When will the Arctic Go Ice Free?

[dnem]

Alternate explanation. This is a very short time series that includes essentially meaningless year to year variability. The 2013-14 rebound could just be noise and reset the level high and the decline kicked right back in again after.  Obviously this is a cherry-picked explanation of the data, but I think the inter-annual weather variability makes it hard to discern any true trend.

[Glen Koehler]

Phoenix - I pulled the 2029 estimate for "BOE by September volume at ~0.8K km3" and 2024-2036 95% confidence interval from the Wipneus linear September volume chart at
https://sites.google.com/site/arctischepinguin/home/piomas

[oren]

Here's the CAB volume chart updated to include mid-Sep, which is quite similar to end-Aug. And here's another one for good measure, the average CAB thickness, where 2019 took the lead during late summer.
Note 1: Big thanks to Wipneus, without whose efforts all this regional data would not exist.
Note 2: Wipneus divides the volume data according to the Cryosphere Today regions, same as he divides the AMSR2 extent and area data. Thus to calculate average thickness one must divide by AMSR2 area (available since Aug 2012), not by NSIDC area.

Edit: I will add that IMHO, had 2019 followed on the heels of 2016, i.e. with the starting thickness of 2017 in May, things could have been much different and a new record could have been achieved without much fuss. The same could apply had the late season of 2019 not been so cold. IOW, it is not so difficult these days to achieve a new record, one extra factor could be enough.

[Glen Koehler]

Wipneus rocks!

[The Walrus]

Alternate explanation. This is a very short time series that includes essentially meaningless year to year variability. The 2013-14 rebound could just be noise and reset the level high and the decline kicked right back in again after.  Obviously this is a cherry-picked explanation of the data, but I think the inter-annual weather variability makes it hard to discern any true trend.

That is possible.  If we remove the high melt level from 2010-12, the linear decline from 2000-2009 would be 0.33k km3/yr, about 25% slower that the 2000-12 trend.  Extrapolating that trend to 2019, would yield less than 3.0k km3 last year, much lower that the plotted value.  I suspect there is more involved.  Not that there is a wall per se, but the ice in the CAB is likely to behave differently than the periphery ice.

[Phoenix]

First, thanks to Oren for the updated charts. Adding day 258 does show the unusual staying power of the 2012 melt season.

Personally, I don't see a convincing trend taking us to BOE by 2030. For those that do see that trend, I'm wondering if there is any near term interim milestone which you are looking at which would cause you to question your assumption? Would we need to see a new record minimum volume by a certain year  in order to maintain your conviction that we are still on a trend line for BOE by 2030?

There is of course a hypothetical non-trend route to BOE in which God / nature serves up a freakish weather distribution which isn't predictable.

fwiw - I ain't no physicist or serious quantitative scientist. I'm just someone who has looked at the ice at the minimum and wondered what separates the ice remaining at the end of recent seasons from the melted areas. The remaining ice is always far from any large continental land mass excluding Greenland and the Atlantic side remainder seems to be pretty correlated with the location where the shallow shelf drops off into the abyss. My feeling that those factors represent a greater degree of difficulty to overcome is mostly a hunch which is supported by the limited data available to me.

The one piece of corroborating evidence that I would suggest people look at in the current melting season is the correlation of 2m surface temperatures and proximity to continental land masses. My anecdotal observation is that there is a connection.

I acknowledge a bias of hoping that I'm right. I don't want the Arctic to melt out. I'm still invested in the hopium of political transformation and hope my kids will be spared the worst of AGW.

[gerontocrat]

And yet another way of looking at the data.

The graph attached shows that if you take the period 2001 to 2012 as separate from 2013 to 2019 you can argue that after 2012 volume loss has slowed.

If you take the 41 year linear trend as your benchmark, then 2019 is pretty near bang on the long-term linear trend of a volume loss of around 320 km3 per annum. Then yearly variations are just deviations from that trend.

So looking at, say just 5 years ahead to 2025, volume at the linear trend would be a tad above 2,000 km3. Five of the years from 2007 to date have had deviations more than 1,000 km3 below the trend value. Result - 5 possible years for a volume BOE.

On the other hand, the Arctic could stop getting warmer and volume will increase.
On another hand (GM rules, OK?), the volume loss has stalled and a BOE in volume terms is a distant maybe.

Me, I go for a probability of a BOE on one measure or another before October 1 2030.
_____________________________________________
ps: You can do the same sort of exercise with extent and area. Attached is the same graph format and calculation but for NSIDC Area. The story that emerges is somewhat different.

To maximise the speed of area loss before 2013 - choose 1996 as your start year.
The trend value of area in 2025 is 2.5 million km2. The maximum deviation from that would not give a BOE by 2025.

[The Walrus]

Very nice plot gerontocrat.  My only concern is that the variations are not random.  The first seven years are all lower (higher melt), two by a large amount.  That is followed by two decades of larger higher variations (16 were higher, three were lower, and two were neither).  That was followed by another seven years that were lower.  The last five years no so general trend yet.  My point is that the scatter is not random, but systematic.

[Phoenix]

Let's employ an analogy. Let's assume Novak Djokovic is AGW. We know in advance that Djokovic will eventually become the #1 player in the world just as we know that AGW is on a trajectory to melt all of the sea ice.

If we look at the linear trend of Djokovic's rise in the ranking as a teenager, we can project him becoming the #1 player. But we can't predict the 3-4 year stall at #3-5 in the rankings before he becomes #1 because the relatively smooth rise on the way to #3 doesn't tell us anything about how much he will have to improve before he can ultimately surpass Nadal and Federer.

What we're doing now in trying to predict the demise in sea level rise is equivalent to trying to predict how quickly Djokovic would rise from #5 to #1 back in 2007/8. The evaluation of that had everything to do with the gap between Djokovic and the players AHEAD of him and nothing to do with the players he had already surpassed.

Looking at a linear trend of the ice that AGW has already conquered is a mechanical process. Pop in the #'s and a formula and it gives you a best linear fit. It doesn't involve any scientific evaluation of what's ahead or reflect an expectation of the future.

We haven't reached a new record for 7 years. Why is that? Is it just the random walk of weather or is there something different about this layer that makes it tougher for AGW to penetrate, just as Nadal and Federer were much more difficult for Djokovic to overcome? That's an interesting discussion IMO.

I think everyone understands that this is not a linear process. The earth and the Arctic are not configured symmetrically and the AGW / earth interface ebbs and flows.

[oren]

looked at the ice at the minimum and wondered what separates the ice remaining at the end of recent seasons from the melted areas. The remaining ice is always far from any large continental land mass excluding Greenland and the Atlantic side remainder seems to be pretty correlated with the location where the shallow shelf drops off into the abyss. My feeling that those factors represent a greater degree of difficulty to overcome is mostly a hunch which is supported by the limited data available to me.

The one piece of corroborating evidence that I would suggest people look at in the current melting season is the correlation of 2m surface temperatures and proximity to continental land masses. My anecdotal observation is that there is a connection.

I know I should have listened in class a few years ago when A-Team explained about editing, overlaying and manipulating images, but here is my terribly crude attempt at putting the volume map for 31-8-2016 (greenish stuff) overlaid on the Arctic ocean bathymetry (bluish).
IMHO it can be seen that water depth, distance from land masses and latitude don't tell the whole story, and are overused as an explanation for ice resilience or lack thereof. The ESS arm has much better staying power over shallow water than the Beaufort over deep water at the same latitude. In addition, the surviving ice mass is adjacent to land along a very large region. The CAA islands can easily reach over 10^oC in mid-July / early August. The same applies to Ellesmere (Alert), Greenland (Kap Morris Jessup) and Svalbard (Longyearbyen). If proximity to land was a deciding factor, the ice should have been better off away from the CAA-Greenland line. As it is, the thickest ice is found against Ellesmere.
Obviously, latitude plays the strongest role, with the North Pole region insolation and bottom melt season too short to melt the ice thickness generated over the freezing season. However that is not the whole story.
I think a much better explanation is the combination of winds, currents and salinity. Being distant from incoming currents from the Atlantic and Pacific is an important factor. Being prevented from export south is another important factor, thus the bunching of the ice against the CAA-Greenland-FJL line.
Increased mobility over the years contributes to the risk to the ice. 2016 saw export of the thickest ice through the CAA channels, AKA the "garlic press", thankfully that has not repeated since then.
When will the Arctic go ice-free? I don't know. But I think the risk is high.

[Phoenix]

The one piece of corroborating evidence that I would suggest people look at in the current melting season is the correlation of 2m surface temperatures and proximity to continental land masses. My anecdotal observation is that there is a connection.

...
here is my terribly crude attempt at putting the volume map for 31-8-2016 (greenish stuff) overlaid on the Arctic ocean bathymetry (bluish).

IMHO it can be seen that water depth, distance from land masses and latitude don't tell the whole story, and are overused as an explanation for ice resilience or lack thereof. The ESS arm has much better staying power over shallow water than the Beaufort over deep water at the same latitude. In addition, the surviving ice mass is adjacent to land along a very large region. The CAA islands can easily reach over 10^oC in mid-July / early August. The same applies to Ellesmere (Alert), Greenland (Kap Morris Jessup) and Svalbard (Longyearbyen). If proximity to land was a deciding factor, the ice should have been better off away from the CAA-Greenland line. As it is, the thickest ice is found against the CAA and Ellesmere.

First...thank you for taking the time to create these images. Wish I had the skills and patience for that. The big image actually does a fantastic job of reinforcing the points I've been trying to make. The correlation of sea ice and ocean depth is very clear with the lone exception being the Beaufort Sea adjacent to N. America (a large heat advecting continental land mass).

There is a nuance which I obviously haven't been able to effectively communicate thus far regarding proximity to a large / continental / heat advecting / land mass. I'll try again.

Yes, it's true that an island in the CAA or a small region on the PERIMETER of Greenland or Ellesmere may heat up to 10C or more. These are areas of solid earth absorbing heat.  But the ability of these small pieces of earth to project heat out over the ice covered ocean is extremely limited. These are puny little pieces of earth and the the heat needs to transferred from the land to
 the ice. That heat will dissipate quickly when leaving these little hot places.

Siberia or Alaska are example of giant pieces of land that may have warm regions many tens of thousand of square miles in size which are adjacent to the Arctic. The ability of these very large warm places to project heat out over distance into the ocean is infinitely greater than a little edge of Greenland or an island in the CAA.

It seems I wasn't clear on the correlation I was trying to point out and I take responsibility for the failure to communicate that.

When a place like Siberia heats up, you might see 2m temperatures like 6-8 C OVER WATER / ICE in the adjacent Laptev Sea or ESS. That's an example of heat originating over land and projecting out over the ice. You won't see 2m temperatures like that over ice near Greenland and in the CAA. Those temperatures are maxing out at ~ 2C.

The temperature on the small piece of land itself on the edge of Greenland or on a CAA island is kinda irrelevant to sea ice melt. It's the temperature where the ice is that matters.

This is something we can easily look at on a daily basis during the melt season. Climate Reanalyzer provides 2m temperatures and everyone can easily sea that the 2m temperatures over the ice near Siberia get higher than the 2m temperatures over the ice near Greenland and the CAA.

Apologies for the failed communication. There is some nuance here that isn't easy for me to communicate.

[Phoenix]

Perhaps another analogy is in order....

Let's imagine you have a room which is 50 cubic meters in size and full of ice and the temperature in the room is exactly 0C, If you add a small rock which is 10 cm in diameter and is heated to 50C, it will not raised the temperature of the room very much.

If you add a much larger rock which is 2m in diameter to the room and also heated to 50C, it will give off a lot more heat than the small rock and raise the temperature of the surrounding room much more.

When they get heated in summer, Siberia and N. America are big hot rocks that are going to give off more heat and contribute a lot to melting the surrounding ice. The islands in the CAA are small rocks. Greenland is a pretty big rock, but only a small part of the rock (the snow free perimeter) heats up. The snow covered interior of Greenland doesn't get hot so Greenland behaves like a little rock in the example.

So at the minimum, there is no ice within 5 degrees latitude of the big hot rocks. The small rocks don't give off enough heat.

So where is the ice at the minimum?

1) Eliminate all of the ice south of 75N
2) Eliminate all of the remaining ice within 5 degrees latitude of a big rock (Siberia, NA)
3) Eliminate all of the ice on the Atlantic side that rests over the shallow shelf that sticks out to 82N.

What you have left is basically the minimum as depicted in the image Oren has provided above.

How do you get further reduce the ice from the current minimum at this point?

1) As Oren explained earlier, you could have some souped up wind driven export to blow it out. That would be a weather event that one probably couldn't easily predict. You could also have an extreme weather event like a GAC. Also, not easy to predict.

2) The Atlantification and Pacification bring warm water further into the CAB. The complexity of all the science behind that is above my pay grade. I just note the coincidence and correlation between the current penetration and the ocean depth and smell a connection.

3) Somehow, you get a  lot more air borne heat into the CAB. How that might happen is something of a mystery. Greenland is going to remain a snow covered rock for a long time. Siberia and NA aren't getting any closer to the CAB.

Looking at the DMI temperatures for the last 60+ years, the summer temperature range from June to July hasn't shown a lot of variance as it has the rest of the year. One doesn't see spikes to 5C at the north pole.

I think I'll rest my case for now. BOE could happen in any give year due to weather variability. But I would be much later than almost everyone else in predicting when its likely to occur.

[oren]

Phoenix - you are very clear, and discussion with you is a pleasure.

However, I am still not convinced. The lone exception of the Beaufort vs. ESS should give a hint that bathymetry isn't the deciding factor. Alaska and Siberia are both huge, Beaufort and ESS are both in southern latitudes within the Arctic Basin, Beaufort is deep and ESS is shallow, ergo you would expect Beaufort to melt after ESS. But it's always (or almost always?) the reverse. Why?
Based on my limited knowledge I would say the difference is currents, winds, salinity.

In addition, in 2016 next to the ESS arm you can see deep water with no ice, north of the 80^o line. Why is that region not ice-covered, but the region north of the CAA but south of 80^o all the way to 75^o is ice-covered? Proximity to land and a southerly latitude relative to the other region I mentioned should dictate differently. Ergo currents. winds, salinity.

Not to mention the ice within the CAA itself, surrounded by land and in shallower water. This can be marked as an exception due to low mobility during winter and spring, but it would be the third exception.

The bottom line is that I think, IMHO, that the bathymetry issue is almost coincidental, except along the Barents/CAB line where the incoming AMOC dives below the fresh waters of the Arctic thanks to the deep bathymetry. But even there check out 2019 where currents and winds caused a pushback of the Atlantic and saw persistent ice in the shallow region between FJL and Svalbard.

This is why I believe the deep bathymetry of the Central Basin will not save the ice. OTOH, it is protected by other factors:
A. First and foremost, its latitude which provides a long freezing season and a short melting season. Without mobility North Pole ice cannot melt out - there's simply not enough energy in the system.
B. Second, the surrounding islands decrease mobility of the ice pack and prevent its mass export.

These are strong defenses, but increased mobility in the CAB, lower winter volume/thickness, and earlier meltout of the surrounding seas all signal towards increased risk. Should area drop further,  the ice around the North Pole could move away more easily, towards less defended latitudes.

[johnm33]

Looking at the images posted by oren, the ice clears where warm southern waters enter. On the Atlantic side it moves through Barentz and off the shelf, but also around Svalbard over Yermak and along the shelfs slope creating turbulence as it passes. On the Pacific side it lacks power since it's only surface waters passing through Bering and this water flows according to currents and mlsp at any given time, so both east and west.
The ice is pushed around by shifting winds and currents, if it's forced north it also moves east and thickens as it's compressed, if it's forced south it moves west opens up and new thin ice is formed. The shortest distance to the pole, from land, is N. Greenland so this is where the eastward movement delivers and thickens the ice. If ice is 3m thick then the chances are some keels will project 30m into the ocean, the deeper the keel goes the easier it is to melt, the freshwater melt rises and yeilds a steady supply of easyfreeze water to repair openings. 30m keels also inhibit currents flowing directly beneath the ice so I'm guessing 3m thick ice has an inertial mass an order of magnitude greater than the ice alone. The disconnect 30m+ beneath the ice means currents flow without gaining purchase leaving the wind alone to act on the ice.
 IF this is right then for the ice to disappear the winds need to rotate the ice clockwise, the ice against Greenlands coast has to be both mobile and weak, and the ice moving towards Fram has to have shallow keels minimising the ?ground effect? on passing currents.
    A couple of big gifs courtesy uniquorn
https://gph.is/g/aQyRDGb
https://gph.is/g/aQOx0Jx

[Phoenix]

Phoenix - you are very clear, and discussion with you is a pleasure.

However, I am still not convinced. The lone exception of the Beaufort vs. ESS should give a hint that bathymetry isn't the deciding factor. Alaska and Siberia are both huge, Beaufort and ESS are both in southern latitudes within the Arctic Basin, Beaufort is deep and ESS is shallow, ergo you would expect Beaufort to melt after ESS. But it's always (or almost always?) the reverse. Why?
Based on my limited knowledge I would say the difference is currents, winds, salinity.

In addition, in 2016 next to the ESS arm you can see deep water with no ice, north of the 80^o line. Why is that region not ice-covered, but the region north of the CAA but south of 80^o all the way to 75^o is ice-covered? Proximity to land and a southerly latitude relative to the other region I mentioned should dictate differently. Ergo currents. winds, salinity.

Not to mention the ice within the CAA itself, surrounded by land and in shallower water. This can be marked as an exception due to low mobility during winter and spring, but it would be the third exception.

The bottom line is that I think, IMHO, that the bathymetry issue is almost coincidental, except along the Barents/CAB line where the incoming AMOC dives below the fresh waters of the Arctic thanks to the deep bathymetry. But even there check out 2019 where currents and winds caused a pushback of the Atlantic and saw persistent ice in the shallow region between FJL and Svalbard.

This is why I believe the deep bathymetry of the Central Basin will not save the ice. OTOH, it is protected by other factors:
A. First and foremost, its latitude which provides a long freezing season and a short melting season. Without mobility North Pole ice cannot melt out - there's simply not enough energy in the system.
B. Second, the surrounding islands decrease mobility of the ice pack and prevent its mass export.

These are strong defenses, but increased mobility in the CAB, lower winter volume/thickness, and earlier meltout of the surrounding seas all signal towards increased risk. Should area drop further,  the ice around the North Pole could move away more easily, towards less defended latitudes.

Hi Oren,

Lot to unpack there. I'll give it a shot.

1) Why does Beaufort begin melting before ESS?

I'm only focusing on the areas that have not yet melted. The Arctic is an infinitely complex beast. I'm going to pass on trying to explain the past.

2) Why is there a deep water patch w/no ice near the ESS arm in 2016?

I'm proposing that bathymetry is an obstacle that slows down the progression of ice loss, not an impenetrable fortress. I expect some exceptions and expect they will grow over time. Unchecked, AGW will melt everything in time. We're in agreement about AGW winning the war. I'm just proposing that the remaining sections will be SLOWER to disappear than the regions that disappeared in recent decades.

3) Why is there ice remaining in the CAA between 75N and 80N?

Because a) that ice is >= 5 degrees latitude away from the nearest continent. The closest continent is N.America at 70N. The other land in the vicinity counts as "small rocks" per my previous analogy.  And b) because bathymetry is mostly irrelevant in a place with extremely limited ocean circulation.

4) Nitpicky point - I don't think the Atlantic water that enters the Arctic is considered to be part of the AMOC. My understanding is that the northernmost point of the AMOC is in the N. Atlantic.

5) Your contention that bathymetry is pretty much relevant only on the Atlantic side.

I mostly agree. But I think it's also relevant on the Pacific side. Not relevant to CAA. This is an area where my communication was deficient.

6) You're not convinced.

LOL. Neither am I. But I do have a strong hunch which runs counter to the conventional wisdom here as most in the poll believe BOE is coming by 2030. This is a role reversal for me. As an acknowledged AGW doomer, I spend most of my time trying to convince people that things are worse and coming sooner than expected in order to create urgency for the Green New Deal. But near term BOE isn't one of the things I warn then about. I don't feel it.

[The Walrus]

Phoenix, the seasonal Arctic temperature is quite telling.  The largest temperature increases have occurred during the first three and last two months of the year.  Temperature spikes of more than 5C are becoming more common.  However, the summer months are showing very little variation, and may possibly be slightly cooler.  Whether this is a leading or trailing factor, I cannot say.  Hence, I tend to agree with your conclusions.

[Phoenix]

Phoenix, the seasonal Arctic temperature is quite telling.  The largest temperature increases have occurred during the first three and last two months of the year.  Temperature spikes of more than 5C are becoming more common.  However, the summer months are showing very little variation, and may possibly be slightly cooler.  Whether this is a leading or trailing factor, I cannot say.  Hence, I tend to agree with your conclusions.

Very cool to have someone paying attention and agreeing. The lack of variation in midyear temperature is definitely something that caught my eye. If you go to the DMI 80N site they have every year since the late 1950's. The portion of the year where the average goes above freezing has shown almost no variation in 60+ years. It's like a damn machine that caps out at ~ 2C.

To be clear, the DMI 80 reflects air temperature. It leaves me with the strong conviction that the CAB isn't going to hit with an air borne heat wave anytime soon. The GAC showed the CAB is vulnerable to heat from below rising to the surface.

This leads me to try and explain something which I haven't articulated previously. Why is bathymetry important in some places

If you look at the layers of the Arctic ocean, you have basically three layers of water. You have the freshwater surface layer, the mid layer of warm Atlantic intrusion and the bottom layer of legacy cold salty Arctic. Over time, Atlantification brings more and more of the mid layer into the Arctic. In the shallower areas which are just a few tens of meters deep, the growing mid layer is being forced closer to the surface relatively quickly. In the deep basins of the CAB (the B in CAB is for "Basin" after all), there is 1-2 orders of magnitude more room for that warm layer to expand into before forcing its way to the surface.

Thanks for your prompt here which gives me an opportunity to reflect more and perhaps communicate more clearly.

[oren]

Your end quote is missing a "]" at its end, that's why it all appears as quoted.

"

"     this is the start
"

"    this is the end

[Glen Koehler]

The following statements are excerpts from:
Dirk Notz, 2015.  How well must climate models agree with observations?
Phil. Trans. R. Soc. A.37320140164.  http://doi.org/10.1098/rsta.2014.0164

     “…the tuning of individual models to match specific metrics can cause a false sense of adequacy of the model for a particular purpose. If, say, a model is tuned to match the observed sea-ice area over the past 10 years, this does not imply that the model's physics allows the model to give credible projections of the future evolution of the sea-ice cover. ”
 
     "…the tacit assumption that we can infer the quality of simulating the future evolution of a climate observable from the quality of simulating its past evolution for which observational data might be available. However, the link between the model performance for a past evolution and the performance for the future evolution of the system is often not clear. "

     “… the multi-year sea ice that used to be the prevalent ice type in the Arctic until some years ago has distinctly different properties compared with the first-year ice that covers much of the Arctic Ocean today. If the processes that a particular model represents well are parametrized based on the properties of the multi-year sea ice, such a model is likely to suffer in performance when simulating the future evolution of the ice pack. ”

     “To establish a clear link between the past and the future evolution of a specific observable, one ideally uses observational data from periods where the climate state was similar to the one that one aims at simulating. This is reflected by the growing interest in paleo data in particular from those periods where the climate state of the Earth was distinctly different from what it is today. Observational data from such periods allow us to evaluate the model performance over a large range of climate states...”

     “Unfortunately, for many observables we currently have no reliable data that would allow us to test climate-model performance for different climate states. "

     “It is often assumed that a period of 30 years is sufficient to neglect internal variability of the Earth's climate system, but, as also shown in the following, this assumption usually does not hold during periods of a rapidly changing climate.”

     “A 30-year-long averaging period can be insufficient to substantially reduce the impact of internal variability, in particular during periods of a significant change in external climate forcing.”

-----------------
    Notz 2015 is a corrective reminder that just because I have a hammer, not everything is a nail.  (Regression is one form of statistical analysis I understand better than others, so I like to use it.)  My predilection for extending a linear regression trend forward is not necessarily legit.  The experts don't rely on such a simplistic method (then again, they sort of do sometimes, see Part 2 below.)
   
     Notz provides caveats about the difficulties of making projections about the future state of a highly variable, possibly chaotic system, that is undergoing rapid change, therefore preventing comparison to and trend analysis extracted from a rather short chronological record. Looking back, the Earth System models discussed by Notz have underestimated the rate of Arctic sea ice decline vs. observations.  Looking ahead, there was a wide spread in Earth System model estimates for when ASI extent would go below < 1M km2, as shown in chart below:



Figure 1. (a) Each dot specifies for a particular simulation the first year during which Arctic September sea-ice area drops below 1 million km2 for CMIP5 simulations under the scenario RCP 8.5. The respective model is identified along the y-axis. (b) Each dot specifies the CO2 concentration at which the Arctic summer sea-ice area drops below 1 million km 2 for the RCP 8.5 scenario. No dot indicates that the specific model only gets ice-free after the year 2100 and a CO2 concentration of more than 900 ppm.

     It will be really interesting to see what the new generation of CMIP6 models say about the future trends for annual ASI extent and volume.

********************
Part 2.   Notz and Julienne Stroeve coauthored a 2018 paper that associated global average surface temperature with when the Arctic sea ice Extent would begin to fall below 1M km2.
https://link.springer.com/article/10.1007/s40641-018-0113-2
 
      That temperature works out to be about 1.7C above the 1850-1900 average, which on our current trajectory we could reach between 2034-2046, with a narrower "best guess" range of 2036-2043. (Temperature estimates are based on a mixture of RCP4.5, RCP4.5 adjusted for observations, ditto RCP8.5 and 8.5 adjusted, and two methods to extend recent observed temperatures and rate increases into the future).  Note - these are the dates I came up with, they are not from Notz and Stroeve. 

      That study also correlated ASI extent going below 1M km2 with cumulative CO2 emissions.  Future emissions are of course unknown, but using an International Energy Agency projection for future emissions trend, the "CO2 - ASI extent" correlation gives an estimate of 2033-2036 for going below 1M km2.  Again - my dates, not from Notz and Strove.

[Freegrass]

Why does Beaufort begin melting before ESS?

Let me give that a go...

I think the Beaufort begins melting out before the ESS because of the warm pacific water that enters through the Bering Strait taking a right turn when it enters the Chukchi sea. I added an image with the currents in the Arctic Ocean to make that visual. It clearly shows the pacific water going to the Beaufort sea.

The reason why it takes so long for the ESS to melt out is because it's not getting warm water from the pacific, or salty water from the Atlantic. The ESS is basically protected from outside influence. It just has to melt out all on its own, near the coldest part of Siberia.

[Glen Koehler]

.... on the other hand
   - Simple linear regression has performed better at estimating future ASI volume over the past 10 years than the Earth System models.

   - As far as I can tell, reinforcing "positive" feedback mechanisms (thinning ice with less structural integrity, mobility, and susceptibility to export; longer wind fetch for more wave action; albedo reduction during melt season and higher water vapor (a powerful GHG) from more open water; weakening Arctic isolation from warm air masses and possible increase of cyclone activity from warming waters etc.) either match or exceed the number and impact of suppressive "negative" feedbacks (greater winter heat loss from open water, rapid recovery of thin first-year ice extent).

   - Higher latitude of the remaining September ice does not appear to be a strong protective factor.  The remaining ice is not centered around 90N.  CAB volume has been decreasing almost as fast as the Arctic as a whole.

    What is the argument that simple linear regression of ASI volume is not a reliable predictor?  What mechanism is going to kick in now that September volume has decreased by 75%(!) in just the last 40 years, from 16 down to 4K km3, to interrupt continued progression along this trend until the cumulative Sept. volume loss reaches 95% and BOE at 0.8K km3?

    The X2 polynomial curved regression (labeled "exponential" on the Wipneus graph page  https://sites.google.com/site/arctischepinguin/home/piomas)  shows an accelerating rate of decline that is a statistically better fit to the data than the straight-line trend.  (That said, I don't trust it for future trajectory).

    Global GHG gas emissions not only continue, until 2020 they continued to increase at an increasing rate.  Even if there is an emissions drop because of economic calamity in 2020, that is not a reduction in atmospheric CO2, just a slowing of the rate at which we are adding more CO2 into the air.  Earth system response for global average surface temperature shows signs of acceleration as carbon sinks are weakened and new CO2 sources (e.g. permafrost) activate.  Yada yada.

    It seems ridiculous for me to opine about this when there are expert analyses available. A recent comprehensive review (IPCC, Special Report on Cryosphere, Sept. 2019) is at https://www.ipcc.ch/srocc/chapter/chapter-3-2/  (top section is good, but also scroll down to section 3.2.2.1).

    Here is the key sentence (edited to remove citations):
     "There is a large spread in the timing of when the Arctic may become ice free in the summer, and for how long during the season as a result of natural climate variability, scenario uncertainty, and model uncertainties related to sea ice dynamics and thermodynamics. Internal climate variability results in an uncertainty of approximately 20 years in the timing of seasonally ice-free conditions, but the clear link between summer sea ice extent and cumulative CO2 emissions provides a basis for when consistent ice-free conditions may be expected (high confidence). "

    God bless IPCC, but that report while detailing the driving factors and current status, doesn't stick its neck out to estimate a specific date for first and regular BOE.  The experts don't have a crystal ball either, and the Wipneus linear volume trend looks compelling. 

    If you know of reasons why we should not expect to continue following that path (with the inevitable ups and downs around it due to annual variability of course) please educate me.  I see September ASI as reaching its first BOE (defined as ca. 0.8K km3 volume associated with 1M km2 extent) between [edit 2024 and 2036] 2021-2030, with a best-guess of [2028- 2029] 2025. 

[oren]

Besides the arguments Phoenix raised above that might preclude early BOE (latitude, distance from land, not enough energy to melt the more defended ice), there is the greater argument that the volume trend is composed of the loss of MYI and its replacement with FYI, joined with the loss of FYI as area/extent drop, as well as the thinning of FYI due to shorter/weaker freezing season and longer/stronger melting seasons. However as MYI volume is now quite low, the MYI trend should end soon and only the two FYI trends will remain, leading to a slowing down of the overall volume trend. This is the Chris Reynolds "Slow Transition" argument, a very good argument discussed on the forum in 2013-2014, but which assumed/expected a nearly stable volume in the Inner Basin at the end of the freezing season.
Whenever a trend is a superimposition of several trends, one must be careful in extrapolation because one of the trends might hit zero and not be able to continue.

Personally I haven't been convinced, I think 2016 was a great example of:
* How the supposed defenses of the Arctic are compromised - terrible ice condition near the Pole, MYI from the "last refuge" heavily exported through the CAA, 4m "Big Block" in the Beaufort completely melted out, Atlantic invading the CAB.
* How the 2012 August GAC was not a fluke.
* How the 2012 CAB area loss can be repeated, and in fact the overall area record of 2012 can be almost broken in a non-fluke season.
* How the freezing season can be much shorter than it used to be, negating the "always the same FYI" argument of the Slow Transition, with open water in the Chukchi into January, and 2017 starting the melting season with record low volume in the Inner Basin.

And the first half of 2019 showed that melting seasons can still be "hot" despite the claim of open water leading to cool cloudy conditions. And that volume banked in the previous year can disappear very quickly with the right weather, thus removing the requirement for consecutive strong years as in 2010-2011-2012. And that it's only the luck of weather that prevents new ASI records.

I still expect a BOE by 2030. But I just want to stress that the counter-arguments certainly have merit and there is significant uncertainty surrounding the issue.

[Phoenix]

For those who think we are trending toward by BOE by 2030, are there any interim milestones that you are looking to which are important to maintaining that outlook?

Let's say we don't have a new record by September 2021. We'd be at the halfway point between the old record and 2030.

Of course, a BOE will always be possible. But is there a point at which you might say the data trend no longer supports it?

[JNap]

Multi-year reader of the ASIF, first time poster.  Please pardon any newbie mistakes or ignorance, I am happy to learn from more experienced members and get better educated.

My Background:  As a layman that first got interested in AGW in 1980’s when I noticed material changes where I had lived in the upper Midwest of the US, i.e. hotter summers, less cold winters, etc.  About the same time, I read about Dr. Hansen’s testimony and have been following the subject since then.  Then the arctic sea ice drops in 2007 and 2012 mentally accelerated the timeline that increasingly extreme impacts will occur.  And a BOE event is one of the seminal impacts of this.

To this topic’s purpose, I think that we will have the first ice free arctic (less that 1,000,000 Km2) between 2028 and 2033.

Looking at gerontocrat’s post:   
https://forum.arctic-sea-ice.net/index.php/topic,2348.msg260070.html#msg260070

I take away the following two key points from the volume graph:

1)  The linear trendline shows that September ice will reach the 3,000,000 Km3 mark approximately 2025. 
2)  The yearly variance from the trend shows that there has been variance of 2.5 to 3.0 million Km3 in years 2010, 2011, and 2012.

The combination of these two would indicate that starting in approximately 2025, we are entering a period where a lower yearly starting volume will combine with a year with high negative variance, resulting in drop of volume below 1,000,000 Km3 – and probably extent below 1,000,000 Km2 threshold as well.

Obviously, these are based on extending current trends and recent variations.  The CAB ice could have different trend dynamics that the peripheral seas.  For example, FYI vs. MYI could have different trend dynamics.   Other factors include the degree / extent of “Atlanticification” of the eastern CAB and “Pacificication” of the western CAB.  The impacts of CAB bathymetry (e.g. Lomonosov Ridge), etc.

Bottom Line: 
It seems that even with some different dynamics for the remaining CAB ice, we will likely encounter one year between 2028 and 2033 where lowered starting ice combined with a medium negative variance will take it to the first instance on an Ice Free Arctic.

Would love to get more experienced perspectives on this.

[Glen Koehler]

    1) Good points Oren.  I had forgotten the MYI aspect of the "slow transition" discussion and focused on the "faster early winter heat loss from open water" and the "quick-growing thin recovery ice" aspects.  FWIW (i.e. pure guesswork), I think the lack of MYI export to support the volume loss rate is likely to be compensated for by degradation of the structural integrity of the remaining weaker thinner ice being more vulnerable to shattering by wave action, thus more mobile, and therefore more susceptible to flushing out the FRAM.
   
    2) Just to prove that I am big-hearted, fair-minded guy, I will throw a bone to after-2030 crowd and say that I do not agree with "2012 August GAC was not a fluke."  An analysis of the 2012 GAC found that it truly was an extreme outlier event, "The most extreme August Arctic cyclone (out of a population of 1618)."  The authors did not say so, but I presume that also means it is not likely to be repeated anytime soon. 

      They also noted that in addition to being a meteorological oddity, the GAC 2012 was perfectly timed to wreck ice already weakened by a strong melt season.  (The great Arctic cyclone of August 2012.  Ian Simmonds and Irina Rudeva,  Geophys. Res. Ltr. 39, L23709, 2012.)
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2012GL054259.

    That said, anything that happened once can surely happen again.  Even more so in a rapidly changing Arctic.


    3)  RE Phoenix interest in markers for when to throw in the towel for a 2030 BOE prediction.  The stats around the linear trend allow for estimating a cumulative probability for at least one year having gone below the 2012 volume record (3.79K km3).  Those values are:
2020    47%
2021    75%
2022    90%
2023    97%
   
    Translated to English, that means a new volume record is entirely possible but not expected this year.  But it is expected by Sept. 2021, strongly expected by 2022, and almost certainly by 2023.  If the record has not been broken by 2023 then one who worships regression religion would have to conclude that the reason is very likely due to failure of the linear trend (as stated in 2020), not just annual variability (for which the Arctic sea ice is infamous.) 

   Same reasoning gives a 50% cumulative chance of a first BOE (@ 0.8K km3 volume, 0.8m thickness, and 1M km2 extent) by 2025, and a 95% chance by 2030.  By 2030 each individual year has less than a 50% chance (and declining each year) of avoiding a September BOE.  By 2040 just about every year would have a September BOE.  This is one bet I'd like to lose.

[Phoenix]

    3)  The stats around the linear trend allow for estimating a cumulative probability for at least one year having gone below the 2012 volume record (3.79K km3).  Those values are:
2020    47%
2021    75%
2022    90%
2023    97%
   
If the record has not been broken by 2023 then one who worships regression religion would have to conclude that the reason is very likely due to failure of the linear trend ...

   Same reasoning gives a 95% cumulative chance of first BOE ... by 2030.  This is one bet I'd like to lose.

Hi Glen,

Can you check the 2012 volume record of 3.79k km3 above? In post #1545, you indicate 3.399k km3 for 2012 and 3.794k km3 for 2019.

If the linear trend leaves a 95% chance of BOE by 2030, that translates to 19:1 odds if the linear trend is true. I'll take bets on no BOE by 2030 at 5:1 odds. That's a crazy good bargain for people who truly believe in the linear trend.

[oren]

Multi-year reader of the ASIF, first time poster.  Please pardon any newbie mistakes or ignorance, I am happy to learn from more experienced members and get better educated.

1)  The linear trendline shows that September ice will reach the 3,000,000 Km3 mark approximately 2025. 
2)  The yearly variance from the trend shows that there has been variance of 2.5 to 3.0 million Km3 in years 2010, 2011, and 2012.

The combination of these two would indicate that starting in approximately 2025, we are entering a period where a lower yearly starting volume will combine with a year with high negative variance, resulting in drop of volume below 1,000,000 Km3 – and probably extent below 1,000,000 Km2 threshold as well.

First, welcome JNap. The first post is the hardest, I hope you continue to contribute with such well thought out posts.
A small correction, ASI volume is in 1000 km3, not 1,000,000 km3. A common confusion of which I've had my share in the past.
Very good point about the variability, people focus on the trends but forget the fluctuations. 2012 did not break the record by simply following the trend.

[oren]

I do not agree with "2012 August GAC was not a fluke."  An analysis of the 2012 GAC found that it truly was an extreme outlier event, "The most extreme August Arctic cyclone (out of a population of 1618)."  The authors did not say so, but I presume that also means it is not likely to be repeated anytime soon. 

      They also noted that in addition to being a meteorological oddity, the GAC 2012 was perfectly timed to wreck ice already weakened by a strong melt season.  (The great Arctic cyclone of August 2012.  Ian Simmonds and Irina Rudeva,  Geophys. Res. Ltr. 39, L23709, 2012.)
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2012GL054259.

    That said, anything that happened once can surely happen again.  Even more so in a rapidly changing Arctic.

I'm not saying the 2012 GAC was not extraordinary, but reminding that another August GAC hit 4 years later, though its effect was smaller for several reasons.
http://nsidc.org/arcticseaicenews/2016/09/arctic-sea-ice-nears-its-minimum-extent-for-the-year/
I also recommend to those who weren't around back then to read Neven's updates from Aug and Sep 2016.
https://neven1.typepad.com/blog/2016/08/index.html

    3)  RE Phoenix interest in markers for when to throw in the towel for a 2030 BOE prediction.  The stats around the linear trend allow for estimating a cumulative probability for at least one year having gone below the 2012 volume record (3.79K km3).  Those values are:
2020    47%
2021    75%
2022    90%
2023    97%

I think these numbers are too high. Chance of breaking 2012 record is IMHO 25% for a given year now.

To Phoenix:
I keep looking for new volume records, not just in September but in April-August. And not necessarily for the whole Arctic but for the Inner Basin and for the CAB itself. As long as these records are being broken it means the volume trend has not stabilized despite September appearances. If no new records are made for several years it will mean the trend has finally stabilized and the BOE strongly delayed.

[Archimid]

I'll take bets on no BOE by 2030 at 5:1 odds. That's a crazy good bargain for people who truly believe in the linear trend.

If the BOE happens by 2030 you won't be able to pay that bet. It is likely that all your plans will change. Like C19, but much much worse.

So that is not a good bet even if the odds look good to you.

[johnm33]

Freegrass that second image is particularly interesting as it shows different origins of Pacific waters which turn west

and east, from?
 The first animation from 20190514-0612, compared to the current one from hycom looking specifically at the movement along CAA It seems a different dynamic is operating whether temporary or persistant remains to be seen, but without the thick ice being recycled through Beaufort the whole arctic will become more mobile, more susceptible to wave action both peripheral and internal, and without the usual inertial mass will flow more freely towards Fram and more readily through the CAA.

[JNap]

Multi-year reader of the ASIF, first time poster.  Please pardon any newbie mistakes or ignorance, I am happy to learn from more experienced members and get better educated

First, welcome JNap. The first post is the hardest, I hope you continue to contribute with such well thought out posts.
A small correction, ASI volume is in 1000 km3, not 1,000,000 km3. A common confusion of which I've had my share in the past.
Very good point about the variability, people focus on the trends but forget the fluctuations. 2012 did not break the record by simply following the trend.

Thanks for the welcome oren.   And also for the magnitude correction in the volume.

For future projections, I think it is very helpful to look at the great set of graphs on Wipneus' page as Glen Koehler referneced above:  https://sites.google.com/site/arctischepinguin/home/piomas

The one in particular that seems most likely for the ice decline to follow is the Gompertz curve which shows the best fit to the data, i.e. the smallest residuals.  Also, that as the numbers get smaller and approach the zero bound, the rate of decline slows somewhat. 

This makes intuitive sense to me given that:
1) As we get to smaller volumes, the remaining ice is made up of a greater degree of MYI vs FYI.
2) The thicker MYI has a less surface area per volume and thus melts slower because:
(i) Less surface area for bottom melt from ocean heat content
(ii)Less surface area for surface melts from melt ponding, solar insolation, (and to a lesser extent, warm surface air temperatures given the low heat content of air vs. the oceans)

[The Walrus]

    3)  The stats around the linear trend allow for estimating a cumulative probability for at least one year having gone below the 2012 volume record (3.79K km3).  Those values are:
2020    47%
2021    75%
2022    90%
2023    97%
   
If the record has not been broken by 2023 then one who worships regression religion would have to conclude that the reason is very likely due to failure of the linear trend ...

   Same reasoning gives a 95% cumulative chance of first BOE ... by 2030.  This is one bet I'd like to lose.

Hi Glen,

Can you check the 2012 volume record of 3.79k km3 above? In post #1545, you indicate 3.399k km3 for 2012 and 3.794k km3 for 2019.

If the linear trend leaves a 95% chance of BOE by 2030, that translates to 19:1 odds if the linear trend is true. I'll take bets on no BOE by 2030 at 5:1 odds. That's a crazy good bargain for people who truly believe in the linear trend.

I guess that depends in which trend the people are believing.  The very short-term, since 2014, would certainly gets us to a BOE.  The short-term, since 2010, has no chance.  The intermediate short-term, 2001-2012, is similar to the very short term.  The long term, entire satellite data, yields about a 50:50 chance.  The very long-term, pre-satellite included, shows no chance.  I would not bet against Phoenix.

[Phoenix]

If no new records are made for several years it will mean the trend has finally stabilized and the BOE strongly delayed.

I agree. If no new records for next 2 years, then we are not trending to BOE 2030.

Personally, if something catastrophic is coming out of the Arctic in the near future, I think it will be release of the fresh water accumulation from the Beaufort Gyre which slows the AMOC. That can happen anytime and a doom scenario I believe in.

[The Walrus]

If no new records are made for several years it will mean the trend has finally stabilized and the BOE strongly delayed.

I agree. If no new records for next 2 years, then we are not trending to BOE 2030.

Personally, if something catastrophic is coming out of the Arctic in the near future, I think it will be release of the fresh water accumulation from the Beaufort Gyre which slows the AMOC. That can happen anytime and a doom scenario I believe in.

Yes.  That will cool Northern Europe and disrupt the North Atlantic fishing industry.

[Glen Koehler]

Can you check the 2012 volume record of 3.79k km3 above? In post #1545, you indicate 3.399k km3 for 2012 and 3.794k km3 for 2019.

Post #1545 was about ice volume in just the CAB - Central Arctic Basin, not the entire Arctic.
     The point being discussed there was that the decline of CAB volume is almost as steep as the Arctic as a whole.  That argues against the CAB being able to resist melting because of high latitude.   Though it is still possible that as peripheral parts of the CAB melt out, continued losses from the CAB will stall when the only remaining ice is at the highest latitudes.  But I don't see evidence for that in the observations so far.
     Post #1574 just above is about volume for all of the Arctic sea ice.
https://forum.arctic-sea-ice.net/index.php/topic,2348.msg260423.html#msg260423


RE GAC 2012 - Agreed, Arctic cyclones are a regular occurrence.  The analysis cited was about the combination of unprecedented intensity and scale of the GAC 2012.  The fact that it hit at the worst possible time for damaging the ice was icing on the cake.


RE Linear trend probabilities for being below the current low volume record for the entire ASI
          The multiplicative effect that generates a cumulative probability for at least one year going below a threshold value makes those probabilities much higher than for any single year.  It is the same math that makes compound interest on savings so powerful over time (or drives up the interest charges on a 30-year mortgage).

    Here are details for the probabilities posted in #1574
Threshold:  Current low record ASI (entire Arctic) September volume of 3.787K km3
    Values rounded to the nearest whole % for display.

Year    Chance of that single year       Chance of at least one year since 2019
          being below 3.787K km3         having gone below 3.787K km3
2020      47%                                    47%
2021      54%                                    76%
2022      61%                                    91%
2023      68%                                    97%
2024      74%                                    99%
2025      80%                                    99.8%
2026      84%                                  100%
2027      88%
2028      91%
2029      94%
2030      96%
2031      97%
2032      98%
2033      99%

*****************************
Threshold:  Using 0.8K km3 as arbitrary definition of BOE based on volume.
Derived from consensus Extent BOE threshold of 1M k2,
and E - V/T. 
    And the observation that linear trends for thickness and volume reach 0.8m and 0.8K km3, respectively, at roughly the same year.  Thus E = 1.0 when V/T = 1.0, which 0.8K km3 / 0.8m fulfills.
    Values rounded to the nearest whole % for display.

Year    Chance of that single year     Chance of at least one year since 2019
          going below 0.8K km3           having gone below 0.8 km3
2020      4%                                       4%
2021      6%                                     10%  because 1 - [(1-0.04) * (1-0.06)]
                                                                = 1 - (96% * 94%) = 1 - 90% = 10%
2022      8%                                     17%   same as above but keep adding multiplier terms
                                                                  e.g 1 - (96%*94%*92%) = 1 - 83% = 17%   
2023    12%                                     27%
2024    15%                                     38%
2025    20%                                     51%
2026    26%                                     63%
2027    32%                                     75%
2028    38%                                     84%
2029    45%                                     92%
2030    52%                                     96% 
Starting in 2030, each single year has >50% chance of BOE
2031    59%                                     98%
2032    66%                                     99%
2033    73%                                     99.8%
2034    78%                                   100%
2035    83%
2036    87%
2037    91%
2038    93%
2039    95%
2040    97%

[Glen Koehler]

FYI - The straight line linear trend used to calculate the probabilities in previous past was derived by regression of all the PIOMAS September average volumes vs. year for 1979-2019.  That trend was then extended into the future to get future estimated Sept. average volume.

     The linear model explains 88.6% (adjusted R2 88.3%) of the variation between individual years.  The standard deviation for the iinear model estimates minus observations was 1.38 K km3.  Probabilites were calclulated by comparing the number of standard deviations between the value estimated for each single year and the respective threshold, and evaluating that number of standard deviations against a normal distribution z score. 

      I did a visual check of residuals to evaluate the assumption of normal distribution.  They looked OK to me, and they were virtually, if not completely, identical to the residuals chart shown on the linear Wipneus Sept. volume graph at https://sites.google.com/site/arctischepinguin/home/piomas .  The pattern of residuals suggest that a curved trend would be a better fit.  I use the straight-line trend because it is more reliable and conservative for making future projections beyond the range of data.  A curved trend would create much earlier estimates for dates when Sept. volume goes below a threshold volume, as shown on the Wipneus exponential chart.  I did not do any formal testing of the residuals or the assumption of normality. 

    I think the trend estimate based solely on volume trend gives dates 5 years earlier than the previous attempt to estimate 1.0M km2 Extent (https://forum.arctic-sea-ice.net/index.php/topic,2348.msg252395.html#msg252395) because the volume trend (being a value based on two multiplicative declining trends) is steeper than thickness trend used for that estimate, so the combination with thickness trend diffused the steep volume trend in the attempt to estimate Extent. But this probably just confusing you and is making my head hurt.

     And I hereby disavow my attempts in posts 1477, 1478 and 1480 to accommodate those who think Extent should be treated as a driving factor.  It just ain't.  As the great Juan C Garcia has as his tagline "Volume is harder to measure than extent, but 3-dimensional space is real, 2D's hide ~50% thickness gone.  -> IPCC/NSIDC trends [based on extent] underestimate the real speed of ASI lost."
   
    But lest I lead anyone into sin, I must acknowledge that the experts come up with significantly later dates for BOE, e.g. https://interactive.carbonbrief.org/when-will-the-arctic-see-its-first-ice-free-summer/  --- If temperatures rise by 2C, then:  "...the Arctic has a one-in-five chance of seeing its first ice-free summer in 2035, according to the study. The chances of an ice-free summer in any given year rise to one-in-two by 2045."

    Sorry for the longwinded post(s), but there it is, my best number crunching shot at using the good ole, brute force but robust straight-line volume trend to estimate when the ASI will
a) break the 2012 record (proposed answer = soon, good chance in 2020, and almost certainly by 2023), and
b) will have had at least one BOE year with volume <= 0.8K km3 (proposed answer = sometime
 in 2025-2030, sooner than most folks seem to think). 

[binntho]

2) Distance from a large heat advecting land mass. The second criteria is the more dominant variable and has a tad more nuance.

I'm not sure if being close to a landmass is better than being close to open ocean. The oceans absorb much more radiation than land, and the heat transfer of ocean currents has vastly larger potentiality than wind-driven heat transfer by air.

As discussed upstream, the density gradient is a roadblock to ocean heat making it to the surface where the ice is located. The salt in the warm Atlantic ocean water makes it too dense. There's abundant heat in the Arctic Ocean lurking in the subsurface layer below the freshwater lens.

Which is basic knowledge in this forum. And has nothing to do with my (perhaps badly) made point. Which is this:  Open ocean absorbs heat from the sun, no matter what the density. Open land absorbs much less heat.

Your earlier post seemed to indicate that ice being close to a landmass would somehow receive a warming boost because of air advecting over the warmed-up land and out over the ice. But I would say that being far away from a landmass would give added boost to melt during summer, by air advecting over open ocean which, as I've said, absorbs much more heat than dry land and has therefore more heat to give to any passing air (besides being able to transfer the heat itself to some extent).

[Phoenix]

2) Distance from a large heat advecting land mass. The second criteria is the more dominant variable and has a tad more nuance.

I'm not sure if being close to a landmass is better than being close to open ocean. The oceans absorb much more radiation than land, and the heat transfer of ocean currents has vastly larger potentiality than wind-driven heat transfer by air.

As discussed upstream, the density gradient is a roadblock to ocean heat making it to the surface where the ice is located. The salt in the warm Atlantic ocean water makes it too dense. There's abundant heat in the Arctic Ocean lurking in the subsurface layer below the freshwater lens.

Which is basic knowledge in this forum. And has nothing to do with my (perhaps badly) made point. Which is this:  Open ocean absorbs heat from the sun, no matter what the density. Open land absorbs much less heat.

Your earlier post seemed to indicate that ice being close to a landmass would somehow receive a warming boost because of air advecting over the warmed-up land and out over the ice. But I would say that being far away from a landmass would give added boost to melt during summer, by air advecting over open ocean which, as I've said, absorbs much more heat than dry land and has therefore more heat to give to any passing air (besides being able to transfer the heat itself to some extent).

Let's agree to disagree binntho.

In a place like Siberia, you may have a 100k km2 land area immediately adjacent to the Arctic where the 2m air temp is all heated to 20C+.

One does not see 2m air temperatures over the open ocean approaching 20C in areas adjacent to the Arctic ice. The 2m air temps over the open ocean are moderated by the sea surface temperature and by definition the SST's adjacent to the ice are close to zero C. 

The ocean may be a superior absorber of heat, but I don't see examples where it is comparable to Siberia or N. America in terms of being able to produce a large area of highly warmed surface air which is in relatively close proximity to the ice. If there is such a section of ocean that meets this definition, please provide it.

[Phoenix]

If no new records are made for several years it will mean the trend has finally stabilized and the BOE strongly delayed.

I agree. If no new records for next 2 years, then we are not trending to BOE 2030.

Personally, if something catastrophic is coming out of the Arctic in the near future, I think it will be release of the fresh water accumulation from the Beaufort Gyre which slows the AMOC. That can happen anytime and a doom scenario I believe in.

Yes.  That will cool Northern Europe and disrupt the North Atlantic fishing industry.

A slowed AMOC is a game changer. Lessening the ocean's role as a vehicle for balancing heat between the equator and the poles will mean the atmosphere will have to take on a larger role and that means weather conditions above the surface will get screwy. Details of that are probably off-topic for this thread.

[pleun]

In a place like Siberia, you may have a 100k km2 land area immediately adjacent to the Arctic where the 2m air temp is all heated to 20C+.

One does not see 2m air temperatures over the open ocean approaching 20C in areas adjacent to the Arctic ice. The 2m air temps over the open ocean are moderated by the sea surface temperature and by definition the SST's adjacent to the ice are close to zero C. 

But heat is not only temperature. I think humid air over the ocean packs a bigger punch than warmer dry air.

[Freegrass]

But heat is not only temperature. I think humid air over the ocean packs a bigger punch than warmer dry air.

I think so too Pleun. That's why I like the "Total Precipitable Water" setting on Nullschool. It shows moisture and heat. I simply call it "energy" entering the arctic.

Here's the five day forecast.

(this probably belongs in the melting thread, but I'm posting it here in relation to your argument)

[Phoenix]

For purposes of making it easier to remember, I'll toss out a name for the theory which explains the location of the remaining ice at the minimum.. I'll call it DHAC-SOO.

It stands for Distant to: Heat Advecting Continents & Shallow Open Ocean.

I'm striving for simplicity.

1) There is basically no ice south of 75N at the minimum.

2) With very small exception, there is basically no ice at the minimum within 5 degrees of a continental land mass.

3) With very small exception, there is basically no ice remaining over the shallow shelves in open ocean. CAA is the exception to the shallow rule because its not in the open ocean.

The remaining ice at the minimum is in the other places for now. Hopefully, AGW won't get it too quickly.

Not Duck Soup....DHAC-SOO !!

[Freegrass]

For purposes of making it easier to remember, I'll toss out a name for the theory which explains the location of the remaining ice at the minimum.. I'll call it DHAC-SOO.

It stands for Distant to: Heat Advecting Continents & Shallow Open Ocean.

I'm striving for simplicity.

1) There is basically no ice south of 75N at the minimum.

2) With very small exception, there is basically no ice at the minimum within 5 degrees of a continental land mass.

3) With very small exception, there is basically no ice remaining over the shallow shelves in open ocean. CAA is the exception to the shallow rule because its not in the open ocean.

The remaining ice at the minimum is in the other places for now. Hopefully, AGW won't get it too quickly.

Not Duck Soup....DHAC-SOO !!

The arctic ocean has many secrets, but in the end, it's a simple system. It's a bathtub that gets heated in summer, and cooled down in winter. If it cools more than it warms, we get an ice age. If it warms more than it cools, you lose the ice. Rocks and things can have local effects, but in the end, it's still a simple system. So don't get lost in the details! They're only valuable for modeling...

[oren]

DHAC-SOO: As the 2nd lowest year on record had a large swath of ice down to 73^o over shallow water in the open ocean near a heat advecting continent, I think this theory is too simplistic and cannot help in making forecasts for the coming years.

I think the best localized predictor for ice-covered and ice-free regions at minimum and on other dates is the statistic for that particular region/sub-region/pixel over the past 10-15 years using NSIDC area data, and for the past 8 years using the high resolution AMSR2 UH area data, coupled with the statistics for adjacent regions/sub-regions/pixels using some kind of weighted average. This requires a combination of programming skills and time, which I have been lacking for the past few years.
As the statistics change over time this method can also show trends, but still cannot easily predict when a whole region first becomes ice-free.

[Phoenix]

DHAC-SOO: As the 2nd lowest year on record had a large swath of ice down to 73^o over shallow water in the open ocean near a heat advecting continent, I think this theory is too simplistic and cannot help in making forecasts for the coming years.

I think the best localized predictor for ice-covered and ice-free regions at minimum and on other dates is the statistic for that particular region/sub-region/pixel over the past 10-15 years using NSIDC area data, and for the past 8 years using the high resolution AMSR2 UH area data, coupled with the statistics for adjacent regions/sub-regions/pixels using some kind of weighted average. This requires a combination of programming skills and time, which I have been lacking for the past few years.
As the statistics change over time this method can also show trends, but still cannot easily predict when a whole region first becomes ice-free.

1) RE: The "large swath" of ice in the ESS in 2016 between 73N and 75N.

You're reaching here. That large swath is maybe 3% of the remaining volume in 2016? It's not there at the minimum in 2019. The location of the ice at the current minimum and it's correlation to certain variables isn't the theory. The high degree of correlation is a fact. The theory is that the proximity to land and bathymetry are important elements of causation.

2) The theory is too simplistic for making forecasts for future years.

I'm not attempting to make forecasts for future years. I'm just offering something for people to think about which gives them reason to question to validity of linear models being offered here. That's an important part of the scientific method. Kicking the tires.

Honestly, I wouldn't have much clue about which areas would go next or when they go.

3) Pixel level analysis

It would be super interesting to see more granular data. I'm sure there are some research professors or PhD students out there in the cryoscience field who are doing more detailed analysis. An AMA session here with some cryo-scientists would be cool if it could be arranged. 

4) Why do I even care?

I'm a doomer who hasn't given up yet. I retain the hope that mankind will rally and do what it needs to do to ensure the continuity of human civilization. To that end, I think it's important to give people the best possible information about likely ecological collapse in order to motivate them to avoid said collapse.

If a near term BOE is likely, then I think we should be shouting it from the rooftops. If a near term BOE is not likely, then I think it shouldn't be emphasized. If we hype outcomes that don't materialize, it gives deniers ammunition to call us alarmists and makes it easier for them to discredit legitimate science.

My goal is public awareness of the truth of AGW. My enemy is the denier / obfuscator who I compete with. Near term BOE is a pretty common thing that people talk about. If / when it doesn't happen, it makes my job harder. I believe that you want my job to be as easy as possible.

We're heading in the direction of civilization collapse regardless of BOE. There's a shitload of solid science in that regard. BOE is still a topic worthy of consideration, but the risk / reward of presenting it to the world as an extremely likely near term outcome doesn't seem to be a good one.
If it happens we're fucked anyway and if it doesn't, we've given the deniers more ammo to discredit the science.

I'm extremely hopeful that the remaining ice is much more resilient than the ice we've lost over the past 40 years. In the absence of any science demonstrating that hope as being misplaced, I think that hope is the pragmatic position to take. I hope you'll join me in that.

[The Walrus]

I would not make too much of where exactly the ice remains at minimum.  Many factors, including winds, currents, and local temperatures influence these remaining floes.  With regards to a BOE, I would agree that the focus such be on the ice north of 75 degrees.  If ice exists below that latitude, the odds of a BOE are nearly nil.

Your goal of public awareness of AGW is laudable.  I believe this becomes easier if one sticks to the truth, refraining from emotional or exaggerated claims.  If a near term BOE is likely, then it should be communicated as such, and the most likely consequences associated with it.  If not, then it should not be hyped for the exact reasons that you mentioned.  Oftentimes, when we feel that the general public is uninformed, we rush to educate them on the most dire consequences.  Unfortunately, these are the precise claims that the press rushes to print.  A slow, steady decline in sea ice over several decades does not garner the same headlines as the claim that the Arctic will be ice-free in a year or two.  In essences, some of us in the AGW community are our own worst enemy. 

As scientists, we can discuss them properly, and assign probabilities to their occurrences.  However,  I genuinely feel that we should portray the most likely scenarios to the general populace, less they become disillusioned with claims that fail to materialize in the short term.  Let others present such claims, and witness their demise.  The truth will prevail, and whoever is proclaiming it, will be viewed as the true prophets.

[oren]

Phoenix, deniers will always deny. No use guiding the conversation by seeking the side of least drama, although I understand your position perfectly.
About when will the Arctic go ice-free, this is a big unknown. There have been many threads about it in this section over past few years. So don't shout it from the rooftops until it actually happens, but don't think it can't happen either. Only time will tell, and by time I mean several years or a decade. It's a slow road.
OTOH, the ASI has crashed enough that anyone who can be alarmed is already alarmed. Those that close their ears will close their ears anyway.

[Phoenix]

Phoenix, deniers will always deny. No use guiding the conversation by seeking the side of least drama, although I understand your position perfectly.
About when will the Arctic go ice-free, this is a big unknown. There have been many threads about it in this section over past few years. So don't shout it from the rooftops until it actually happens, but don't think it can't happen either. Only time will tell, and by time I mean several years or a decade. It's a slow road.
OTOH, the ASI has crashed enough that anyone who can be alarmed is already alarmed. Those that close their ears will close their ears anyway.

I don't subscribe to the cynical view. Opinion polls show that climate is a rising issue in the US where I live. Yale conducts annual climate surveys. In my home state of Nevada, 93% of Democrats polled before the February caucus supported a Green New Deal. The % of Democrats who listed climate as the #1 issue increased by 3-4x from the 2016 primary to the 2020 primary. Climate caught health care as a leading issue for Democratic voters in 2020.

People are being converted through relentless messaging,  exposure to increasing and irrefutable evidence and peer pressure. The GOP is losing young voters.

ASIF as a whole is a great resource for climate science discussion. I value it. Thanks for your role in moderating.

[Glen Koehler]

 Last time I promise...
        ugh, I now realize that I can't use whole-Arctic values for linear regression because as peripheral seas melt out they can no longer contribute to the ice loss rate.  Thus a trend based on earlier years when those seas contributed to year-to-year losses is invalid because doing so contributes fictional negative volumes into the trend.

      The correction is to focus on the CAB trend only.  But even that combines earlier melting of the peripheral CAB areas with the areas where the final 1M km2 ice extent (the consensus threshold for BOE) is likely to occur.  The best approach I can think of is to partially correct for that by ignoring CAB data from 1979-1999 or 1979-2004 to discount the earlier years when the peripheral CAB areas still had September ice, and thus base the trend on losses after those areas were no longer contributing to the year-to-year losses.

      I looked at CAB Sept. average volume linear, polynomial and logarithmic trends for 2000-2019 and 2005-2019.  Starting in 2005 gives a flatter slope, and a lower R2 than starting in 2000.  Defining BOE threshold as 0.8K km3, the estimated date for when there is a 50% cumulative chance for having had a first year of BOE is 3.5 years earlier when using the 2000-2019 data.  The estimate for 95% cumulative chance of BOE is 5.3 years earlier using the 2000-2019 data.  And the date for when each Single year has a 50% chance of being <BOE threshold is 6.2 years earlier when using the 2000-2019 data.

      I also looked at CAA as a potential last refuge area, which was interesting because it brought to my attention the fact that in 2012 the CAA Sept. volume was almost down to zero. But then I found this map of the where the last 1M km2 ASI extent is expected posted by Tor Bejnar at https://forum.arctic-sea-ice.net/index.php/topic,417.msg122285.html#msg122285

     Comparing that "final ice extent map" to the NSIDC Arctic seas boundary lines at https://nsidc.org/data/masie/browse_regions shows that the final ice refuge is almost (perhaps entirely) within the NSIDC defined boundaries of the CAB.  In addition, the CAA trend has it blinking out shockingly soon.  So the estimates below are derived from the 2005-2019 CAB-only volume trend.  The CAB volume data are from PIOMAS, which defines the CAB boundary somewhat differently than NSIDC, but that should not affect the resulting date estimates. 

      I am using 0.8K km3 as the volume threshold proxy for BOE. If you don't like that, we all at least agree that 0 volume means zero extent.  This is the most demanding possible definition for BOE.  Using a 0 volume threshold only delays the 50% and 95% cumulative chance dates, and the 50% Single year date, by 3.0 to 3.3 years.  Thus, whether we define BOE as 0.8K km3 or Zero km3 doesn't make much difference.

      Enough fussbudgetry, here are the numbers for my latest and last attempt to use linear regression to estimate BOE dates.
 
*************************
DATA: PIOMAS Central Arctic Basin (CAB) September average volume estimates based on linear trend from 2005-2019.
Threshold: 0.8K km3  as an arbitrary definition of BOE based on volume.
Values rounded to the nearest whole percent.

Year    Chance of that single year     Chance of at least one year since 2019
            going below 0.8K km3           having gone below 0.8 km3
2020          0%                                        0%
2021          0%                                        1%
2022          1%                                        1%
2023          1%                                        2%
2024         2%                                         4%
2025         3%                                         6%
2026         4%                                       10%
2027         6%                                       15%
2028         8%                                       22%
2029       12%                                       31%
2030       16%                                       42% 
2031       21%                                       55%  First BOE more likely than not by 2031
2032       28%                                       67%
2033       34%                                       78%
2034       42%                                       87%
2035       49%                                       94%
After 2035, each single year has >50% chance of BOE
2036       57%                                       97%
2037       65%                                       99%
2038       72%                                     100%
2039       78%
2040       83%

    Finally, several points that seem worth repeating. 
     1) Some of the difference in intuitive impression for when the ASI will reach BOE status is due to the difference between the cumulative chance of a first year BOE (almost certainly to be followed by subsequent non-BOE years), and when each individual year has >50% of going below a BOE threshold.
   
     2) As experienced ASIF members have commented, melt season weather will determine when the first BOE occurs, and the Arctic may have already reached a state where BOE is within reach of an extremely warm and significant cyclone year.  But the trend analysis shown above estimates that ia first BOE year s unlikely until after 2030.

     3) Scientists who study ASI decline say that any such estimates are subject to +/- 20 years error due to high variability in the ASI!  (In other words "You don't know, and we don't know either!").

     4) BOE is an arbitrary marker, and the world won't end when there is a first BOE.  But it serves as an important observable and easily understood marker with visual impact to communicate climate change progression.  As the late-summer ASI ice decline continues and eats into August (which trails September by only a few years) and into July, the effects on albedo, weather, ecosystems etc., and the consequent impacts on human civilization, will increase.   

[Tor Bejnar]

It was actually a Ned W map that I cross-posted on the Arctic Maps thread.