2018 sea ice area and extent data

[Cid_Yama]

Well, someone's not right.

https://seaice.uni-bremen.de/data/amsr2/today/extent_n_running_mean_amsr2_regular.png

Appears to be tracking 2007, and significantly below 2017.

[gerontocrat]

Well, someone's not right.

https://seaice.uni-bremen.de/data/amsr2/today/extent_n_running_mean_amsr2_regular.png

Appears to be tracking 2007, and significantly below 2017.

I understood JAXA and Univ Bremen use the same data from the AMSR2 sensor. But what is meant by "running_mean". All is confusion. JAXA and NSIDC is what I use and I cannot change that. Is anyone able to explain?

Meanwhile, I have invented a new measure (unless someone else has already done it). See my next post.

[gerontocrat]

Towards a single number to show change in Arctic Sea Ice over time.

The trigger for this attempt came from one word “Atlantification”, which refers to the northward  retreat of the ice edge along a front from North Greenland through North of Svalbard, Franz Josef Land the New Siberian Islands deep into the ESS. The effect of this “Atlantification” is to further change the nature of the climate of that part of the Arctic Ocean and bordering land masses (and beyond) from one dominated by an ice surface to one dominated by open water, i.e. from a sea-ice climate to a maritime climate.

We all, scientists, amateurs, the media, look at maxima and minima to determine changes in the 14 seas of the Arctic. But this can be misleading. This year, the Kara Sea was very late in starting to melt, but then melted very fast to an area and extent well below the 2010’s average (and almost a record low). The obvious conclusion is that it showed sea ice to be declining. But perhaps that is wrong, as for a large part of that melting season there was more ice in the Kara Sea than the 2010’s average. In other words,  Atlantification of the Kara Sea in 2018 maybe went into reverse. The opposite may apply in other seas, where extent and/or area and/or volume decline might appear to be stalled or even reversed, but actual ice in the sea over the year has declined.

Calculating the measure of Atlantification (or trend to a maritime sea environment).
This is the simplest part.
1.    Average Daily Ice Area for each year = Sum of the area for the 365 days of the year and  divide by 365
2.   Divide by the 1980’s maximum to give the proportion of sea covered by ice during the year.
3.   Then Open Water Percentage is 1 minus the proportion of sea covered by ice during the year expressed as a percentage

The first results are attached below. The 2018 figure is only to the 7th September. When and how quickly freeze kicks in will determine the final outcome.

EDIT: Greenland Sea data corrected in the table (Thanks, Stephan)

[oren]

Great new index. (Note another version of it could multiply by insolation, but simpler is often better.)
I do think that proper comparison with previous years can only be made if past years' index is calculated until the same date as 2018.

[gerontocrat]

Great new index. (Note another version of it could multiply by insolation, but simpler is often better.)
I do think that proper comparison with previous years can only be made if past years' index is calculated until the same date as 2018.

Early days yet, Oren -  Getting Excel to change the ranges to sum automatically as the date changes is the one trick I have not yet cracked. (Using the date to change everything else has been pretty straightforward). Maybe my index goes into cold storage until the end of the year.

Adding Insolation? No.  The value of my measure, if it has any, is to give an idea of the progress towards changing from an ice sea to a maritime sea. Adding insolation would mask that indicator. Surely an open water sea in winter changes the impact on climate compared with an ice-covered sea just as much as in summer. The insolation effect is all about ocean heating potential and tealight has done all that superbly, and once again is all about part of the year - the brief Arctic summer.

I am still doing a write-up on why I think the measures we use at the moment inadequately measure change in the Arctic Ocean environment. I will put it up in a day or two, though perhaps not in this thread.

[Juan C. García]

[ADS-NIPR-JAXA] ASI Extent.

September 8^th, 2018: 4,609,439 km², a drop of -53,615 km².
2018 is the 8^th lowest on record.

[Brigantine]

2018 is now below the 2010 minimum.

2003     5933760
2004     5683663
2006     5625046
2005     5179300
2009     5054055
2014     4884120
2013     4809288
2010     4622092
2018     4609439 [to date]
2008     4500623
2017     4472225
2011     4269199
2015     4257003
2007     4065739
2016     4017264
2012     3177455

[gerontocrat]

JAXA ARCTIC EXTENT 4,609,439 km2（September 8, 2018）

Just to add to Juan's post...
- Extent is now 228 k km2  (4.9%) above the 2010's average extent and 123 k (2.7%) above 2017 on this date,,
- Extent loss to date is now a slightly less very large 616 k km2 (6.2 %) below the 2008-2017 average, with 99.4% of the average melting season done.

Resulting minimum from average remaining melt is down a tad to 4.55 million km2. Range of results from last ten years remaining melt is 4.46 to 4.61 million km2.  An end result significantly above 2017 now seems inevitable (or is it).

That 2017 feeling has waxed consistently for what seems a long time - melting has slowed to below average for many days, but on these last 2 days, extent loss has been much above average. There is, on average, under 1% (4 days) of further low and highly variable extent losses to go. Could the melting season last a bit longer than that with a late burst of extent loss - Yes.  However further extent loss would have to be quite large for 2018 to shift from 8th place (See post from Brigantine). On the other hand, could extent loss become gains ? Yes. Indeed, a minimum of 4.61 million km2 (September 8 extent) is possible.

[Rob Dekker]

Gerontocrat, that is an interesting new index.
I have a couple of questions about it :

Calculating the measure of Atlantification (or trend to a maritime sea environment).
This is the simplest part.
1.    Average Daily Ice Area for each year = Sum of the area for the 365 days of the year and  divide by 365
2.   Divide by the 1980’s maximum to give the proportion of sea covered by ice during the year.
3.   Then Open Water Percentage is 1 minus the proportion of sea covered by ice during the year expressed as a percentage

Do I understand this correctly that your index is calculated over one year (Jan 1 - Dec 31) ?
And that the index calculated for a running year (like 2018) is calculated based on the Daily Sea Ice Area from Jan 1, 2018 on to current date ?

If so, wouldn't your index show a lot of variability in the early months of the year (Jan, Feb), and mellow out at the last months of the year (Nov, Dec) ?

Also, since your index represents "open water" integrated over a year, we could see an BOE (ice free Arctic) in September and that would not register on your index as anything special.

So I wonder which "information" your index actually would provide.

Does this make any sense ?

[gerontocrat]

Gerontocrat, that is an interesting new index.
I have a couple of questions about it :

Calculating the measure of Atlantification (or trend to a maritime sea environment).
This is the simplest part.
1.    Average Daily Ice Area for each year = Sum of the area for the 365 days of the year and  divide by 365
2.   Divide by the 1980’s maximum to give the proportion of sea covered by ice during the year.
3.   Then Open Water Percentage is 1 minus the proportion of sea covered by ice during the year expressed as a percentage

Do I understand this correctly that your index is calculated over one year (Jan 1 - Dec 31) ?
And that the index calculated for a running year (like 2018) is calculated based on the Daily Sea Ice Area from Jan 1, 2018 on to current date ?

If so, wouldn't your index show a lot of variability in the early months of the year (Jan, Feb), and mellow out at the last months of the year (Nov, Dec) ?

Also, since your index represents "open water" integrated over a year, we could see an BOE (ice free Arctic) in September and that would not register on your index as anything special.

So I wonder which "information" your index actually would provide.

Does this make any sense ?

I think I was wrong to add 2018 to the table. The index needs 365 days of data to show the position - maybe using the previous 365 days from the current month-day for each year would allow a useful index calculation during the year. I'll think about that.

 " we could see an BOE (ice free Arctic) in September and that would not register on your index as anything special." Yes - but only if that BOE was not something special. We were all excited by the 2007 and 2012 melts when they happened - but in the grand scheme of things they were not the Great Leap Forward to Atlantification of the Arctic. A few days of very little ice makes a difference but not a lot.

The intention was not to create a measure that would replace all others as in Douglas Adams' "42" but to create a measure that gives a better idea of progress in the conversion of the Arctic Ocean from an ice desert to a maritime open water sea over the last 40 years.  The climate works on an annual basis, so a measure for the whole year is needed. I am doing a write up on my thoughts on why such a measure is needed, but decided to chuck the results into the arena to see the reaction. Already I am having a think about reactions received so far..

[Dharma Rupa]

The intention was not to create a measure that would replace all others as in Douglas Adams' "42" but to create a measure that gives a better idea of progress in the conversion of the Arctic Ocean from an ice desert to a maritime open water sea over the last 40 years.  The climate works on an annual basis, so a measure for the whole year is needed. I am doing a write up on my thoughts on why such a measure is needed, but decided to chuck the results into the arena to see the reaction. Already I am having a think about reactions received so far..

I'd rather see two (or three) numbers.  One calculated from the beginning of September to the end of March and one from the end of March to the beginning of September (and your current calculation or a variation with the year starting end of March).

(I think the Winter number would be the real Atlantification number.)

[Steven]

NSIDC sea ice area is currently 3.20 million km2, which is 9th lowest for the date:



Comparison with previous years' minima:

1.  2012   2.23
2.  2016   2.46
3.  2011   2.92
4.  2017   2.94
5.  2007   2.95
6.  2010   3.07
7.  2008   3.08
8.  2015   3.14
9.  2018   3.20 (year to date)
10. 2009   3.55
11. 2014   3.58
12. 2013   3.61
13. 2006   4.08
14. 2005   4.13
...

[Shared Humanity]

Great new index. (Note another version of it could multiply by insolation, but simpler is often better.)
I do think that proper comparison with previous years can only be made if past years' index is calculated until the same date as 2018.

Early days yet, Oren -  Getting Excel to change the ranges to sum automatically as the date changes is the one trick I have not yet cracked. (Using the date to change everything else has been pretty straightforward). Maybe my index goes into cold storage until the end of the year.

Adding Insolation? No.  The value of my measure, if it has any, is to give an idea of the progress towards changing from an ice sea to a maritime sea. Adding insolation would mask that indicator. Surely an open water sea in winter changes the impact on climate compared with an ice-covered sea just as much as in summer. The insolation effect is all about ocean heating potential and tealight has done all that superbly, and once again is all about part of the year - the brief Arctic summer.

I am still doing a write-up on why I think the measures we use at the moment inadequately measure change in the Arctic Ocean environment. I will put it up in a day or two, though perhaps not in this thread.

Great metric and your chart already reveals that every region is seeing a steady increase in open water by decade. It also shows a dramatic drop in 2018 for open water in the Hudson.

(cue bbr2314 to shout hysterically about the approaching ice age in 3...2...1...)

Since this measure is intended to capture open water percentage for the year, couldn't you simply measure it using a calendar year? This would serve to capture the trend towards early melt and late freeze, particularly in the peripheral seas.

(edit: just read your response above so I now understand it is calendar year.)

Question: Would having this measure tracked quarterly as well as annually reveal anything?

[oren]

Getting Excel to change the ranges to sum automatically as the date changes is the one trick I have not yet cracked. 

You could try this workaround - add another column with an IF that checks if the date is past the last date available for 2018. If so it gives a zero, if not it gives a 1.
In your calculation, multiply the summed data by this column. All later dates not available for 2018 are automatically removed. Your sum range remains the full year, but the actual summing is over your desired range, the rest is zero.
I think this is important as the index is accumulation-based, and summing over a shorter range gives lower numbers if I understand correctly (hence the low HB number for example). OTOH it would be a pity to wait until year-end to have a number for 2018.

[Pmt111500]

I'd rather see two (or three) numbers.  One calculated from the beginning of September to the end of March and one from the end of March to the beginning of September (and your current calculation or a variation with the year starting end of March).

(I think the Winter number would be the real Atlantification number.)

I'll add one to three to get four and suggest going by celestial days, equinoxes and solstices, you'd still get a winter number after december solstice. If you go by two please do it by equinoxes so there's a physical reason to it, there's probably more use to such an index that takes in account the available sunlight. You could also calculate it for only the period when it's dark or wholly lighted say... Above 75N?

[Dharma Rupa]

I'd rather see two (or three) numbers.  One calculated from the beginning of September to the end of March and one from the end of March to the beginning of September (and your current calculation or a variation with the year starting end of March).

(I think the Winter number would be the real Atlantification number.)

I'll add one to three to get four and suggest going by celestial days, equinoxes and solstices, you'd still get a winter number after december solstice. If you go by two please do it by equinoxes so there's a physical reason to it, there's probably more use to such an index that takes in account the available sunlight. You could also calculate it for only the period when it's dark or wholly lighted say... Above 75N?

I was thinking about going by the Sun, but there is the problem of a delay, and I can't think of any physical justification for "equinox plus 15."  You want to more or less capture the usual max and the usual min as the boundaries -- but I can't say how.

[gerontocrat]

Thanks for all the comments everybody - I think. I've got a headache.

and Dammit Oren, I used a similar trick with Lotus 1-2-3 to switch entire sheets on and off. (Such a better package for a number cruncher). Mind you, it was back in 1996 in Indonesia. memory failure .

The measure is not one for a daily update - the change in one day is minuscule. But I will put it in and use it e.g. when Neven declares the melting season over.

I also see value in Mid-September to Mid-March (freezing season) and Mid-March to Mid-September melting calculations and a half-year to June 30 calculation. I can promise no more than that. (I still have to earn some loot part-time to eke out my pathetic pensions - rolling stones gather no moss).

Anyway - here is a table just for Total Arctic Area. No graphs yet. Only 14 seas to go.

Comments welcome ( as I must be a masochist)

EDIT I changed the maximum area I used from the absolute maximum to the maximum of the 1980's average daily values- which is a slightly lower figure. This would avoid any accusations of cherry picking data for the starting point. The effect is to slightly increase the percentages of open water throughout the time series but does not affect the trend.

[Dharma Rupa]

Thank you!  Now all I have to do is try to absorb that...

(first impression is that it is pretty much year-round, but a graph or two might help.)

[Dharma Rupa]

P.S. If we are going to talk about Atlantification then eventually we are going to have to distinguish Pacifying.  Probably should get the basic idea sorted out first.

[Michael Hauber]

Early days yet, Oren -  Getting Excel to change the ranges to sum automatically as the date changes is the one trick I have not yet cracked. (Using the date to change everything else has been pretty straightforward). Maybe my index goes into cold storage until the end of the year.

Put dates in column a, values in column b, first date in c1, last date in d1.  Formula: sumifs(b:b,a:a,">=" & c1,"<=" & d1) will add up everything between those dates.  Just watch for potential problems such as converting 8/2/2018 from 8th Feb to 2nd Aug.

[tvansant]

"Soon after NASA’s Ice, Cloud and land Elevation Satellite-2, or ICESat-2, launches on Sept. 15 from Vandenberg Air Force Base in California, it will start collecting a terabyte of data a day to monitor the height of Earth’s surface below. The mission will measure the amount of sea ice floating above the ocean surface to within an inch. It will measure annual changes across ice sheets to within the width of a pencil."

https://www.nasa.gov/feature/goddard/2018/icesat-2-scientists-to-investigate-icy-mysteries

Overcoming many challenges, ICESat-2 is currently scheduled to launch on Saturday, September 15th 5:46 am PDT (8:46am EDT) from Vandenburg AFB in California.

[slow wing]

Wonderful! Thanks for posting, tvansant.

A couple of snippets...


Nathan Kurtz, NASA Goddard
"What I’m really excited to see are the month-to-month changes in sea ice thickness, because we didn’t have that with the first ICESat..."

"..right now there’s a lot of assumptions made about how snow accumulates on the ice over the year. With ICESat-2, we should be able to see things like the effect of a passing storm."


Thorsten Markus, NASA Goddard
"What I’m really most excited about is seeing these data come in at such a high resolution...
Over sea ice you’ll see all these gaps and ridges, and discover just how rough that remote ice is."

[Juan C. García]

[ADS-NIPR-JAXA] ASI Extent.

September 9^th, 2018: 4,603,771 km², a drop of -5,668 km².
2018 is the 8^th lowest on record.

[gerontocrat]

JAXA ARCTIC EXTENT 4,603,771 km2（September 9, 2018）

Just to add to Juan's post...
- Extent is now 236 k km2  (5.1%) above the 2010's average extent and 132 k (2.9%) above 2017 on this date,,
- Extent loss to date is now a slightly less very large 624 k km2 (6.3 %) below the 2008-2017 average, with 99.5% of the average melting season done.

Resulting minimum from average remaining melt is 4.55 million km2. Range of results from last ten years remaining melt is 4.46 to 4.60 million km2.  An end result significantly above 2017 now seems inevitable.

That 2017 feeling has waxed consistently for what seems a long time - melting has slowed to again after a brief 2 days of above average loss. There is, on average, just 0.5% (3 days) of further low and highly variable extent losses to go. Could the melting season last a bit longer than that with a late burst of extent loss - Yes.  However further extent loss would have to be very large for this time of year for 2018 to shift from 8th place (See post from Brigantine). On the other hand, could extent loss become gains ? Yes. Indeed, a minimum of 4.60 million km2 (September 9 extent) is very possible.

ps: Somewhat of a contrast with the other end of the planet

[binntho]

Yes. Indeed, a minimum of 4.01 million km2 (September 9 extent) is very possible.

Should presumably be 4.60 million km2? Unless I've misunderstood something?

[gerontocrat]

Yes. Indeed, a minimum of 4.01 million km2 (September 9 extent) is very possible.

Should presumably be 4.60 million km2? Unless I've misunderstood something?

How did that get into the post? Damn

[gerontocrat]

NSIDC Total Area as at 9 September (5 day trailing average) =  3,232,605  km2

This is down to 185 k above the 2010-2017 average total area for this date

Total Area loss             36 k , 6 k below average for the day, (28k on 8 Sep)

Central Seas loss         37 k,                                 (29k on 8 Sep)
Peripheral Seas gain       2 k, all seas <5% of max. (2k on 8 Sep)
Other Seas loss              1 k, all seas <5% of max.  (0k on 8 Sep)

Analysis of individual seas.
Pacific Side
- The Bering Sea - finished,
- Chukchi Sea loss 1 k, area 32 k, < 5% of maximum.
Atlantic Side
- Baffin Sea loss 2 k, area 19 k, <5% of max,
- Greenland Sea gain 2k km2 (suspicious?), area 52 k,
- Barents Sea - finished,
- The Kara Sea area loss 1 k, area 27 k, < 5% of maximum.
- The Laptev Sea area loss 1 k, area 12 k, <5% of maximum.

CAB
- Beaufort Sea loss 9 k,
- The Canadian Archipelago loss 0 k,
- East Siberian Sea loss 11 k .
- The Central Arctic Sea loss 13 k,

Other seas
- St Lawrence -finished,
- Hudson Bay area loss 0 k, Area  7 k,
- The Okhotsk Sea - finished .

On average, this is when daily area loss declines sharply and wobbles from day to day. The area loss for the last 2 days has been well above average, 25+ k above average on the 9th. Extent loss (also 5-day trailing average)) was 34 k, 8k  above average for the day. Daily extent loss 26k after loss 39k, gain 6k, loss 65k ,gain 32k, in the previous days.  End of season wobbles continue.

East Siberian Sea continues high late season losses (Graph Attached)

[gerontocrat]

PERCENTAGES OF OPEN WATER IN THE ARCTIC

I started with one figure - the calendar year. After including all the comments - too many figures

Graph and revised table for total Arctic attached. But what I wanted to look was at differences sea by sea. Rethink on what data to post required and when. My laptop was wheezing before I started adding all this into the spreadsheet. Anybody got a spare Cray lying around?

[Dharma Rupa]

PERCENTAGES OF OPEN WATER IN THE ARCTIC

I started with one figure - the calendar year. After including all the comments - too many figures

Dividing it into freezing season and thawing season told me less than I expected.  It looks like it might be losing ice slightly faster in Winter, but not enough to write home about.  Fine with me if you lose that measure.

Getting it by sea, or by Atlantic versus Pacific, seems much more important.

P.S.  Neven's division into 4 sections such that two of them are the Atlantic and Pacific sides and two are the continents might be nice.

[gerontocrat]

PERCENTAGES OF OPEN WATER IN THE ARCTIC

I started with one figure - the calendar year. After including all the comments - too many figures

Dividing it into freezing season and thawing season told me less than I expected.  It looks like it might be losing ice slightly faster in Winter, but not enough to write home about.  Fine with me if you lose that measure.

Getting it by sea, or by Atlantic versus Pacific, seems much more important.

P.S.  Neven's division into 4 sections such that two of them are the Atlantic and Pacific sides and two are the continents might be nice.

Still very much a work-in-progress. Still staggering through sea by sea.
But the attachments show what I am after.

Bering Sea - now a maritime sea, (that sometimes has a bit of ice in it)
Central Arctic - still an ice desert, (that sometimes has a bit of open water in it)

[Dharma Rupa]

The graphs are a bit hard to interpret.  Is the software smart enough to put a vertical line between the averages and the last three years?

[Dharma Rupa]

Bering Sea - now a maritime sea, (that sometimes has a bit of ice in it)
Central Arctic - still an ice desert, (that sometimes has a bit of open water in it)

The chart makes it rather clear that the CAB is less of a desert than it used to be.

[gerontocrat]

The graphs are a bit hard to interpret.  Is the software smart enough to put a vertical line between the averages and the last three years?

No, but maybe I am

[Dharma Rupa]

The graphs are a bit hard to interpret.  Is the software smart enough to put a vertical line between the averages and the last three years?

No, but maybe I am

Hehehe...You see what my problem was then.

[magnamentis]

Wonderful! Thanks for posting, tvansant.

A couple of snippets...


Nathan Kurtz, NASA Goddard
"What I’m really excited to see are the month-to-month changes in sea ice thickness, because we didn’t have that with the first ICESat..."

"..right now there’s a lot of assumptions made about how snow accumulates on the ice over the year. With ICESat-2, we should be able to see things like the effect of a passing storm."


Thorsten Markus, NASA Goddard
"What I’m really most excited about is seeing these data come in at such a high resolution...
Over sea ice you’ll see all these gaps and ridges, and discover just how rough that remote ice is."

great news all this, will most probably and hopefully put an end to some volume discussions and perhaps even improve extent and area measurements.

i'm looking forward with great interest to see how the measured the ice-thickness will compare against modeled ice-thickness of the past. i expect significant improvements of perhaps hope and the wish are fathering the thought ?

[gerontocrat]

The graphs are a bit hard to interpret.  Is the software smart enough to put a vertical line between the averages and the last three years?

No, but maybe I am

Hehehe...You see what my problem was then.

Will the attached do? ('cos it's all you're getting)

[Dharma Rupa]

The graphs are a bit hard to interpret.  Is the software smart enough to put a vertical line between the averages and the last three years?

No, but maybe I am

Hehehe...You see what my problem was then.

Will the attached do? ('cos it's all you're getting)

It makes it clear that there is a difference...I guess noting the way the lines bend at the decade is a bit too pedantic.

This is fine.

[gerontocrat]

It makes it clear that there is a difference...I guess noting the way the lines bend at the decade is a bit too pedantic.

This is fine.

The data is arranged so I can do the time series for each year from 1980 without too much grief.
Probably the second effort when phase 1 is done. But also (Neven's 4 regions) vs. Atlantic periphery, Pacific periphery, the CAB and those 3 seas not really connected to the Arctic Ocean (Lawrence, Okhotsk, Hudson)) to consider.

But here a taster of phase 2

But cut some slack to this poor clapped out analyst and his poor clapped out laptop. Weep, sniffle, reach for handkerchief.

[Dharma Rupa]

But cut some slack to this poor clapped out analyst and his poor clapped out laptop. Weep, sniffle, reach for handkerchief.

My poor slave driven....whatever

I will say that you might have found a way to capture WACCy.

[oren]

Great analysis gerontocrat, capturing and integrating lots of data in a few charts. And I think your metric shows nicely that the 2007-2018 short term trend has not been flat.

[Juan C. García]

[ADS-NIPR-JAXA] ASI Extent.

September 10^th, 2018: 4,619,083 km², an increase of 15,312 km².
2018 is the 8^th lowest on record.

[Cid_Yama]

Once again, someone is not right.

https://seaice.uni-bremen.de/data/amsr2/today/extent_n_running_mean_amsr2_previous.png

https://seaice.uni-bremen.de/data/amsr2/today/extent_n_running_mean_amsr2_regular.png

[gerontocrat]

JAXA ARCTIC EXTENT  4,619,083 km2（September 10, 2018）

Just to add to Juan's post...
- Extent is now 244 k km2  (5.3%) above the 2010's average extent and 146 k (3.2%) above 2017 on this date,,
- Extent loss to date is now a very large 633 k km2 (6.4 %) below the 2008-2017 average.

The resulting minimum from average remaining melt is 4.57 million km2. Range of results from last ten years remaining melt is 4.46 to 4.62 million km2.  An end result significantly above 2017 now seems inevitable.

That 2017 feeling has waxed consistently for what seems a long time - melting has switched to gain after a brief 2 days of above average loss. There is, on average, just 50k km2 (2 days) of further low and highly variable extent losses to go. Could the melting season last a bit longer than that with a late burst of extent loss - Yes.  However further extent loss would have to be very large for this time of year for 2018 to shift from 8th place. On the other hand, could extent loss become gains, as it was on this day ? Yes. Indeed, a minimum of 4.60 million km2 (September 9 extent) is very possible.

Now maybe is the time to switch a lot of the attention to which seas continue melting and which commence refreeze.

ps: Somewhat of a complete contrast with the other end of the planet. (I wonder if there is a "bi-polar see-saw" going on?)

[gerontocrat]

NSIDC Total Area as at 10 September (5 day trailing average) =  3,214,414  km2

This is down to 173 k above the 2010-2017 average total area for this date

Total Area loss             18 k , 12 k above average for the day,

Central Seas loss         22 k,                                 
Peripheral Seas gain      4 k, all seas <5% of max.
Other Seas loss             0 k, all seas <5% of max.

Analysis of individual seas.
Pacific Side
- The Bering Sea - finished,
- Chukchi Sea loss 0 k, area 31 k, < 5% of maximum.
Atlantic Side
- Baffin Sea gain 2 k, area 21 k, <5% of max,
- Greenland Sea gain 2k km2 , area 54 k,
- Barents Sea - finished,
- The Kara Sea area loss 0 k, area 27 k, < 5% of maximum.
- The Laptev Sea area loss 1 k, area 12 k, <5% of maximum.

CAB
- Beaufort Sea loss 8 k,
- The Canadian Archipelago loss 5 k,
- East Siberian Sea loss 8 k .
- The Central Arctic Sea gain 1 k,

Other seas
- St Lawrence -finished,
- Hudson Bay area loss 0 k, Area  7 k,
- The Okhotsk Sea - finished .

On average, this is when daily area loss declines sharply and wobbles from day to day. Even at 18k, this day's loss is 12k more than the average on this date (end of season). Extent loss (also 5-day trailing average)) was 35 k, 20k  above average for the day. Daily extent loss 73k after loss 26k, loss 39k, gain 6k, loss 65k , in the previous days.  End of season wobbles continue. But it looks like melting may still have some legs left.

[Juan C. García]

[ADS-NIPR-JAXA] ASI Extent.

September 11^th, 2018: 4,571,173 km², a drop of -47,910 km².
2018 is the 8^th lowest on record.

[gerontocrat]

JAXA ARCTIC EXTENT  4,571,173 km2（September 11, 2018）

Just to add to Juan's post...
- Extent is now 185 k km2  (4.1%) above the 2010's average extent and 70 k (1.5%) above 2017 on this date,,
- Extent loss to date is now a very large 633 k km2 (6.4 %) below the 2008-2017 average.

The resulting minimum from average remaining melt is 4.53 million km2. Range of results from last ten years remaining melt is 4.42 to 4.57 million km2.  An end result significantly above 2017 now seems inevitable.

That 2017 feeling has waxed consistently for what seems a long time - but this day's extent loss was very much above average. There is, on average, just 1 day of further low and highly variable extent loss to go. Could the melting season last a bit longer than that with a late burst of extent loss - Yes.  However further extent loss would have to be very large for this time of year for 2018 to shift from 8th place. On the other hand, could extent loss become gains, as it was on this day ? Yes. Indeed, a minimum of 4.57 million km2 (September 11 extent) is possible.

Now maybe is the time to switch a lot of the attention to which seas continue melting and which commence refreeze.

[oren]

2018 is now below the 2010 minimum.

2003     5933760
2004     5683663
2006     5625046
2005     5179300
2009     5054055
2014     4884120
2013     4809288
2010     4622092
2018     4609439 4571173 [to date]
2008     4500623
2017     4472225
2011     4269199
2015     4257003
2007     4065739
2016     4017264
2012     3177455

Reposting Brigantine's useful post. With a 70,500 km2 further loss 2018 can move to 7th, and with a 102,000 km2 loss it can move to 6th. Both are still possible, though probability unknown. Lower than 6th is practically impossible.

[gerontocrat]

NSIDC Total Area as at 11 September (5 day trailing average) =  3,210,074  km2

This is down to 167 k above the 2010-2017 average total area for this date

Total Area loss             4 k , 12 k above average for the day,

Central Seas loss         6 k,                                 
Peripheral Seas gain    1 k, all seas <5% of max.
Other Seas loss           0 k, all seas <5% of max.

Analysis of individual seas.
Pacific Side
- The Bering Sea - finished,
- Chukchi Sea loss 1 k, area 30 k, < 5% of maximum.
Atlantic Side
- Baffin Sea gain 0 k, area 21 k, <5% of max,
- Greenland Sea gain 1k km2 , area 55 k,
- Barents Sea - finished,
- The Kara Sea area loss 0 k, area 27 k, < 5% of maximum.
- The Laptev Sea area gain 1 k, area 12 k, <5% of maximum.

CAB
- Beaufort Sea loss 0 k,
- The Canadian Archipelago loss 3 k,
- East Siberian Sea loss 11 k .
- The Central Arctic Sea gain 8 k,

Other seas
- St Lawrence -finished,
- Hudson Bay area loss 0 k, Area  7 k,
- The Okhotsk Sea - finished .

Eastern Siberian Sea still losing ice - graph attached.

The season is nearly over. Nowt to do but watch day by day.

[lurkalot]

No-one seems to have addressed Cid_Yama's two posts pointing out that Bremen's charts show 2018 having a lower extent than 2017 and being closer to 3rd place than 8th. Furthermore, today's Charctic shows 2017 and 2018 neck and neck, with 2018 looking likely to be the lower in the next day or two.

As always, comparisons are made more difficult by different sorts of averaging. Could anyone (Wipneus?) with access to Bremen, Hamburg and NSIDC actual figures check how they compare with JAXA? Thanks.

[Wipneus]

As always, comparisons are made more difficult by different sorts of averaging. Could anyone (Wipneus?) with access to Bremen, Hamburg and NSIDC actual figures check how they compare with JAXA? Thanks.

The ADS/NIPR numbers are as they are. They undoubtedly use Jaxa data, but which data and how it is processed I have not been able to figure out. There is Jaxa sea ice concentration data, from which extent and area can be fairly easily calculated but that gives different numbers.

Here are the lowest NSIDC extent numbers:
NH Extent
1 2012-09-16 3.339905
2 2016-09-07 4.082952
3 2007-09-14 4.146931
4 2011-09-08 4.332572
5 2015-09-08 4.387092
6 2008-09-18 4.548265
7 2010-09-19 4.589683
8 2017-09-12 4.611109
9 2018-09-11 4.624741
10 2014-09-16 4.988244

The Uni Hamburg AMSR2 extent numbers are only available since 2012:
1 2012-09-16 3.066197
2 2016-09-07 3.736899
3 2015-09-08 4.014171
4 2018-09-11 4.294476
5 2017-09-11 4.334096
6 2014-09-14 4.615457
7 2013-09-09 4.694411

Bremens AMSR2 3 .125km data is incomplete, no September data for 2012-2015:
1 2016-09-07 3.763777
2 2018-09-11 4.321027
3 2017-09-12 4.355720
4 2012-07-31 5.787364
5 2014-10-20 7.580769
6 2015-07-01 9.180398
7 2013-11-23 10.180262


I will post the area numbers in a minute.