shifts.

[Random_Weather]

jdallen,

Correct, GFS is strongly warm biased with time, since years i watch this, therefore i think its an issue with snow cover..

also see: http://www.karstenhaustein.com/reanalysis/gfs0p5/ANOM2m_arctic/verification/ANOM2m_bias_past07_arctic.html

to the +168h forecast, there is a massive warm bias to the past 7 days

Also evident over the full northern hemisphere: http://www.karstenhaustein.com/reanalysis/gfs0p5/ANOM2m_mollw/verification/ANOM2m_bias_past07_mollw.html

[dnem]

Incredibly useful graphic, Random.  Thank you for posting.

[Neven]

Definitely useful. GFS as usual running hot beyond 3 days out.

As for ECMWF sea level pressure, D1-6 looks just a tiny bit worse than it did yesterday, all days are 1030+ hPa. But on the flip side D7-10 is looking up, even though high pressure still dominates the CAB (but we try not to look too much beyond Day 6, so I'm not posting it):

[gerontocrat]

Longyearbyen, Svalbard weather forecast attached. Is this warm for the time of year, or just average?

[uniquorn]

The wave forecast got a bit stronger for thursday north of Svalbard

[Aluminium]

Longyearbyen, Svalbard weather forecast attached. Is this warm for the time of year, or just average?

In Barentsburg the weather is much warmer than average. I use this site (in Russian). The average temperature is normally below zero until the end of May.

[JayW]

Beaufort Sea May 9-13, 102 hours. Contrast boosted for detail, I also find it makes it easier to distinguish the clouds from the surface.

http://feeder.gina.alaska.edu/npp-gina-alaska-truecolor-images?search%5Bfeeds%5D%5B5%5D=1&search%5Bsensors%5D%5B3%5D=1

[BenB]

I'm well aware of the problems with GFS at those time scales, but 3.6 seemed pretty extreme for this time of year, even by its standards. Anyway, the point was more that the next week looks consistently warm, and even subtracting a 1.6 correction factor for 168 hours, you get a 2 degree anomaly. Speaking of which, the more reliable 3-day forecast is for a 2.2 anomaly:

[Neven]

Ben, you may be right. There may be something there. Another heat pulse, while high pressure keeps dominating, would be another early blow to the ice.

[BenB]

And I may be wrong  But May seems interesting this year - whereas in recent years DMI has tended to follow the mean temperature very closely at this time of year (or even dip below it), this year we've stayed consistently warmer. Now this heat pulse may be coming at a time when temperatures and insolation are high enough to cause some real damage, and potentially get the melt season in the CAB off to the fast start that it hasn't really seen recently in terms of melt-ponding and pre-conditioning the ice.

At the same time, we have very high sea surface temperature surface anomalies, which if they continue to build will cause damage at the end of the season. It will be interesting to watch how things develop.

[S.Pansa]

Here is another model-view on the things that might come to the Arctic. The first pic compares ESRL & GFS forecasts for temps and SLP.
In agreement with the ECMWF-forecast posted by Neven above they show the high pulling back to the Beaufort. That could make room for a little heat intrusion over the ESS and Laptev. The forecasted SLP is in good agreement, the surface temps differ slightly 

The two other pics show the predicted melt pond fraction. First pic from today has some ponds in the Chukchi/Beaufort. The second shows the possible consequences of the predicted little heat wave: widespread meltponding in the ESS/Laptev.

Bad news for the ice? Who knows. Fortunately the SPIE of the great late professor Slater is back online (maintained now by Andrew Barret's team at NSIDC as I gather from this guest post at the ASIB).
Will be interesting to see if this forecast materializes and how this will effect Slaters model.

PS: A late thanks to Dryland for the effort and the link to his site, wehre he has put together the most interesting ESRL model outputs

[Neven]

Beaufort Sea May 9-13, 102 hours. Contrast boosted for detail, I also find it makes it easier to distinguish the clouds from the surface.

And we'll have at least another week of this... Spectacular images, but bad for the ice.

[Shared Humanity]

Is that clouds or is there a little bit of refreeze in the Beaufort as the ice fractures and separates?

[FishOutofWater]

The Greenland low pressure vortex, Arctic ocean high pattern will continue for the next week according to the ECMWF model. This weather pattern enhances transport of both atmospheric and ocean heat from the subtropics in the Atlantic towards Europe's subpolar seas. This pattern has already rapidly warmed the waters in the Norwegian and Barents seas in the first two weeks of May.

Now it will be warm enough and the sunlight will be intense enough for melting of snow on top of the ice to begin under the dome of high pressure over the Arctic.

Slater's model is incorporating this pattern into its prediction of an ice area/extent cliff in the first week of July.

[numerobis]

Looks to me like the high Arctic is getting hammered this May, whereas the lower-latitude Arctic is running a bit cool and cloudy.

Down here it'll all melt out regardless.

[Lord M Vader]

Wrt to Slaters model which foresees an SIE of 9,18 Mn km² by July 4. It's worth to remember that the SIE for July 4 last year was 9,174 Mn km².... A number of 9,18 is just a fourth place.

The main question is whether the high pressure is able to remain in the CAB and adjacent areas through June. The GFS weekly forecasts hints of a high pressure dominated weather over the Siberian side by the end of May and beginning of June. That would be a rather interesting set up as the ice this year is considerably thicker on the Siberian side while thinner at the North American side and close to the pole.
The forecast for week 4 (June 5-11) is very far away but GFS have been rather consistent for the last few days and believes in a continuation of HP over the Siberian side.

[Gray-Wolf]

I think once High Pressure is established changes occur only slowly. Lows are far more 'dynamic' and so more difficult to forecast for any area?

Should there be a physical reason why we see HP dominance over winter in the north Atlantic across low solar then it would be interesting to see if such a forcing maintains in summer but migrates further north with the changing of the seasons?

Any 'extra push' toward H.P. over summer in the basin would be most unwelcome!!!

[romett1]

Small update about FDD (Freezing Degree Days) anomaly north of 80° N between Jan 01 and May 14. May has been significantly warmer than 2017 and looking at the weather forecast it is possible that 2016 level (since start of the year) may be reached. Image: https://sites.google.com/site/cryospherecomputing/fdd

[JayW]

Is that clouds or is there a little bit of refreeze in the Beaufort as the ice fractures and separates?

There's certainly low clouds, especially today.  I wouldn't rule out some freezing over of leads locally, but it's my opinion that it's mostly clouds, perhaps entirely.

Here's the "landcover" band from VIIRS.  May 12-15, ~81 hours

http://feeder.gina.alaska.edu/npp-gina-alaska-landcover-images?search%5Bfeeds%5D%5B6%5D=1&search%5Bsensors%5D%5B3%5D=1

[FrostKing70]

Interesting.   Those are some large gaps developing....

[Michael Hauber]

Visually on MODIS, I cannot find nearly as much red (surface melting) in the channel 3-6-7 as 2016 at same date.  But a lot more cloud - foggy conditions under high pressure?  Not sure if this fog is hiding surface melting or if its reflecting sunlight and delaying the start of surface melt.

I'm sure temps are warm, and melting early, compared to the longer term average.  But compared to recent years?  Not sure.  But I can't imagine it being at the cooler end, either middle or warmer.

[A-Team]

Up-forum, the NSIDC's sea ice age comparison for 'week nine' of 1984/2018 includes the Barents, Kara and even White Sea [Белое море] as part of the Arctic Ocean whereas we are primarily exercised with the Arctic Ocean proper: north of the Svalbard - FJL - SZ line (or a bit more inclusively north of 80ºN) in this region.

http://nsidc.org/arcticseaicenews/files/2018/05/Figure_4ad_correctedV2-985x1024.png

Their graph makes its point well enough: the older ice classes are pinching out to zero as first and second year ice come to predominate. However what is the quantitative effect of diluting the AO ice class areas with these peripheral seas that never have so much as SYI?

The Barents etc comprise some 31% of the total area but as the consolidated graphic below shows, it was almost entirely open water already in 1984 on this date, The Kara was filled with FYI in both years. However there's tremendous variability in the Barents ice cover and 1984 is not all that representative.

Thus an overly broad definition of 'Arctic Ocean' contributes quite a bit of noise to the ice age trend, just as it does to extent, area, concentration and volume trends. At some venues (such as Piomas), individual sub-regions have been dissected out and plotted separately. Elsewhere, whole northern hemisphere tracking remains at cross-purposes with near-term prediction.

Note open water (0YI?) at 'week nine' is not allocated space on the NSIDC graph. That really should be fixed because even restricting to the Arctic Ocean proper, 0YI now occurs all year above Svalbard (not just during mid-winter 'week nine' of 2018). Indeed, Jaxa shows open water there in all years archived, back to 2003.

Open water, averaged over this date for 14 years, is shown in last frame. No trend is seen (or expected) because wind-blown ice can easily predominate in the northern Barents.

Up-forum, I've explained how and why NSIDC should refine the FYI and SYI ice age classes down to  a monthly basis (or age since attainment of 0.5m and 1m SMOS thicknesses). Tracking ice provenance is quite important to understanding melt season susceptibility.

NSIDC makes these ice age classes by tracking parcels in Ascat imagery (ie radar roughness development), so I've attached the 2018 time series from day 1 to day 63 (Mar 4th) through day 135 (May 15), along with the Jaxa counterpart. Ascat data has had a very bumpy ride to NOAA archives lately so data gaps in MetopA are repaired here with MetopB in the mp4 below.

Jaxa is complementary to Ascat in terms of wavelength and polarity but provides less surface feature acuity. That's offset by its much richer channel dimensionality and potential for detecting snow melt, melt ponds, drained melt ponds, or wave-washed floes.

While fiducial training sets are in short supply, a peculiar region north of FJI showing persistent pink (resp. green in inverse; final rest frame) for the last five days -- ruling out weather artifacts -- deserves further interpretative scrutiny. It is shown as thin ice in the PR89 polarization ration product at Jaxa.

https://ads.nipr.ac.jp/vishop/#/monitor

http://www.ospo.noaa.gov/data/atmosphere/ascat/MetopA/ICE/msfa-NHe-a-2018135.sir.gif

[numerobis]

Local news mentions Frobisher Bay’s flow edge is closer to torn than usual:
http://nunatsiaq.com/stories/article/65674floe_edge_near_iqaluit_thinner_and_closer_than_usual_canadian_/

Along with generic ice-is-thinner-and-younger that we all know.

[Telihod]

I wonder if the ice north of Greenland could melt completely this year (if it is not replaced from somewhere else), since positive temperatures are a regular thing there (and the ice is thinner than usual)?

[subgeometer]

I think we'll see signs of melt ponding soon on the fast ice by the the ESS coast, maybe starting in the big bay. The ECWMF forecast shows well above freezing temps at 925hPa all week, as well as at the surface. The GFS also has Siberia warming a lot

[JayW]

925mb reanalysis temp anomaly from April 14-May 14, 2018.

https://www.esrl.noaa.gov/psd/data/composites/day/

[charles_oil]

JayW - it looks like the temp scale maxes out at 5+ degrees on the 14th temp anomaly so I imagine there are contours of even greater anomalies within the bullseyes too....

[Shared Humanity]

I wonder if the ice north of Greenland could melt completely this year (if it is not replaced from somewhere else), since positive temperatures are a regular thing there (and the ice is thinner than usual)?

While temperatures are warm and the ice is thin, it is the migration of ice that will be the key factor. Paradoxically, increased Fram export which is bad for the ice is more likely to push ice up against northern Greenland, preventing an ice free coast from setting up.

[A-Team]

The content-rich Jaxa site has the cleverest design of all the Arctic data portals that we use. It is written in Angular.js; the screenshot below shows the available options. The view is 'Bering Strait down', or 135º cw from our usual 'Greenland down' orientation.

https://ads.nipr.ac.jp/vishop/#/monitor

The first mp4 shows the first 135 days of an ambitious triple product: land snow pack thickness, open water sea surface temperature and sea ice concentration. The latter was far too faint in the original and so had to be isolated and contrast-enhanced separately. (To make the land mask, grays from mid-summer low-snow days were combined until the mask was entirely snow free.)

Snow coverage drops precipitously near the end of the time series. Snow-on-land has its own forum as it is only indirectly relevant to sea ice melt season through its loss-driven warming effect on land surface temperature and thus on local air temperature (there being a vast land area in the northern northern hemisphere impacted). Both land snow and sea ice are important reflectors back to space of incoming solar radiation.

The second mp4 shows the first 136 days of Jaxa ice thickness. While this may have some merit, there's no immediate link to methodology and some artifacts of passing weather can be seen.

[VeliAlbertKallio]

Another paradoxy is that when overall ice covered part of the ocean is getting smaller, the remaining ice can be more easily redistributed by storms and winds to unusual places. A thin and soft ice can also pack up like harmonica. So we could have all sorts of scenarios as the jet streams have weakened and the winds occasionally push air across the North Pole from the Atlantic to the Pacific. But any such oddities are likelier towards end of season if ice area is small. 

I wonder if the ice north of Greenland could melt completely this year (if it is not replaced from somewhere else), since positive temperatures are a regular thing there (and the ice is thinner than usual)?

While temperatures are warm and the ice is thin, it is the migration of ice that will be the key factor. Paradoxically, increased Fram export which is bad for the ice is more likely to push ice up against northern Greenland, preventing an ice free coast from setting up.

[A-Team]

ice cover getting smaller, more easily redistributed by winds to unusual places. A thin and soft ice can also pack up like harmonica. the jet streams have weakened and the winds occasionally push [ice] across the North Pole. soft ice compresses like an accordion.

Right. The relationship between wind strength and ice movement is indirect, mediated primarily by 'form drag' or the frictional coupling between rough surfaces such as compression ridges and floe edges, summing with ocean-ice 'form drag' from keels and bottom skin. Recent papers by P. Lu and A. Petty provide details on the physics there:

https://scholar.google.com/scholar?client=opera&oe=UTF-8&um=1&ie=UTF-8&lr&cites=3522928423510465463

http://alekpetty.com/papers/

It is very difficult to get at those parameters, much less compare them year on year or predict ice movement in advance, so for our purposes it suffices just to look at net ice movement itself. The two main tools for gridded daily displacements are Jaxa and OSI-SAF sea ice drift products.

The first 135 days of 2018 ice motion are compared below. The Jaxa product has wind speed coloring so a static png can give an overview of both storm events and local sites of motion. OSI-SAF shows dodgy ice and derives better vectors from a two-day motion determination. Ascat, up-forum, provides the best summation of these daily motions and the deformations of large-scale trackable features.

The Arctic Ocean ice pack, even in the dead of winter, does not move as a rigid body (restricted to translations and rotations). On the contrary, forcings are very unevenly applied and the resulting motion is regionally distinct: in 2018, quite pronounced off the Alaskan coast, episodic out the Fram, persistent and noteworthy south across the Svalbard-FJL line, westward along the Siberian coast, but with hardly any motion at all (since mid-September!) in the central Arctic. 

As Veli notes above, thermodynamic considerations alone are insufficient for an understanding of melt season and have zero prospects for predicting September minimums or overall timing of ice disappearance.  Passive intra-basin transport of ice to warmer or sunnier locations can play just as important a role as storm fragmentation of floes or conventional out-of-basin export.

[gerontocrat]

I know A-team thinks the DMI North of 80 graph is not significant but despite that here it is.
Later than 2017 but is the above average temperature trend coming to a halt?

[Neven]

Not if the GFS temperature forecast is correct. Most of the positive anomaly is centred around the Pole:

[A-Team]

>the DMI North of 80 graph is not significant (?) but despite that

80ºN is kinda Dane-o-centric whereas the 67ºN Chukchi-o-centric is needed too. I was thinking blizzard92 has some way of making contoured mean temperature maps for the whole AO, as well as anomaly relative to post-2010? If that could be done on a rolling monthly (semi-weekly?) basis instead of roman calendar months ... ditto for EMWF 7-day forecast  ... ditto mslp and winds, then we could all go home, the idea being using the anomaly forecast as a multiplicative overlay on concentration to determine coming at-risk areas.

https://twitter.com/ZLabe/media

[Neven]

The ECMWF SLP forecast is looking a bit worse again now, with high pressure remaining relatively high and quite extensive. D7-10 has the high pressure moving over to the Siberian side of the Arctic, but forecasts that far out tend to be volatile, and so there's no use in posting them. Here's D1-6:

[A-Team]

Here is the last of the Jaxa offerings, PR89 (radar polarization ratio at 89 GHz). The mp4 shows the first 137 days of 2018 as contrast-enhanced UL, daily difference UR, glow contrast-enhanced LL, and sepia daily difference LR.

The artifacts seem just overwhelming -- both weather and furrowing arcs -- and it's never been entirely what PR89 is good for as a standalone tool. However 'PR89 polarization ratio' provides a distinctive google scholar search term, yielding at two rounds the following titles for recent years (and the conclusion use SMOS products):

Estimation of thin ice thickness for coastal polynyas in the Chukchi and Beaufort Seas
Estimation of thin ice thickness from AMSR-E data in the Chukchi Sea
Thin ice detection in the Barents and Kara Seas with AMSR-E and SSMIS radiometer data
Thin sea ice identification in the Kara Sea using AMSR-E data
A modular ridge randomized neural network applied to the estimation of sea ice thickness
An assessment of sea-ice thickness along the Labrador coast from AMSR-E and MODIS
Improvement and sensitivity analysis of thermal thin-ice thickness retrievals
On the accuracy of thin-ice thickness retrieval using MODIS thermal imagery over Arctic FYI
SMOS-derived thin sea ice thickness: algorithm baseline
Estimation of sea ice freeboard from SARAL/AltiKa data
Sea ice production and water mass modification in the eastern Laptev Sea
Observations of supercooled water and frazil ice formation in an Arctic coastal polynya
Lead detection in Arctic sea ice from CryoSat-2: quality assessment
Long-term variation in sea ice production and intermediate water in the Sea of Okhotsk
Global view of sea-ice production in polynyas and its linkage to dense/bottom water formation
Spatio-temporal variability of polynya dynamics and ice production in the Laptev Sea
Near-bottom water warming in the Laptev Sea in response to atmospheric and sea-ice conditions
Origin of freshwater and polynya water in the Arctic Ocean halocline in summer 2007
Helicopter-borne observations with portable microwave radiometer in the Sea of Okhotsk
Sea ice surface temperature estimation using MODIS and AMSR-E data along the Labrador Coast
Impact of Laptev Sea flaw polynyas on the atmospheric boundary layer and ice production
A wind‐driven, hybrid latent and sensible heat coastal polynya off Barrow, Alaska
Changes in distribution of brine waters on the Laptev Sea shelf in 2007
Impact of Siberian coastal polynyas on shelf‐derived Arctic Ocean halocline waters
Variability and trends in Laptev Sea ice outflow between 1992-2011

[Greenbelt]

The ECMWF SLP forecast is looking a bit worse again now, with high pressure remaining relatively high and quite extensive. D7-10 has the high pressure moving over to the Siberian side of the Arctic, but forecasts that far out tend to be volatile, and so there's no use in posting them. Here's D1-6:

This morning's GFS has the big central arctic 1030mb dome weakening slightly and sliding toward Canada, with lots of low pressure from Greenland to northwestern Russia. Longer range seems to have strong surface high between Iceland and Scandinavia, and also big high building in far northeastern Siberian arctic coast.

[gerontocrat]

Global Ocean Heat Content from NOAA updated for 1st quarter 2018. Not a surprise that the only way is up.
https://www.nodc.noaa.gov/OC5/3M_HEAT_CONTENT/

[subgeometer]

I know A-team thinks the DMI North of 80 graph is not significant but despite that here it is.
Later than 2017 but is the above average temperature trend coming to a halt?

The forecasts (GFS, ECWMF) are showing another lurch upwards about to begin, with very warm air intruding via the Laptev and ESS  especially after midweek. Maybe 0C wiil be reached a few days early this year if the trend continues

[Alexander555]

https://robertscribbler.com/2018/05/20/potential-historic-arctic-warming-scenario-in-the-gfs-model-forecast-for-late-may/

[Shared Humanity]

Global Ocean Heat Content from NOAA updated for 1st quarter 2018. Not a surprise that the only way is up.
https://www.nodc.noaa.gov/OC5/3M_HEAT_CONTENT/

Curious about the different slopes of the curve for 0-700 m and 0-2000 m. Does the more rapid warming of 0-2000 m suggest an efficient heat transfer or mixing to depth?

[FishOutofWater]

Multiple things are going on in the depths of the ocean that explain the 2000m heat content increase. The subtropical gyres have expanded polewards in both hemispheres. Even a small northwards displacement of the north wall of the Gulf Stream involves massive amounts of heat. There's a still a thermal gradient below 700m depth.

The coldest water in the oceans, Antarctic bottom water is not forming like it used to. The Weddell sea polynya hasn't been active for decades. As less bottom water forms in both hemispheres the oceans warm in the deepest levels as the volume of the ocean's coldest water decreases with time.

In addition, new bottom water is warmer than old AABW.

[bbr2314]

I would think Barentz' imminent total melt-out is going to create a bit of a mini-cliff for area/extent #s. It has been absolutely battered by incoming heat thanks to Scandinavia's melt prior to Siberia / Alaska/ Quebec (at least I suspect this has something to do with it). Without snowpack to modify incoming airmasses, the torch has been sustained and worsening, and this should only continue.

ATL SSTs are running incredibly warm along the northern edge of the Gulf Stream, and there is plenty of heat content available for transport broiling up off the US Eastern Seaboard.

In any case, Barentz' #s are currently being held up by a slush of .25-.5M ice that is not going to resist the next two weeks (IMO). With the actual front E of Greenland already extremely far north, the entire front may take on a rather odd appearance come 6/1-6/15 as Kara holds on, while the main advancing melt front begins taking out chunks of the actual CAB.

HYCOM also shows Hudson Bay's NW edge giving out, and satellite seems to somewhat confirm this, but I don't think things will progress rapidly there as the ice is thick and largely still snow-covered. Both Hudson and Kara may prove resilient this summer vs. normals (but will still melt out entirely by August at the latest). Maybe we see a particularly potent end-of-season drop as well?

[Alexander555]

Is the Ice in Hudson Bay so thick ? If you look on the picture from Nasa Worldview, in the middle it looks brite white. But the west side has a different colour. A little bluish. Or is it a light effect.

And does somebody knows, if you take the area of Hudson Bay. Places like the Hudson Strait and Fox Basin. Are they included in the Hudson Bay area.

[oren]

And does somebody knows, if you take the area of Hudson Bay. Places like the Hudson Strait and Fox Basin. Are they included in the Hudson Bay area.

AFAIK these locations are indeed included in Hudson Bay data.

[bbr2314]

Is the Ice in Hudson Bay so thick ? If you look on the picture from Nasa Worldview, in the middle it looks brite white. But the west side has a different colour. A little bluish. Or is it a light effect.

And does somebody knows, if you take the area of Hudson Bay. Places like the Hudson Strait and Fox Basin. Are they included in the Hudson Bay area.

I believe it is quite thick and Canucks confirm.

http://iceweb1.cis.ec.gc.ca/Prod/page3.xhtml



The west side is definitely beginning to show some melt but most elsewhere is holding firm. This will begin to change over the next 30 days, of course.

It should also be noted that the above chart includes a handy 7-day mean normal vs. actual temp guide on the lower-left for adjacent Canuck weather stations. All are handily or exceedingly below normal, at least in the week preceding 5/14.

[bbr2314]

Here is 2017 vs 2018 (2018 is... extremely... brighter/whiter)