Greenland 2014 Melt Season

[crandles]

Melt area is now well above normal and will likely go much higher and stay above normal for the foreseeable future.

well above?

2012 had reached 30% by this time of year so I am not sure that 10% vs long term average of 9% should be seen as 'well above'. Or did the 2012 graph need redoing on change of methodology?

[RaenorShine]

I wouldn't go as far as well above normal, unless you count the previous week as well below normal, its just normal seasonal variability so far. NSIDC area melt is just above the seasonal average, but not even record melt levels for this year yet (the peak in May was higher), not everything is a 'Mother of God' moment!

DMI shows some precipitation in the south, and the daily SMB going back to around zero.

[Frivolousz21]

The models showing the melt layer reaching 2000-2500FT at times this week over GIS.

With lots of sun.

It's going to reach the 20-30% this week at least for a time.

They have been showing for two days now a big warm up in the late short range.

In two days it will at least be 20% I say single day.

Before NSIDC went out today it was well above normal.  Did you guys not see this?

It was on a straight up trajactory and the real warmth starts

[Frivolousz21]

Then it just gets nasty.

[RaenorShine]

Before NSIDC went out today it was well above normal.  Did you guys not see this?

It was on a straight up trajactory and the real warmth starts

I think we need to be careful to differentiate between fact (measurements), comment and forecasts. From your posts I'd have stuck with "NSIDC melt area has now climbed above the seasonal average, but is likely to increase further in the coming days if the weather forecasts hold", but I do appreciate your colourful style!

The melt area was on a upwards trajectory from a below average value, it may continue upwards over the next week or so to a 'well above average' value but it was not there yet.  You wouldn't call the first 22C temperature against a 20C average 'well above average' or 'what a scorcher', so we shouldn't with this.  If your forecasts hold, we'll be there soon.

Looking at the data on Polar Portal, we seem to be below average maximum SMB for the season to date (510Gt Vs 540Gt maximum gain, from reading off the graph).  The main reason for that though is not melt, but the lack of precipitation through May. melt has consistently looked more patchy than average.

Over the past week, the albedo anomaly vs 2000-09 has fallen across much saddle (compare May 18 to the latest slide of 31 May), and is even showing as positive along the east and west coasts (meaning the ice sheet is whiter than normal, i.e. less melt)

2012 continued to grow to around 540Gt SMB max gain until late May when it started to fall.  We are now equal to the YTD value for 2012, but the ice sheet is likely to be in a far better shape as we have had no significant melt to date.  Even with a warm spell I think we are going to struggle to keep up with 2012 as it starts it's decent.

Given a couple of weeks of good melting weather this could change, but we'll see how it goes. I'm one of the forum members that struggle to read the weather charts you post. I rely on the commentary to interpret them, but do take any of the longer range ones with a pinch of salt as things so often change.

I've concentrated on posting what I see happening in the graphs and maps to increase my limited understanding rather than looking forward. Doing these daily postings has helped my understanding of what is going on a lot.

To that end, DMI shows the precipitation has moved away, and the daily SMB is down again by around 1Gt, melt seems to be increasing around the southern edges of the sheet.

NSIDC is undertaking maintenance for a couple of days, and we'll see where that is tomorrow evening.

[Frivolousz21]

I can help you with that a bit.

It's almost certain if the NAO is negative GIS is warm and sunny.
It has been positive during the period of cooler times.

It's about to change.

[JayW]

I'm an amateur, but it looks like the -NAO (according to the GFS) would initially be east based, before losing it's grip. Until what looks like a traditional Greenland block attempts to establish itself in 2 weeks.  I'm not necessarily endorsing it, just reporting, but I'm posting it for comparison later (verification).

http://www.cpc.ncep.noaa.gov/products/precip/CWlink/MJO/block.shtml

[RaenorShine]

NSIDC is still down for maintenance.

DMI shows heavy melt on the western Greenland periphery on 5/6.  The precipitation is back though in the south, and overall SMB has Yo-yo'd back to close to zero.

[RaenorShine]

NSIDC is back up, but has not updated yet.

DMI shows a storm dumped a load if snow and rain over the south, leading to the largest daily SMB gain in over a month of 4Gt. Melt has increased around the coasts though.

[Frivolousz21]

The freeze level was like 2000-2500M.

Is there actual conformation of snow?

[RaenorShine]

@Greenlandicesmb (this account is linked to the DMI Polar Portal) tweeted this yesterday

Large snowfall in south more than balances melt in west, #SMB not always straightforward!

The southern dome reaches around 3000M according to Wikipedia, so that is above your freeze level. Even if it rains, on the higher altitudes this will be freezing on contact with the sheet.

NSIDC still has not updated, I guess it will be Monday before this is fixed now.

DMI shows more precipitation yesterday, with another 3Gt SMB daily gain, the last two days have more than compensated for the recent small losses, and we are now at the YTD highest SMB balance, above the 2012 value for this date, but still below the 1990-2011 mean.

[Frivolousz21]

You are right but they have that snow cover right up to the edge of the ice line.

There are mountains really close to ice line on the SE side that quickly jump above 1500M but there should of been melt below that without a doubt.

That is a hallmark good melt pattern for GIS.  It's June 8th now.  I can't see how snow is accumulating or water freezing on contact with the ice anywhere below 2000M with that kind of flow as well as the sun being strong now.  Even if it's cloudy there is still going to be extra insolation.

GIS is forecasted to be torched starting tomorrow with a large ridge and freeze levels rising to near 3000M at times around GIS.

I don't think the older years are updated right.  If this was the old model it would be showing large melting so far.

2013 should not be well lower then 2014 going by the new model at this time.

I don't know.  Never the less the models show major wall to wall torching up coming.

Freeze level reaching 3000M a couple days.  Anything at 1500-1800M or below should have 24 hour melting those days or 18-19 hour.

[RaenorShine]

I think this is probably the last gasp of the freeze season. I can see clear areas of red (meaning negative SMB) along the south and west coast. Looking at the weather map http://polarportal.dk/en/weather/ it show winds coming off the ice sheet to the east. Could this be something in my to do with the difference? I.e the west is cooling the warm air and melting, but on the east it has cool air flow from the ice sheet.

Remember also that the new model doesn't count the loss until it leaves the sheet. Having very wet snow could well in increase the melt in coming days. I would not be surprised from your analysis to see that we will have a fairly high melt area once NSIDC comes back on line.

[RaenorShine]

NSIDC has still not updated.

The DMI shows the precipitation in the south has moved away yesterday.  The daily SMB gain has lowered to around 0.5Gt, and melt can be seen around the Southern Peninsula.

[RunningChristo]

Currently Close to +18 C at Kangerlussuaq, western Greenland, so today the rivers will soar and albedo fall as a rotten tomato to the ground .

http://www.dmi.dk/groenland/maalinger/vejret-lige-nu/stations/vis/4231

[Frivolousz21]

There is currently no end in sight on the models for the GIS warm spells.

Wall to Wall -NAO with above average heights and sea level pressure.

[RaenorShine]

NSIDC has still not updated.

DMI SMB has continued it downward trend to a 1Gt loss. with increased melt in the south, and light precipitation being limited to the west coast.

The albedo anomalies on the southern peninsula shows a noticeable increase in reflectivity along the west coast on the 7th (when the storm hit, and the latest map to be shown) in comparison to the values from the 6th. The albedo anomoly on the SW coast is still positive vs the 2000-9 average, but has dropped down several percentage points.  In a couple of days this should start to increase again, it will be interesting to see how long it takes to get back to the same values as the 7th.

[Frivolousz21]

A brutal assault on GIS starts tomorrow.

Without a massive pattern change the GIS melting season is about to explode everywhere.  It has likely been above normal on NSIDC the last wee.

With this forecast coming anywhere near accurate that will lead to way above the 2SD mark.  Probably pulling 50%+ days.

[RaenorShine]

NSIDC is still down.

DMI show a lot more melt (both ablation and melt around the edges of the sheet) in the south, but this has been offset by precipitation in the North West, leaving a net daily SMB of zero.

@greenlandicesmb have highlighted a new paper in Geophysical Research Letters about firn (partially melted, compacted and refrozen snow - http://en.wikipedia.org/wiki/Firn) . @greenlandicesmb said:

Firm warming effects of refreezing really important to mass budget. Our model includes it, but only few observations

Here's a link to the abstract.

http://onlinelibrary.wiley.com/doi/10.1002/2014GL059806/abstract

Field measurements of shallow borehole temperatures in firn across the northern Greenland Ice Sheet (GIS) are collected during May 2013. Sites first measured in 1952-1955 are re-visited, showing long term trends in firn temperature. Results indicate a pattern of substantial firn warming (up to +5.7C) at mid-level elevations (1400-2500 m) and little temperature change at high elevations (>2500 m). We find that latent heat transport into the firn due to meltwater percolation drives the observed warming. Modeling shows that heat is stored at depth for several years and energy delivered from consecutive melt events accumulates in the firn. The observed warming is likely not yet in equilibrium with recent melt production rates, but captures the progression of sites in the percolation facies toward net runoff production.

[solartim27]

Any idea what the line to the north of Jacobshavn is?  Melt water channel?  Seemed to form in a day or two.
http://lance-modis.eosdis.nasa.gov/imagery/subsets/?subset=Arctic_r02c02.2014162.terra

[RaenorShine]

Solartim27, the big black line is a satellite image join. If it's any other line I'd suggest posting in the jacobshavn thread also.

NSIDC melt area is finally back on line. It shows the melt area for yesterday at around 16%. Looking at the graph it climbed slightly higher than today's value at the end of last week before falling back to the seasonal average (because of the storm?). It's climbing again now, and melt is still increasing.

[solartim27]

I should say it formed in about a week, since the last clear day was 6/2.  It is not from the satellite, it could be a regular feature, like a canyon or chasm.

[RaenorShine]

Yesterday in DMI SMB model,  melting increased particularly in the South East along the coast, and the precipitation in the NW has eased slightly. This looks to have led (just) to the greatest SMB loss so far this year of just over 1Gt.

[crandles]

I should say it formed in about a week, since the last clear day was 6/2.  It is not from the satellite, it could be a regular feature, like a canyon or chasm.

On 11 June there seems to be a single strip (perfectly straight) taken at a different time to the rest. I also see a cloud feature. Neither appears on 10 June image nor can I see any other feature 'north of Jacobshavn' that looks like a black line.

Is it present on 10 June as well as 11 June? Can you indicate more clearly where on the image you are talking about?

[icefest]

@cradles, I'm fairly sure the feature is clouds. It moves between terra/aqua and in 3-7-6 you can see the shadow.

http://forum.arctic-sea-ice.net/index.php/topic,154.msg28289.html#msg28289

[RunningChristo]

More than 20 C in Southern Greenland today, for the first time this summer:-).
http://www.dmi.dk/groenland/maalinger/vejret-lige-nu/stations/vis/4271

[Frivolousz21]

Small melt ponds are visible on the Western edge of the ice sheet.

[RaenorShine]

Two YTD records to report today.

Firstly, NSIDC have melt area at almost 20% for 11/6, with melt around the whole coast.

DMI have the daily SMB loss of 2Gt for 12/6, doubling the previous record daily melt totals.

[RaenorShine]

Melt intensified again yesterday.

NSIDC melt area is now at 20% (12/6).

DMI daily SMB increased again to over 2Gt(13/6). Melt has intensified in the west, but some light precipitation fell along the south east coast.

[RaenorShine]

NSIDC did not update for 13 June.

DMI shows the precipitation has moved north along the east coast and increased. This has been more than made up by heavy melt in the south, leading to a 3Gt daily SMB loss.

[RaenorShine]

No NSIDC update again today.

DMI polar Portal Shows the melt intensifying, especially down the west coast. for the net daily SMB however this has been offset by precipitation in the NW and SE, leading to a slight retreat from yesterdays highs.  It's still well above 2Gt loss for the 15/6 though.

[Rubikscube]

Am I the only one who thinks the DMI maps for the past week look a little suspicious? In one place there appears to melting all week which is continuously intensifying as temperatures gears up, but just a notch further inland there is no melt. As a result the border between intens melt and no melt becomes suspiciously abrupt and immobile, and looking at Polar Potal's archives one can see a similar phenomenon on the 9-15th of July 2012 just much further inland.

I expect this to be a model artifact (especially those pics from 2012 look bugged), but I see no obvious reason to why it should be so difficult to model surface mass loss further inland. Has DMI tried to calibrate its model this year, but seemingly ended up with an even narrower "melting zone", or is the surface melt really progressing that slowly towards the interior? I do expect that NSIDC will be showing surface melt much further inland on June the 15th.

[RaenorShine]

Am I the only one who thinks the DMI maps for the past week look a little suspicious? In one place there appears to melting all week which is continuously intensifying as temperatures gears up, but just a notch further inland there is no melt. As a result the border between intens melt and no melt becomes suspiciously abrupt and immobile, and looking at Polar Potal's archives one can see a similar phenomenon on the 9-15th of July 2012 just much further inland.

I expect this to be a model artifact (especially those pics from 2012 look bugged), but I see no obvious reason to why it should be so difficult to model surface mass loss further inland. Has DMI tried to calibrate its model this year, but seemingly ended up with an even narrower "melting zone", or is the surface melt really progressing that slowly towards the interior? I do expect that NSIDC will be showing surface melt much further inland on June the 15th.

DMI have modified their model this year, and rerun previous data to match. It is a model however so artifacts are probable.

I would not be surprised of increasing melt on NSIDC, which is a measurement (rather than a model).  DMI does not show all melt as its a SMB calculation (rather than melt). It takes into account that surface melt initially remains as wet snow, and not all runoff reaches the coast.

The model has been updated in 2014 and now gives a better picture of what happens with the meltwater. Earlier a large amount of the meltwater was treated as loss in the form of runoff from the ice sheet. The new model is better at taking into account the part of the meltwater that refreezes on its way to the coast, and this then remains a part of the ice sheet.

They've stated that the map will be less red than before, but I agree that a harsh edge like is showing at the moment is suspicious unless there is a topographical reason for the boundry.  It may be that we are seeing the boundry of the models regions (ice sheet edge/interior) and it treats these differently at different times of the year.

I have thought that having a breakdown of the different processes (melt, ablation, refreeze, precipitation etc) would be a benefit as we could see under the bonnet.

[Rubikscube]

The model has been updated in 2014 and now gives a better picture of what happens with the meltwater. Earlier a large amount of the meltwater was treated as loss in the form of runoff from the ice sheet. The new model is better at taking into account the part of the meltwater that refreezes on its way to the coast, and this then remains a part of the ice sheet.

They've stated that the map will be less red than before, but I agree that a harsh edge like is showing at the moment is suspicious unless there is a topographical reason for the boundry.  It may be that we are seeing the boundry of the models regions (ice sheet edge/interior) and it treats these differently at different times of the year.

Thank you RS, I was not fully aware of those updates although you might perhaps have mentioned them before. I doubt very much there is a topographical reason for the distinct edge and lean more towards the idea of a model boundary, however, when taking a closer look, the edge appears to be somewhat mobile despite its suspicious distinctness (at least more mobile than it was in 2012). My theory would be then that the boundary is, as you suggest, created because the DMI model ignore substantial surface melt until it assumes that the surface layers are so soaked that ponds and rivers starts forming and thus transporting some of the meltwater away through crevasses and moulins instead of it just refreezing together with those top layers of snow.

[Andreas T]

not sure whether this belongs into another thread, but I have noticed this at other times and places: in the image below (Keyser Franz Josef Fiord I believe)  shows inland areas with more bare earth and ice (with melt ponds?) while nearer the sea land and ice are more white, i.e. snow covered. What is the reason for that? higher temperatures inland? Less snowcover to start with  due to lower precipitation near the cold icecap?

image from here:https://earthdata.nasa.gov/labs/worldview/?switch=arctic&products=baselayers,!MODIS_Aqua_CorrectedReflectance_TrueColor,MODIS_Terra_CorrectedReflectance_TrueColor~overlays,arctic_graticule_3413&time=2014-06-15&map=10373.115699,-2212873.838375,1296517.115699,-1320969.838375

[Andreas T]

the same area last year at 15 june for comparison

[RaenorShine]

NSIDC have updated, and it is finally well above average Friv Melt area extent is at around 30%, just inside of the 2sd area, but not by much.

My theory would be then that the boundary is, as you suggest, created because the DMI model ignore substantial surface melt until it assumes that the surface layers are so soaked that ponds and rivers starts forming and thus transporting some of the meltwater away through crevasses and moulins instead of it just refreezing together with those top layers of snow.

That would be my guess also, If I read their notes correctly, melt ponds would not be counted as negative SMB until they start moving either in moulins or rivers.  Even then they will only count if the model thinks the water will actually get to the sea.

The model (given the limited real data points available) is never going to be a 100% accurate map or graph of actual loss. If over the year the model gives an acceptable measure of net SMB (which can be checked against GRACE and the Glacier Outflows), say +/-20% I'd say that it's doing a good job.  As more research is done (and money and computer power becomes available) they'll continue to update it to better reflect reality.

[Frivolousz21]

We are really close to the darl ice layer fully revealing itself

[Timothy Astin]

Andreas T asks about the greater melting further west in the "East Greenland Riviera". The mountains are large, often over 1000m, and both hillsides and rock cliffs warm up in the sun.
Indeed, all over Greenland, the first places to see the snow melt in the Greenland spring are the valley sides.

Along the Atlantic ocean, the white sea ice stays cool, by comparison. Even if the sea ice melts, the generally southern drift of sea water and sea ice tend to keep the temperatures down.

I am fortunate to have personal experience of this from three summer visits to this art of Greenland to look at rocks thanks to CASP (www.casp.cam.ac.uk). A summer spent mainly in Canning Land, and mostly near sea level, was noticeably cooler than a summer based on the large mountains around Kejser Franz Joseph Fjord.

Also I noticed from MODIS images last year, that this part of Greeenland was one area that seemed to have an extensive and quite early thaw in 2013.

[Andreas T]

Thanks Timothy, always nice to hear from people with on the spot experience. I have found this

...A narrow band along the eastern coastline showed significantly greater than average melting, but here as well the surface melt conditions did not extend inland and uphill...

on NSIDCs 2013 review http://nsidc.org/greenland-today/ confirming what you say.
Could the height of the mountains also have an effect through warming fall winds (foehn)? There doesn't seem to be much difference in snowcover on south and northfacing slopes.

[RaenorShine]

The melt continues across Greenland, with DMI showing the SMB daily loss for 16/6 increasing to almost 4Gt (a new YTD record), despite increased precipitation across the southern peninsula.

The high melt area on the NSIDC site (30% and still on the way up) is already higher than anything seen until July last year, but its not on the scale of 2012 yet, which hovered around 40% for most of June and July with the huge peak of 90% of the ice sheet showing melt in July 2012.

Because of the recalibration it's difficult to compare the DMI values. Their 2013 values were more or less matching 2012 throughout June, with an 8Gt daily SMB loss around now.

[F.Tnioli]

I have three questions which are of general nature, but also affect the amount of water discharge from Greenland somewhat, so i hope it's OK if i ask them here.

At http://books.google.ru/books?id=4QJ2HkvFxIUC&pg=PA6&lpg=PA6&dq=greenland+%22bedrock+temperature%22&source=bl&ots=R-K-biGW_g&sig=ExBJ7rnHi0FVm31-2PkLlLJlAnU&hl=en&sa=X&ei=AkugU_rCHIn_ygPciILoBg&redir_esc=y#v=onepage&q=greenland%20%22bedrock%20temperature%22&f=false , i read that bedrock temperature in central Greenland is "somewhat higher" than -13C.

1. How much is "somewhat"?
2. Is there any near-real-time monitoring of bedrock temperatures in Greenland, with data being available to public?
3. Is there any maps for how deep (in meters) Greenland permafrost is?

With the melt going like it does right now and like it did 2 years ago, bedrock temperatures are to increase (with all the water going under the ice sheet, it's inevitable). The big reason behind my interest is not this melt season (it's secondary), - but the possible scenario of ice sheet itself (most of it) eventually sliding down into the ocean, once bedrock temperatures are high enough (as long as it's below ~0C widely enough, the bedrock itself will keep being freezed "into" the ice sheet, preventing it from moving anywhere; but with above 0C temperatures deep enough under the surface - geo gradient they call it, do they? - and with above 0C water coming from above from melts, sooner or later bedrock will happen to get above 0C for most of its surface, i guess, and then much/most of ice sheet may just slide down - amount of potential energy all that ice has due to being that high above sea level - is tremendous).

[Rubikscube]

The first record domino of the 2014 melting season is down.

Weather station KPC_U, which located in the north easternmost corner of Greenland not far from Zachariae Isstrøm, is reporting that 16. June 2014 saw a peak temperature of 5,73 C, which happens to be the highest temperature recorded since the station begun operating in 2008. http://promice.org/WeatherArchive.html?promiceStationStationid=114&stationid=114

One might argue this isn't the record that will be discussed the most by generations to come, but it is never the less a record .

[folke_kelm]

to answer your questions, F. Tnioli,

1. ice core temperatures suggest bedrock temperatures around -10 to -13 degrees at some points
That ist important, because you only have some few points where the temperature is known. It is not very likely that the bedrock is colder, because ice is a very good thermal insulator. it is much more likely that temperature are around minus 2 to minus 5 degrees at many points because oft the heat flow inside the bedrock. (geothermal gradient)

2. No, as long as i know there are no continuous measurements of bedrock temperatures.  This is not necessary either, because the temperature will not exceed a temperature around the melting point of water which is dependent of the overlying pressure and i guess this will be the bedrock temperature  at many points. Melting at these point is only possible due to geothermal heat flow. Meltwater from above will certainly be in thermal equilibrium in this deep due to heat exchange under its way.

3. No, you cant measure it due to far too much ice overlaying the bedrock, but i guess that at most places there will not be much permafrost under the ice shield due to geothermal heat flow.

4. (which you did not ask) The entire ice cap can not just slide downhill, because Greenlands bedrock topography is like a salad bowl, very deep down in the middle (600 m under Sea level) and Mountains around it. Extraction of the ice is only possible through outlet glaciers like Peterman, Zachariae (see this forum) or others or surface melting. Water percolating down will not have above 0 temperatures but exact 0 degrees. It will eventually refreeze inside the ice shield, and warming it from inside, but not melting it. The important fact about this heat accumulation inside the ice shield is, that the viscosity of the ice is very strong dependent of the temperature. The warmer it is, the faster the glaciers are moving.

[RaenorShine]

NSIDC Melt area has increased to around 35% well over the 2sd area. Melt is around all the coasts, and across the saddle between the north and south sheets.

[F.Tnioli]

to answer your questions, F. Tnioli,

1. ice core temperatures suggest bedrock temperatures around -10 to -13 degrees at some points
That ist important, because you only have some few points where the temperature is known. It is not very likely that the bedrock is colder, because ice is a very good thermal insulator. it is much more likely that temperature are around minus 2 to minus 5 degrees at many points because oft the heat flow inside the bedrock. (geothermal gradient)

2. No, as long as i know there are no continuous measurements of bedrock temperatures.  This is not necessary either, because the temperature will not exceed a temperature around the melting point of water which is dependent of the overlying pressure and i guess this will be the bedrock temperature  at many points. Melting at these point is only possible due to geothermal heat flow. Meltwater from above will certainly be in thermal equilibrium in this deep due to heat exchange under its way.

3. No, you cant measure it due to far too much ice overlaying the bedrock, but i guess that at most places there will not be much permafrost under the ice shield due to geothermal heat flow.

4. (which you did not ask) The entire ice cap can not just slide downhill, because Greenlands bedrock topography is like a salad bowl, very deep down in the middle (600 m under Sea level) and Mountains around it. Extraction of the ice is only possible through outlet glaciers like Peterman, Zachariae (see this forum) or others or surface melting. Water percolating down will not have above 0 temperatures but exact 0 degrees. It will eventually refreeze inside the ice shield, and warming it from inside, but not melting it. The important fact about this heat accumulation inside the ice shield is, that the viscosity of the ice is very strong dependent of the temperature. The warmer it is, the faster the glaciers are moving.

Thank you for answers.

An extra bit on 4. Yes, it's somewhat like a salad bowl, but with a few openings around its edges. Yet the general idea stays. The meltwater discharge into the ocean which we see during melt seasons, and massively increased during last few years - i guess it is but a little part of total amount of melt water in greenland; this is the water which happen not to get trapped within the "salad bowl". Most meltwater goes into it, though, and stays "inside" the bowl. Since water is heavier than ice, it goes down, right to the bedrock, and forms under-ice lakes and even seas. Whatever bedrock permafrost is there - freezes part of this water (but by doing so, warms up and eventually melts, with geothermal gradient helping it). The result is larger and larger under-ice-sheet lakes of liquid water (much like those under-ice-sheet lakes in Antarctica). Liquid water is also practically incompressible liquid, and thus it serves as a good lubricant between the ice sheet and the bedrock. Sooner or later, there will be enough liquid water down there, and too little bedrock permafrost left to refreeze it, for the greenland ice sheet itself to become able to move. Even a small move - like, few dozens meters to any direction - of trillions and trillions of tons of ice would probably result in a big hit once "looking down peak(s)" of still-solid ice would meat with "looking up peak(s)" of rock. Which would probably result in fragmentation of the ice sheet, and then, most of ice sheet (in pieces) would probably end up dropping into the world ocean. Parts of it which were below sea level - down in those -600 meters to sea level parts of the "bowl", - will surely stay there, floating in a layer of meltwater; but most of what was above sea level - will find its way into the ocean, simply by gravity. The integrity of the ice sheet (as a single piece) only holds as long as the sheet itself is firmly anchored to the bedrock, and layers of meltwater between the ice sheet and the bedrock are exactly things which remove such anchoring. Granted, this is not likely to happen any soon - hopefully it's at least 5 decades ahead, - but when it does, sea level rise of a few meters would basically disable most trade routes and ports, and possible wave(s?) of unprecedented proportions would destroy lots and lots of ports, cities, infrastructure, and human lives.

I am quite convinced that Greenland ice sheet won't melt peacefully into the water, just shedding some little icebergs via its coastal ice shelves, and discharging meltwater all the way till there is no ice left in summer Greenland. May be a trigger in a form of moderate earthquake will be needed to fracture the ice sheet, may be quite much of ice sheet would stay "in the bowl" and only a minor part of the sheet would slide into the ocean - but even if 10% of the sheet would drop into the ocean, it'd already be devastating for modern globalized civilization, considering improtance of coastal regions, sea trade, and devastating effects of the wave(s) generated by such a drop. That's why i am trying to sort out how it'll go.

[RaenorShine]

DMI Polar Portal show the melt decreasing yesterday, with the precipitation lessening also.  Daily SMB loss is 3Gt, still outside the 2sd range.

This is backed up by the daily temperature highs, Wunderground show the maximum tempurature from their stations falling back to 17C (Narsarsuaq) compared to the day before's 20C (Kangerlussuaq)

[folke_kelm]

 F.Tnioli

Principally i agree with you that the ice sheet will not melt peacefully over several 10000 years. it will be much faster.
But you have to consider physics in your reasoning. some of your scenarios will not happen because of simple physics.
For example:
"Most meltwater goes into it, though, and stays "inside" the bowl."
This will not happen. Meltwater will of course reach the bedrock inside the salad bowl, but it will be under far too high pressure to stay there due to overlying ice. The meltwater which is not refreezing will be squeezed out to point with lower pressure, the edges of the sheet and the sea. The water will find fractures, if not finding any it will create them. There is no chance of accumulating until the ice sheet is thin enough to swim (about 600-700m only).
There is no doubt about, that the entire ice sheet is able to move, but again you have to consider, that ice acts as a liquid with very high viscosity. First when you get to the edges of the sheets, it will act as a solid and fall off the cliff. You see this behavior very clearly at Jakobshavn Ice stream. You are able to follow liquid movement far into the ice sheet, and then discharging the ice due to calving.
This behavior is only possible, when you have liquid water under the glacier, that is not possible now under the ice inside Greenland.

for the ice sheet to disappear there are three ways of significant importance.
surface melt and run off
outlet glacier discharge
melt from below due to circulating sea water
With warming sea water and retreating grounding line glacier discharge and bottom melt will gain in speed. behavior like you see in Jakobhavn Ice Stream will spread to other outlet glaciers, thinning the interior ice sheet. This is a slow process, far slower than you expect, but much faster than we are familiar with.

[F.Tnioli]

F.Tnioli

Principally i agree with you that the ice sheet will not melt peacefully over several 10000 years. it will be much faster.
But you have to consider physics in your reasoning. some of your scenarios will not happen because of simple physics.
For example:
"Most meltwater goes into it, though, and stays "inside" the bowl."
This will not happen. Meltwater will of course reach the bedrock inside the salad bowl, but it will be under far too high pressure to stay there due to overlying ice. The meltwater which is not refreezing will be squeezed out to point with lower pressure, the edges of the sheet and the sea. The water will find fractures, if not finding any it will create them. There is no chance of accumulating until the ice sheet is thin enough to swim (about 600-700m only).
There is no doubt about, that the entire ice sheet is able to move, but again you have to consider, that ice acts as a liquid with very high viscosity. First when you get to the edges of the sheets, it will act as a solid and fall off the cliff. You see this behavior very clearly at Jakobshavn Ice stream. You are able to follow liquid movement far into the ice sheet, and then discharging the ice due to calving.
This behavior is only possible, when you have liquid water under the glacier, that is not possible now under the ice inside Greenland.

for the ice sheet to disappear there are three ways of significant importance.
surface melt and run off
outlet glacier discharge
melt from below due to circulating sea water
With warming sea water and retreating grounding line glacier discharge and bottom melt will gain in speed. behavior like you see in Jakobhavn Ice Stream will spread to other outlet glaciers, thinning the interior ice sheet. This is a slow process, far slower than you expect, but much faster than we are familiar with.

1. Pressure. I don't think so. Ice density is lower than liquid water density. Means, any water which will be replacing solid ice near the bedrock - will be of substantially less volume than the ice of the same mass. Then, to have any pressure applied to those layers of water, the icesheet must actually _drop down_ relative to the bedrock. Which won't happen as long as the ice sheet is anchored to the bedrock and literally "sits" on that bedrock. Overwise, vast under-ice-sheet lakes (more like seas in some cases) wouldn't exist. But they do, hundreds of them in Antarctica, and i wouldn't be surprised if quite many already exist near Greenland's bedrock for a long time already, too.

2. Yes, i didn't mean "ice floating" thing. Of course, kilometers-high mass of ice can't float in barely few-hundreds-meters-deep water. All that ice would still have most of its weight placed right onto the bedrock itself. But with most melting going on at the bottom, - not somewhere in the middle of the ice sheet, - the "bottom parts which serve as weight-bearing columns for the whole ice sheet" will be decreasing in number, in size, and in strength (warmer ice is softer, i hear?). And since the rockbed is not purely horizontal - whenever ice sheet just "drops down a tiny bit", horizontal forces (upon collision/deceleration with/by the rockbed) will also be generated. Thus "sliding into the ocean" bit.

3. You said "This behavior is only possible, when you have liquid water under the glacier, that is not possible now under the ice inside Greenland.". Sicne i suspect it's the opposite - not just possible, but already existant liquid water under the "inside" parts of Greenland ice sheet, - i took some time to search for any possible info on this. I found such info: http://www.cam.ac.uk/research/news/lakes-discovered-beneath-greenland-ice-sheet-0 . Note this is exactly the kind of lakes i am talking about: fed by surface melt waters which go all the way down to the bedrock. It seems to me my physics is not as bad as you suspected it is, eh? =)

[RaenorShine]

NSIDC melt area increased again on 17/6 tp just under 40% of the ice sheet.