Arctic Sea Ice Extent, Area, and Volume

[Typhoon Tip]

Don't we typically observe the nadir this week ?

[ORH_wxman]

On 9/5/2023 at 10:29 AM, Typhoon Tip said:

Don't we typically observe the nadir this week ?

Yeah area tends to be at a minimum near the end of the first week of September or early in the 2nd week.

Right now, we're about 40k above the min two days ago. We'll see if that 9/3 min holds.

[madwx]

NSIDC extent at 4.28 and area at 2.63 with some time to drop more, especially extent. Definitely have had some excitement to end the melting season

[forkyfork]

area dropped to third lowest

[ORH_wxman]

17 hours ago, forkyfork said:

area dropped to third lowest

Unless there was a revision to past area, I have 2023 as 4th lowest. Either way, this is one of the stronger late melts on record post-August 1st.

2012: 2.228429

2016: 2.463209

2020: 2.5859

2023 low so far is 2.63, but it's only at 2.67 right now, so it could fall further into 3rd still.

The area file I used to reference is no longer active, so I don't have a good way to look up revisions. But I do have the old data saved. Is there a good archive of the data somewhere now? I only have cryosphere computing, but that only gives us a single day shot and doesn't seem to archive the daily numbers.

 

[forkyfork]

i got that from a sea ice forum post. it must have been third lowest for the date and not overall min

[ORH_wxman]

1 hour ago, forkyfork said:

i got that from a sea ice forum post. it must have been third lowest for the date and not overall min

Data retrieval has gotten significantly worse in the past 2 years. Pretty frustrating since it should be easier considering the raw concentration daily data is available on NSIDC. It's just nobody is putting into numeric form anymore.

[so_whats_happening]

41 minutes ago, ORH_wxman said:

Data retrieval has gotten significantly worse in the past 2 years. Pretty frustrating since it should be easier considering the raw concentration daily data is available on NSIDC. It's just nobody is putting into numeric form anymore.

It has been extremely hard to find data of recent for many things not just sea ice. I do not like it. The temps are holding rather steadily above average across the Arctic so it will definitely be a slow start this year. The Russian land temps are finally cooling down enough that it should help with starting the refreeze season soon but we may have seen our lowest for the year and just a very slow start as we head to October.

[bluewave]

On 6/24/2023 at 9:48 AM, bluewave said:

The model run by Will Gregory at CPOM based on May conditions is going for an average September extent of 4.46 million sq km. This is solidly within the range since 2013. The model has been very accurate. So the 2012 record should be safe for another year 

https://www.arcus.org/files/sio/panarctic/2023_june_sio_cpom_ucl_gregory_et_al.pdf

Executive summary" of your Outlook contribution (using 300 words or less) describe how and why your contribution was formulated. To the extent possible, use non-technical language.
This statistical model computes a forecast of pan-Arctic September sea ice extent . Monthly averaged May sea ice concentration and sea-surface temperature fields between 1979 and 2023 were used to create a climate network (based on the approach of Gregory et al 2020). This was then utilised in a Bayesian Linear Regression in order to forecast September extent. The model predicts a pan-Arctic extent of 4.46 million square kilometres. Sea ice concentration data were taken from NSIDC (Cavalieri et al., 1996; Maslanik and Stroeve,1999) and sea-surface temperature data were taken from ERA5 (Hersbach et al., 2019)

Average September extents since 2012

 

 

2022….4.87

2021……4.92

2020……3.92

2019……4.32

2018…...4.71

2017……4.87

2016……4.72

2015…..4.63

2014…..5.28

2013…..5.35

2012…..3.60

The model looks like it’s going to be relatively close again based on the May melt pond data. The current daily extent is 4.31 million sq km. So another normal September by post 2012 warmer and thinner sea ice standards. Which means that when September gets averaged out, it should fall somewhere in the 4s. Tough to get below 4 or above 5 these days for a September average extent. 

[bdgwx]

The NSIDC extent went up a bit on 9/18. The 4.21 km2 on 9/17 could be our minimum.

[ORH_wxman]

Area has been pretty flat over the last few days....up a little bit, but not quite enough yet to call the minimum with high confidence. We are currently at 2.72 million sq km which is 90k above the min back on 9/12. If we tick up to 2.8 million sq km in the next couple days, then I think it will be safe to call the min.

[Typhoon Tip]

20 hours ago, ORH_wxman said:

Area has been pretty flat over the last few days....up a little bit, but not quite enough yet to call the minimum with high confidence. We are currently at 2.72 million sq km which is 90k above the min back on 9/12. If we tick up to 2.8 million sq km in the next couple days, then I think it will be safe to call the min.

I'm sure you're likely versed in this but for the general audience ...

This positive AO projection from the GEFs-based CPC source is a relatively new telecon modality. 

If it has legs, you want that at this time of year - if your rooting on recovery rate of sea ice

Also ... land-based cryospheric concerns while we are at it. Like I've said in the past, the rate of recovery along the Siberian to Canadian archipelago sector, wrt to autumn land-snow/cryosphere appears to be a better correlation to N/A American Dec temperature.  

Now, ...that's just one in many correlations in the pallet.  So that is just a coarser sort of perspective. But positive Arctic Oscillations, while being a warm signal for the mid latitudes beneath the 60.. .70th parallels, is actually a cold and stormy one above amidst the arctic domain itself. 

This is a new modality ... it may strengthen, or it may not have legs and settle back.  Frankly, as we tip deeper into the cold, ... the correlation between AO and warm ENSOs begins to favor the negative AO.  But for now that's not yet significant. If the positive AO strengthens some it might be interesting to see if there is a gain/acceleration, however small or large notwithstanding, that temporarily exceeds the normal seasonal growth that occurs post the sea ice nadir, wrt to sea ice. And while we are at it, do we observe land snow pack advance in those polar regions.

[ORH_wxman]

On 7/3/2023 at 12:49 PM, ORH_wxman said:

Official 2023 prediction on Arctic Sea Ice minimums:

Area: 3.1 million sq km (+/- 300k)...this would rank 8th lowest

Extent: 4.7 million sq km (+/- 500k)....this would rank 13th lowest

 

This is based on years of using NSIDC area from the SSMI/S satellite which gets fooled by meltponding into thinking it's open water. This means that we can use NSIDC area as a proxy for meltponding and some of the most robust literature on predicting minimums in advance is using meltponding in May and June.

 

I give more detail on my method here in last year's prediction, but didn't feel the need to repeat it every single year so I will just link it for those who are interested in how I make these predictions:

 

 

I will include the graph I usually post,

Keep in mind that this graph is not a graph of historical minimums....it is simply what 2023 minimum would be if we followed that previous year's melt rate AFTER July 1st. The reason this works fairly well is most of the decadal trend in ice loss has been due to increases in melting prior to July 1st, not after.

But it gives us a nice distribution of possible results if we mirrored previous season's melt rate from post-July 1st.....from this graph, you can see that if we had a post-July 1st melt rate like 2016 (currently the biggest post-July 1st melt on record), then we'd end up with a minimum of 2.65 million sq km....that would still be over 400k above the record minimum in 2012 which is why we can rule out a new record this year. There just wasn't enough melting prior to July 1st to challenge the 2012 record. Likewise, you can see how even a higher retention post-July 1st would still produce a minimum below 4 million sq km....so it is safe to say that we will not have a 4 million sq km min for the first time since 2006. The distribution of this year's possible results looks quite mundane in the post-2007 context...and that is reflected in my predictions above.

 

 

After several solid gains on the weekend, it is now safe to call the NSIDC area minimum....the minimum was set back on 9/12 at 2.63 million sq km. This ranks as the 4th lowest minimum behind 2012, 2016, and 2020.

 

This is 470k below my prediction of 3.1 million sq km and outside my confidence intervals of + or - 300k. You can ironically see in the post above that I said if we matched the record melt after July 1st that occurred in 2016, we'd end up with a min at 2.65 million sqkm....well, we not only matched it, but slightly exceeded it. 2023 is the new record high ice loss after July 1st which blew up my prediction. This is my first real miss since 2016 when we had a similar post-July 1st melt. The odds of a huge post-July 1st melt hadn't really changed much over time based on the empirical data prior to this season, but I might incorporate a very slight handicap in that direction during next year's prediction....though even a slight handicap would not have saved this year's prediction. That is just the nature of trying to use 300k error bars when your standard deviation on this is 375k. So I'd expect to miss about 30-35% of the time using pure statistics....but I usually incorporate medium range weather patterns and also the distribution of the ice into the predictions as well which historically has helped make the misses less frequent than the purely statistical 30-35%.

The extent minimum looks like it was around 4.25 million sq km which is barely still inside my confidence interval on extent.

A very warm blocky pattern on the North American side was the main culprit in the strong post-July 1st melt. Typically, this area from the CAA to the Beaufort side of the CAB struggles to melt down if there isn't a lot of preconditioning in May/June, but this year we able to overcome mediocre preconditioning with the very strong blocking pattern that set up in July and August.

[bluewave]

6th lowest annual daily minimum extent. Most of the daily minimum extents have been in the 4s since the lower Arctic ice sea era began in 2007. 12 finishes in the 4s…3 in the low 5s…and 2 in the 3s.

https://nsidc.org/arcticseaicenews/2023/09/arctic-sea-ice-minimum-at-sixth/

 

1	2012	3.39	1.31	Sept. 17
2	2020	3.82	1.47	Sept. 16
3	2007 2016 2019	4.16 4.17 4.19	1.61 1.61 1.62	Sept. 18 Sept. 10 Sept. 18
6	2023	4.23	1.63	Sept. 19
7	2011	4.34	1.68	Sept. 11
8	2015	4.43	1.71	Sept. 9
9	2008 2010	4.59 4.62	1.77 1.78	Sept. 19 Sept. 21
11	2018 2017 2022	4.66 4.67 4.70	1.80 1.80 1.81	Sept. 23 Sept. 13 Sept. 19
14	2021	4.77	1.84	Sept. 16
15	2014 2013	5.03 5.05	1.94 1.95	Sept. 17 Sept. 13
17	2009	5.12	1.98	Sept. 13

 

 

[Typhoon Tip]

16 hours ago, bluewave said:

6th lowest annual daily minimum extent. Most of the daily minimum extents have been in the 4s since the lower Arctic ice sea era began in 2007. 12 finishes in the 4s…3 in the low 5s…and 2 in the 3s.

https://nsidc.org/arcticseaicenews/2023/09/arctic-sea-ice-minimum-at-sixth/

 

1	2012	3.39	1.31	Sept. 17
2	2020	3.82	1.47	Sept. 16
3	2007 2016 2019	4.16 4.17 4.19	1.61 1.61 1.62	Sept. 18 Sept. 10 Sept. 18
6	2023	4.23	1.63	Sept. 19
7	2011	4.34	1.68	Sept. 11
8	2015	4.43	1.71	Sept. 9
9	2008 2010	4.59 4.62	1.77 1.78	Sept. 19 Sept. 21
11	2018 2017 2022	4.66 4.67 4.70	1.80 1.80 1.81	Sept. 23 Sept. 13 Sept. 19
14	2021	4.77	1.84	Sept. 16
15	2014 2013	5.03 5.05	1.94 1.95	Sept. 17 Sept. 13
17	2009	5.12	1.98	Sept. 13

 

 

Seems that's a week or two later than the 1900 thru 2000 climo?

[forkyfork]

extent gets more and more irrelevant as the ice cover thins out. 15% coverage of thin floes is not the same as 100% coverage of thick ice 

[bluewave]

On 6/24/2023 at 9:48 AM, bluewave said:

The model run by Will Gregory at CPOM based on May conditions is going for an average September extent of 4.46 million sq km. This is solidly within the range since 2013. The model has been very accurate. So the 2012 record should be safe for another year 

https://www.arcus.org/files/sio/panarctic/2023_june_sio_cpom_ucl_gregory_et_al.pdf

Executive summary" of your Outlook contribution (using 300 words or less) describe how and why your contribution was formulated. To the extent possible, use non-technical language.
This statistical model computes a forecast of pan-Arctic September sea ice extent . Monthly averaged May sea ice concentration and sea-surface temperature fields between 1979 and 2023 were used to create a climate network (based on the approach of Gregory et al 2020). This was then utilised in a Bayesian Linear Regression in order to forecast September extent. The model predicts a pan-Arctic extent of 4.46 million square kilometres. Sea ice concentration data were taken from NSIDC (Cavalieri et al., 1996; Maslanik and Stroeve,1999) and sea-surface temperature data were taken from ERA5 (Hersbach et al., 2019)

Average September extents since 2012

 

 

2022….4.87

2021……4.92

2020……3.92

2019……4.32

2018…...4.71

2017……4.87

2016……4.72

2015…..4.63

2014…..5.28

2013…..5.35

2012…..3.60

Another outstanding job by this model with the September average extent finishing at 4.37 million sq km. This is nearly identical to the forecast of 4.46 million sq km issued back in June based on May melt pond data.

 

 

[forkyfork]

 

[Eskimo Joe]

On 9/13/2023 at 3:09 PM, forkyfork said:

area dropped to third lowest

 

What a time to be alive. 

[Prismshine Productions]

Antarctic really having a bad time right now, started the year +1.377m km above last year's mark, currently down to only +558k km above last year, poor penguins...

Sent from my SM-S146VL using Tapatalk

[Prismshine Productions]

Antarctica now only up +513k over 2023... Pace continues we may hit a new record minimum for the 3rd year in a row

Sent from my SM-S146VL using Tapatalk

[FPizz]

On 1/20/2024 at 12:53 PM, Prismshine Productions said:

Antarctica now only up +513k over 2023... Pace continues we may hit a new record minimum for the 3rd year in a row

Sent from my SM-S146VL using Tapatalk
 

My neighbor just went there on a cruise.  Looked really fun.  I didn't even realize they had cruises here. 

[Prismshine Productions]

Antarctic tracker:
1/20: +513k over 2023
1/21: big time melt, +426k over 2023

Extent currently 3.385m km² with 7 weeks of melt season left

Sent from my SM-S146VL using Tapatalk

[Prismshine Productions]

Antarctic 1/22 update:

3.296m km2 extent (-89k)

+393k over 2023, margin continues to shrink

Sent from my SM-S146VL using Tapatalk

[Prismshine Productions]

1/23 Antarctic update:

3.207m km2 (-89k)

+362k over 2023 (31k gap shrinkage)

Sent from my SM-S146VL using Tapatalk

[chubbs]

 

[Prismshine Productions]

 

Funny enough, the arctic currently has the same extent as it did on Valentines Day 2023, so we a few weeks up this year

Sent from my SM-S146VL using Tapatalk

[Prismshine Productions]

1/24 Anarctic update:

3.123m km2 extent (-94k)

+347k over 2023, -15k to the gap

Sent from my SM-S146VL using Tapatalk

[chubbs]

9 hours ago, Prismshine Productions said:

Funny enough, the arctic currently has the same extent as it did on Valentines Day 2023, so we a few weeks up this year

Sent from my SM-S146VL using Tapatalk

 

Good blog article. One thing that interested me is the differences in timing between Arctic and Antarctic. Hints of a see-saw, with the Antarctic having the bigger losses recently.

[bowtie`]

I do not see any spaghetti lines anywhere near where we are right now.