Volume gap to previous lowest to shrink late in melt season?

[crandles]

Over on Stoat, I wrote and William Connolley replies shown in [ xxx -W] form:
http://scienceblogs.com/stoat/2017/05/03/economist-farewell-to-the-arctic/

You seemed to believe that tends in area/extent were better than using volume.

[Perhaps that they are more familiar. And certainly they are more strongly observationally based; I know little of PIOMAS or how much I should trust it. Also, its alone: unlike the area / extent, which many centres provide -W]

Is there a test of that with this years low volume at maximum? If it is the volume trend that matters we should expect the 1.8 k km^3 volume below previous lowest to grow over the melt season due to albedo effects. (I.e. thinner ice melts out quicker allowing albedo to fall faster, more heat absorbed by ocean that can melt more ice.) If it is area or extent that shows a more consistent trend as you have seemed to indicate you believe then as we get down to low volumes shouldn’t that volume get harder to melt and so the gap to previous lowest should shrink towards the end of the melt season?

[Yes, that’s kinda what I think. I also think that the volume “matters” less. For the climate, what really matters is the albedo, to first order; which is a function of area, to first order -W]

I am not sure why you believe this and you haven’t seemed able to explain. Perhaps bathymetry effects?

Anyway do you believe the gap to previous lowest volume will shrink towards the end of the melt season. Could there be a bet about this to test the idea?

[Unfortunately I don’t have a strong opinion on that. I think it likely that the gap will shrink, yes. I’d even put money on that, just to formalise things -W]

I suspect I won’t disagree with you enough to bet but maybe others will be interested to see if you propose anything.



This seemed like an important matter to test this season, so hope it is ok to add this poll and we are not having too many.

Feel free to vote and/or add comments. Further reasoning on option 2 particularly welcome eg is it bathymetry effects or ....

[Jim Williams]

I voted "No gap will increase mainly for reasons other than albedo," but I did that because I took the question as "albedo this season."  I think that albedo is the major ultimate cause after CO2 in making the Arctic warm faster that everywhere else, but I don't think it a main proximate cause.  Thus, I don't think open water early in the year is all that important in determining the volume late in the season.

[jai mitchell]

The expectation of a decreasing gap to low volume toward the end of the melt season is contingent on the previous observation of large volumes of land-fast multi-year ice on the northern Greenland coast. 

Similar to the 'effective ice-free convention' in SIE of < 1M km^2

This multi-year ice no longer exists in the way it was previously understood.

The assumption that later volumes of ice will be harder to melt is no longer supported by conditions in situ nor the physics of the environment.  However, at this stage, there are many larger impacts to projections of minimum volume and extent, through the coming melt period, than albedo.

for example, the early breakup of the Nares bridge. . .

[Richard Rathbone]

Weather May-Sept in 2012 was very favorable to melt, therefore the gap will probably shrink. I might be able to outrun someone average, but that doesn't mean I can stay ahead of Bolt.

[Tor Bejnar]

I understand (or imagine the possibility of) forces that will speed melting (e.g., thinner ice means more bottom melt as more light penetrates the ice and heats the top water) and others that will impede melting (snow cover); therefore I've guessed "gap will remain about the same".  Of course, weather is number 1, and we don't know what that will do.

[Juan C. García]

No idea...

I am not sure if the gap will increase or decrease. I just believe that the gap is big enough, so if it shrinks, will not be that much to have it disappear.

Emotionally, on the other hand, I hope it will not increase, because that will take us to an ice-free Arctic.     But that is just a hope, not based on science.

In fact, I believe that it will depend on weather and it depends on some events that could or could not happen, like if we have a cyclone on early August. At this moment, nobody knows...

[magnamentis]

generally speaking, the less volume we have, the smaller the gap will be, of course with seasonal and year to year variations.

we certainly agree that with a volume of 500 we never can get a gap of 700, just to make sure that there is no misunderstanding as to what i mean.

[oren]

As 2012 was an exceptional melt year, I think there is a higher probability of the gap shrinking than the gap growing, although I still expect a new record low in volume.
If the gap does grow I doubt albedo will be the main reason, as extent and area are high and could remain relatively high for a while due to dispersion and export among other things. But when thin ice is left over large areas, strong storms and GACs could wipe it out.

[johnm33]

It could go either way, but I think it will increase. There's so little thick ice against the CAA, and I suspect it could rapidly disappear, when it does, so long as the ice is driven from the coast, wave action will then act all round the pack, if so area and extent will crash, volume too. We may be lucky but the more melt that takes place for other reasons the less ice can build up to block the straits and the less extreme the conditions needed to bring this about.

[Paddy]

I was tempted by the "don't know" option, but went for "no change". But there are factors that could drive it either way.

Factors that could drive a shrinking of the gap:
- Good snow cover this year -> Greater albedo on land -> cooler arctic temps
- Big drops came later in the season for the record low year of 2012 that may well narrow or close the gap

Factors that could drive a widening of the gap:
- Albedo feedback
- Thin ice fragility

Overall, I'm favouring either no change or a slight widening gap, but only time will tell.

[Hyperion]

All predictions are stabs in the dark. The extremely complex system of geobioatmophysic feedback dynamics we enjoy, is well beyond our current level of scientific understanding. Due to no precedents, for tens of millions of years at least, of such explosively growing greenhouse burden added to cataclysmic biosphere degradation as we have tested this system with, we can expect no less than emphatic and unpleasant rebound behaviour. Though in saying that, I should point out that the chronological period for the response is, from what we do know, anything from days to megayears.
At present however. Large effects are in the house. FE. the meeting of warm salty tropical waters and fresher arctic is no longer a extra arctic basin phenomena. This is now a prime driver of AMOC. as the interface is well established beyond the edge of the continental shelf now, and certainly of far larger and growing length, frictional forces of cooled deep return currents on the shelfs are gone. but  the cooling of high salinity atlantic incoming waters enhanced along the extended front. Creating a subduction sheetcurrent thousands of km long that is only enhanced by extreme weather and other perturbations. And thus we are stuck now with seemingly unavoidable NAD acceleration and consequences for what future is pertinent to the lifespan of the flesh bags we inhabit.

[Archimid]

I voted "No gap will increase mainly for reasons other than albedo", but albedo is important. Specially over land. I think the gap will increase because the remaining ice will be easy to melt, warm air will keep flowing in and I think there is huge potential for export through the garlic press.

[Jim Williams]

I voted "No gap will increase mainly for reasons other than albedo", but albedo is important. Specially over land. I think the gap will increase because the remaining ice will be easy to melt, warm air will keep flowing in and I think there is huge potential for export through the garlic press.

About the only thing I might add to that is that the broken ice is subject to wave action.  (And at some point will become subject to upper ocean mixing -- though I don't know when.)

[Paddy]

I voted "No gap will increase mainly for reasons other than albedo", but albedo is important. Specially over land.

On that note, Northern hemisphere snow cover is currently higher than the 10 year average:

[FishOutofWater]

I think that the CFS v2 model is right and that this summer we will see high pressure over the Pacific side of the Arctic and low pressure on the Atlantic side. This will cause compaction under the high and will enhance ice export towards the Fram. The net effect will be a precipitous decline in extent in June and July if the model verifies. Thus the "volume gap" will decline.

I think the model is getting the large scale subsidence and uplift patterns right because:

1: Persistence. The snow pack is thickest on the Kara sea side of Siberia because the polar vortex has been displaced in that direction for months. After the stratospheric polar vortex collapsed, the 500mb vortex continued to be displaced in that direction.

2: Ocean heat patterns, strong land heating in Alaska and north western Canada, and the enhanced Brewer Dobson circulation favor subsidence over the Beaufort sea where ice converges under the high.

In conclusion I think that we will see a record low September extent based on what's happening in the atmosphere. Of course, this forecast is not high confidence. Long range forecasting is hard if not impossible.

[crandles]

bump not long left to vote

[Tor Bejnar]

My "no change" vote is based on my interpretation of the poll: Whatever the volume gap is in August will be about the same as the gap in September.  (If there is a significant change in gap, I believe it will be due to weather.)

This is independent of the early melt season decline in volume gap.  (I think the recent gap change is due to weather.)

[magnamentis]

the lower we get, the less room for a "gap" there will be, if previous lowest of something is 2, the gap can't be larger than 2 if we get zero, if the previous lowest was 1, the gap can't be greater than 1 if we get zero, hence the gap to previous lowest has to shrink by definition.

[DuraSpec]

the lower we get, the less room for a "gap" there will be, if previous lowest of something is 2, the gap can't be larger than 2 if we get zero, if the previous lowest was 1, the gap can't be greater than 1 if we get zero, hence the gap to previous lowest has to shrink by definition.

Brilliant logic! Love it! Made me chuckle with a good feeling ~ thanks

[Tor Bejnar]

A "gap" could be the other way, of course - a "rebound" - and it could be HUGE.   

[magnamentis]

the lower we get, the less room for a "gap" there will be, if previous lowest of something is 2, the gap can't be larger than 2 if we get zero, if the previous lowest was 1, the gap can't be greater than 1 if we get zero, hence the gap to previous lowest has to shrink by definition.

Brilliant logic! Love it! Made me chuckle with a good feeling ~ thanks

perhaps it's time to mention that sooner or later we should all switch from comparing numbers to comparing percentages. i mean once we gonna have very small numbers of increased losses in sea ice because not that much is left, a loss of 400 something compared to 500 something the year before would make an loss of 20%, while 20 years before we could have had a loss of 3000 of
something which is much more than 400 but those 3000 were perhaps only 5% of the total.

i on purpose am using non-specified units (something) to avoid a back and forth discussion about the different kinds of things that are measured and their appropriate numbers. it does not matter, the concept applies to volume, area and extent and anything else that can be compared while the comparison is made on greatly deferring levels.
 

[magnamentis]

A "gap" could be the other way, of course - a "rebound" - and it could be HUGE.   

sure that LOL but as i understood, perhaps i'm wrong, this is meant about the amount of sea ice that we're lower than before. like a stock zero is the lower limit while there is no theoretical limit on the upper end for stock and a very high limit for sea-ice if we consider the several 1000m thick ice over many areas like scotland etc. during ice ages. again i agree, you're right, just not sure whether this is what we're looking at here, let me know if i got that wrong, always ready to stand corrected

cheers

[TerryM]

perhaps it's time to mention that sooner or later we should all switch from comparing numbers to comparing percentages. i mean once we gonna have very small numbers of increased losses in sea ice because not that much is left, a loss of 400 something compared to 500 something the year before would make an loss of 20%, while 20 years before we could have had a loss of 3000 of
something which is much more than 400 but those 3000 were perhaps only 5% of the total.

i on purpose am using non-specified units (something) to avoid a back and forth discussion about the different kinds of things that are measured and their appropriate numbers. it does not matter, the concept applies to volume, area and extent and anything else that can be compared while the comparison is made on greatly deferring levels.

I like the concept. Perhaps stats in the form of:


A 62% loss since 1979 and a 72% loss over the past year.


Much more informative, particularly to a casual observer than say 44k units since 1979 and 7k units since last year.


Terry

[be cause]

I like succinct titles .. just because ..

[Rob Dekker]

Chris, thank you for posting an interesting and important question.

I'm pretty sure that the answer, all other things assumed equal, is that the volume gap will stay the same. Essentially that the volume loss is essentially unchanged regardless of the actual extent of the ice cover.

Reasoning goes as follows, going to first (physical) principles :

Imagine a slab of ice in an ocean centered around the NP.
Now, know that the amount of energy available to melt a square meter of that ice is limited.
Essentially there is just X number of Joule's available to melt ice.
We know that to be the case, since a limited amount of energy enters that location, and a limited amount leaves, and neither of these two values depends on the thickness of the ice.
So that square meter of ice melts a given amount of ice, independent of its thickness.
It is also not dependent on the temperature, since melting ice is pinned down at about 0 C.

Also, the amount of heat available for melt is in first principle not dependent on the location (latitude) of the ice, since it takes a lot more energy to melt (1.5 meter FYI) ice than it takes to warm up the entire column of air above it.

For ice at the ocean-ice boundary things are not too different. There is a limited amount of energy available that the ice boundary extracts from the ocean water, and this is (again, in first principles) independent on the latitude of that ocean-ice boundary.

So the amount (volume) of ice melting out is not dependent on the latitude of the ice boundary nor the latitude of any specific piece of ice.

Volume melt (in first principle) thus depends ONLY on the amount of heat that gets inserted into the Arctic. And that depends on 1) the temperature of the rest of the planet (global warming) and 2) the heat absorbed by land-based snow cover (albedo feedback).

Specifically, albedo feedback of ice loss itself does NOT affect volume melt over a season. Only albedo feedback of (land based) snow cover does.

I hope that argument makes sense, but if not, just check out the PIOMAS volume graph :

and see that these are sinus waves with almost the same amplitude (no albedo feedback within a season).

This also means that there appears to be no physical reason why volume loss would slow down as it gets closer to zero, and thus Prof. Wadham's remark is still relevant :

In the end, it will just melt away quite suddenly

[Crocodile23]

Specifically, albedo feedback of ice loss itself does NOT affect volume melt over a season. Only albedo feedback of (land based) snow cover does.

All the other you say make sense and are logical, but this doesn't.
 Since ice loss in area, means there would be extra ice-free areas on the sea(because of this ice loss), where the temperature will start climbing and we know that this can happen MUCH MUCH more efficiently than sun melting the ice. So this rise of temperature in sea water, would have an effect in further ice melting.
No?

[crandles]

Specifically, albedo feedback of ice loss itself does NOT affect volume melt over a season. Only albedo feedback of (land based) snow cover does.

All the other you say make sense and are logical, but this doesn't.
 Since ice loss in area, means there would be extra ice-free areas on the sea(because of this ice loss), where the temperature will start climbing and we know that this can happen MUCH MUCH more efficiently than sun melting the ice. So this rise of temperature in sea water, would have an effect in further ice melting.
No?

I agree with Crocodile23. I thought it was well accepted that albedo feedback was the dominant effect in melting season.

Consequently I expected the dominant view would be that the gap would grow late in the melt season. Perhaps this is being reduced by views of expecting such extreme low volume in July that the gap can only shrink.

Consider circle of ice 1m across and 50cm think versus 100m across and 50cm thick. 1m across has plenty of water around it such that edge effects melt it all but this may not be case for 100m across. That is before considering albedo effects such as albedo of 20cm thick ice is lower than of 40cm thick ice.

Less volume means more volume can melt is clear from the data. What is less clear is what happens when the ice is only over deep water and other areas where it is hard to melt it all out eg where movement moves ice into area so open water fraction does not tend to rise or at least not as rapidly as in areas where ice moves out of the area.

[Richard Rathbone]

Albedo feedback is the reason it generally melts a lot less in the 4-8weeks before the solstice than it does in the 4-8 weeks after the solstice. Its also the reason the shapes of those PIOMAS curves Rob quoted have changed over the decades and now melt rather more in the month before the solstice than they used to while being pretty much the same shape for the rest of the year. (which is why the anomaly shows a regular June cliff in the past decade, the trend has been running for 4 decades but its only the last decade that the accumulated effect has got large enough to swamp the yearly noise)

So far the effects of an earlier start and the lower latitude ice melting out earlier have cancelled for the later months and the July August September shape of the volume curve hasn't changed, just shifted down. This isn't something that can go on indefinitely, but there's no definite break point in the trend yet so I'm with Rob, the normal expectation is that volume gaps stay constant late season.

The question then becomes, did 2012 have sufficiently unusual late season behaviour to expect an average 2017 to be different and I don't think it did. The big effect in 2012 was June, and I do expect the gap to close further this month, but late season it might go either way or stay the same, its too early to say what August weather will be like.

[Rob Dekker]

Specifically, albedo feedback of ice loss itself does NOT affect volume melt over a season. Only albedo feedback of (land based) snow cover does.

All the other you say make sense and are logical, but this doesn't.
 Since ice loss in area, means there would be extra ice-free areas on the sea(because of this ice loss), where the temperature will start climbing and we know that this can happen MUCH MUCH more efficiently than sun melting the ice. So this rise of temperature in sea water, would have an effect in further ice melting.
No?

You are right, and I worded that sentence incorrectly.
My point was that since volume melt is bound by energy deficit of the Arctic, and specifically independent of the latitude of the ice-water boundary, that the "gap" between different years does not change over the season.

[Dave C]

A lot of posts in this thread are fairly equivocal, but I'm willing to go out on a limb- The volume gap will almost certainly shrink.

The albedo effect is extremely overrated on this forum. Yes, in the long run albedo effects add up. But in the short run albedo effects are dominated by other factors. By far the two biggest volume melt years were 2010 and 2012. Ironically enough, those were the two highest recent years in April extent- precisely the opposite of what you would expect if the albedo effect were significant.
http://nsidc.org/arcticseaicenews/2015/05/new-tools-for-sea-ice-thickness/

Also, the gap has already started to close significantly-
2010 and 2012 have narrowed the gap by about 1000 km3 from April 30 to May 31.
Currently the gap from 2012 is 1481 km3.

It is virtually certain that this years minimum volume will not beat the record by 1480 km3. Due to 2012's huge melt season(for non-albedo reasons) there is a better than even chance that the volume record will not be broken at all.

[oren]

I refer to my recent post in the PIOMAS thread regarding the gap with 2012 reasonably expected to shrink quite significantly by mid-June. But it could go either way later.
https://forum.arctic-sea-ice.net/index.php/topic,119.msg116420.html#msg116420

[DavidR]

Only two of the past ten years have seen a volume loss sufficient to sustain the gap between the April maximum and the September minimum. Both of these, 2010 and 2012, had warmer starts to the melting season than this year. So I  don't expect the gap to be sustained.  On the other hand only  one year had a loss of volume that  would not break the current volume minimum record easily  so I  expect the record minimum to  be broken by a sizable margin.

[iceman]

A lot of posts in this thread are fairly equivocal, but I'm willing to go out on a limb- The volume gap will almost certainly shrink.
      ....

Normally I would side with the reasoning of anomaly reversion: a shrinking volume gap.  This time I will instead join the equivocating crowd.  The gap will widen during July, then narrow during August.
     First because of the distribution of volume.  At least according to PIOMAS, a disproportionate amount of volume is either already in the Atlantic melt zone or heading towards it.  And the wide swath of thinner ice toward the Siberian side of CAB looks vulnerable to any sustained dipole setup: either by melt ponding under sunny skies (admittedly doesn't look like an unusual amount so far) or heat advection from the open waters in Chukchi and elsewhere, which would be an indirect albedo effect.
    Second - the narrowing stage - because the GAC-2012 was apparently an outsize contributor to that year's record minimum volume.  Even though big cyclones fed by anomalous heat in the northern Pacific seem more likely than usual this year, there is no particular reason to expect an outlier like 2012's - which also took a track that did severe damage to the remaining ice.
    Which bucket to choose, then?  I'll go with "gap will shrink..." which still admits to the possibility of a new record low this year.  Another form of equivocation.

[Jim Hunt]

Due to 2012's huge melt season(for non-albedo reasons)

You appear to be neglecting the reduced albedo of melt ponds?