Arctic Sea Ice Extent, Area, and Volume

[ORH_wxman]

Update:

NSIDC SIA stood at 8.39 million sq km as of 6/20/2022.

 

Here's where other years stand now:

2021: -580k

2020: -720k

2019 -470k

2018 +90k

2017: -50k

2016: -420k

2015: -170k

2014: -80k

2013: +60k

2012: -740k

2011: -330k

2010: -450k

2009: +710k

2008: +70k

2007: -610k

 

You can see there's some pretty clear divergence now between the huge melt years and the higher retention years. There's still some time for movement (2008 lost a lot in the final 10 days of June), but 2022 is quickly trending toward years like 2018/2017/2014/2013. There is a dipole pattern setting up over the next week or so....if 2022 is going to make a move toward the big dog melt years, it has to be during this next week.

[GaWx]

51 minutes ago, ORH_wxman said:

Update:

NSIDC SIA stood at 8.39 million sq km as of 6/20/2022.

 

Here's where other years stand now:

2021: -580k

2020: -720k

2019 -470k

2018 +90k

2017: -50k

2016: -420k

2015: -170k

2014: -80k

2013: +60k

2012: -740k

2011: -330k

2010: -450k

2009: +710k

2008: +70k

2007: -610k

 

You can see there's some pretty clear divergence now between the huge melt years and the higher retention years. There's still some time for movement (2008 lost a lot in the final 10 days of June), but 2022 is quickly trending toward years like 2018/2017/2014/2013. There is a dipole pattern setting up over the next week or so....if 2022 is going to make a move toward the big dog melt years, it has to be during this next week.

 These numbers translate to another nice gain relative to 2007-2021 with 2022 now at +240K vs the mean of those years compared to +120K four days ago and +70K five days ago. Fwiw, the north of 80N temperatures averaged out continue to be slightly colder than the 1958-2002 mean, which can't hurt especially considering every of the last 40 days has been colder than that mean.

[ORH_wxman]

35 minutes ago, GaWx said:

 These numbers translate to another nice gain relative to 2007-2021 with 2022 now at +240 vs the mean of those years compared to +120 four days ago and +70 five days ago. Fwiw, the north of 80N temperatures averaged out continue to be slightly colder than the 1958-2002 mean, which can't hurt especially considering every of the last 40 days has been colder than that mean.

When I first started looking more into NSIDC SIA as a predictor around 10 years ago, I almost chose 6/20 as the cutoff date because it's not that much less accurate than 6/30....so I think most of this season is likely baked into the pie at this point. There have a been a few years (like the previously mentioned 2008) which really went crazy the final 10 days of the month to end up closer to the bigger melt seasons, but a large majority of the seasons were pretty baked by the solstice.

 

I think the best-guess theory (with the assistance of some literature on melt ponding) is that there simply isn't enough time in the melt season to make up for a lack of melt-ponding in the first 3 weeks of June. If you fall behind in the melt-ponding during that period, you pretty much cannot catch back up unless you really have some crazy anomalous weather. Most of the melting in the 2nd half of the summer is from bottom melt rather than surface melt (sfc temps cool back below freezing pretty quickly late summer), so you need as much surface melt momentum as possible early on to achieve a big melt season...and lots of melt-ponding helps enhance sfc melt by lowering albedo and increasing heat transfer, etc.

[ORH_wxman]

Update:

The melt has picked up a bit with the dipole pattern currently in progress across the arctic basin. As of 6/22, the NSIDC SIA stood at 8.09 million sq km.

 

Here's how other years compared on the same date:

2021: -520k

2020: -430k

2019: -390k

2018: +350k

2017: +20k

2016: -480k

2015: +30k

2014: +70k

2013: +270k

2012: -650k

2011: -140k

2010: -480k

2009: +710k

2008: +50k

2007: -460k

 

The dipole pattern looks like it will persist for another 4 days or so and then the low currently in the ESS will migrate toward the Beaufort/CAB and produce a reverse dipole which would slow the sfc melt down again.

So the time to rack up losses is right now and the next few days.

[GaWx]

2 hours ago, ORH_wxman said:

Update:

The melt has picked up a bit with the dipole pattern currently in progress across the arctic basin. As of 6/22, the NSIDC SIA stood at 8.09 million sq km.

 

Here's how other years compared on the same date:

2021: -520k

2020: -430k

2019: -390k

2018: +350k

2017: +20k

2016: -480k

2015: +30k

2014: +70k

2013: +270k

2012: -650k

2011: -140k

2010: -480k

2009: +710k

2008: +50k

2007: -460k

 

The dipole pattern looks like it will persist for another 4 days or so and then the low currently in the ESS will migrate toward the Beaufort/CAB and produce a reverse dipole which would slow the sfc melt down again.

So the time to rack up losses is right now and the next few days.

  This means that the last two days vs the average of the last 15 years has gone from +240K to +140K.

  Recap vs avg. of last 15 years as of:

6/15/22: +70K

6/16/22: +120K

6/20/22: +240K

*6/22/22: +140K

 

*Edited due to typo

[so_whats_happening]

Ain't looking good with another round of warmth showing up after the 4th of July across Alaska and the very northern reaches of Canada.

[GaWx]

 Tuesday 6/28 was the first day in 1.5 months with it not colder than normal north of 80N:

[bluewave]

The early forecasts based on the May conditions in the Arctic are forecasting an average September extent  in the 4.3 to 4.5 million sq km range.The 15 year era from 2007 to 2021 featured 10 out of 15 years finishing in the 4s. Only two years finished below 4.00 million sq km. With just three finishing slightly above 5.00 million sq km. 
 

September 2022 forecast 

 

https://www.arcus.org/files/sio/33269/cpom_ucl_gregory_et_al.pdf

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 2022 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 4.5 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)
Brief explanation of Outlook method (using 300 words or less).
Monthly averaged May sea ice concentration (SIC) and sea-surface temperature (SST) data between 1979 and 2022 were used to create a May SIC-SST climate(complex) network. Individual SIC grid cells were first clustered into regions of spatio-temporal homogeneity (and similarly for SST) by using a community detection algorithm (see Gregory et al, 2020). Links between each of these network regions (covariance) were then passed into a Bayesian Linear Regression to derive an estimate on the prior distribution of the regression parameters. Subsequently a posterior distribution of the regression parameters was then derived in order to generate the forecast of September sea ice extent.

https://www.arcus.org/files/sio/33269/cpom.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.
We predict the September ice extent 2022 to be 4.3 (3.8-4.8) million km2. This is just above the trend line. In spite of the large sea ice extent in May 2022, sea ice thickness and melt pond cover are quite normal with respect to the last decade.
Brief explanation of Outlook method (using 300 words or less).
This is a statistical prediction based on the correlation between the ice area covered by melt- ponds in May and ice extent in September. The melt pond area is derived from a simulation with the sea ice model CICE in which we incorporated a physically based melt-pond model1. See our publication in Nature Climate Change http://www.nature.com/nclimate/journal/v4/n5/full/nclimate2203.html for details2.

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

2011……4.61

2010…..4.90

2009….5.36

2008….4.67

2007…..4.28

 

[ORH_wxman]

Update:

The accelerated melt ponding from the dipole has pushed 2022 more toward the higher melt years of the pack now instead of being near the high-retention years....but still not into the highest tier level like 2012.

NSIDC SIA on 6/29/22 stood at 7.12 million sq km. Below are how other years compared on the same date:

2021: +40k

2020: -50k

2019: -280k

2018: +490k

2017: +310k

2016: +20k

2015: +410k

2014: +400k

2013: +560k

2012: -510k

2011: -30k

2010: -310k

2009: +910k

2008: +550k

2007: -180k

 

I'm away the next few days but I'll have my final prediction for SIA minimum when I get back and the 6/30 and 7/1 data is in. Unless there are drastic changes over the next couple days, this year's prediction is probably going to be somewhere between a top 5 and top 10 melt year....I don't think it has quite enough juice to make a top 3 and the dipole recently has done enough damage to take a result like 2009/2013/2014/2018 out of the equation.

[GaWx]

1 hour ago, ORH_wxman said:

Update:

The accelerated melt ponding from the dipole has pushed 2022 more toward the higher melt years of the pack now instead of being near the high-retention years....but still not into the highest tier level like 2012.

NSIDC SIA on 6/29/22 stood at 7.12 million sq km. Below are how other years compared on the same date:

2021: +40k

2020: -50k

2019: -280k

2018: +490k

2017: +310k

2016: +20k

2015: +410k

2014: +400k

2013: +560k

2012: -510k

2011: -30k

2010: -310k

2009: +910k

2008: +550k

2007: -180k

 

I'm away the next few days but I'll have my final prediction for SIA minimum when I get back and the 6/30 and 7/1 data is in. Unless there are drastic changes over the next couple days, this year's prediction is probably going to be somewhere between a top 5 and top 10 melt year....I don't think it has quite enough juice to make a top 3 and the dipole recently has done enough damage to take a result like 2009/2013/2014/2018 out of the equation.

  This means that during the last nine days vs the average of the last 15 years, 2022 has had a pretty dramatic 380K downturn from +240K to -140K (just over 40K/day).

  Recap of current vs avg. of last 15 years as of:

6/15/22: +70K

6/16/22: +120K

6/20/22: +240K

6/22/22: +140K

6/29/22: -140K

[ORH_wxman]

46 minutes ago, GaWx said:

 

  This means that during the last nine days vs the average of the last 15 years, 2022 has had a pretty dramatic 380K downturn from +240K to -140K (just over 40K/day).

  Recap of current vs avg. of last 15 years as of:

6/15/22: +70K

6/16/22: +120K

6/20/22: +240K

6/22/22: +140K

6/29/22: -140K

Yeah the melt-ponding has been very extensive over the Beaufort/Chukchi/CAA portion of the ice....see below. The first image is the NSIDC (SSMI/S source) one and the second is AMSR2. Note how the concentration is much higher on the AMSR2 image because it's not fooled by melt-ponding nearly as much like SSMI/S is.

 

 

[bluewave]

We have discovered a negative feedback for Arctic sea ice loss that was unknown back in the historic 2007-2012 decline era. Most observers in 2012 thought that the record set that season would probably be eclipsed within several years. But the rapid rebound in 2013 and 2014 began to change that idea. A paper came out in 2013 that showed the much thinner ice would lead to slower rates of extent decline in the near future. A study below published last year builds in this understanding. So given these findings, it’s no surprise that the 2022 melt season so far looks pretty average for the last decade with nothing close to the rates of loss seen in 2012. 
 

Recent Slowdown in the Decline of Arctic Sea Ice Volume Under Increasingly Warm Atmospheric and Oceanic Conditions

Jinlun Zhang

First published: 25 August 2021

https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2021GL094780

4 Concluding Remarks

PIOMAS shows a significant decline of Arctic SIV over much of the period 1979–2020, in line with previous studies (e.g., Kwok & Rothrock, 2009; Lindsay & Schweiger, 2015). However, it also shows that the Arctic SIV decline is slowing down during 2007–2020, which is in line with CS2 observations of largely stabilized Arctic SIV over 2011–2020. The slowdown of Arctic SIV decline occurs despite increasingly warm SAT and UOT. The reason for the slowdown of the SIV decline is because of a change in ice mass balance such that the decrease in ice export at the open boundaries of the Arctic exceeds the decrease in net ice production inside the Arctic.

The magnitude of ice export depends on SIV or thickness and ice motion. The relatively strong decrease in ice export in 2007–2020 is attributed mainly to two factors: (a) Ice is thinning inside the Arctic Ocean and therefore less available for export. (b) The increase in ice speed is lower than the decrease in SIV or thickness and unable to drive more ice out of the Arctic (Zhang et al., 2012). Thus, the behavior of ice export at a time of SIV decline serves as a negative feedback to retard the decline.

The relatively weak decrease in net ice production in 2007–2020 is because of a strong increase in ice growth in fall and winter, which compensates for some of the ice loss in summer due to elevated ice melt associated with ice-albedo feedback. The strong increase in ice growth is attributed mainly to three factors: (a) Thinner and less compact ice has much higher growth rates than thicker ice under freezing conditions. (b) There is an increase in dynamic open water creation due to enhanced ice divergence and shear deformation as a thinner ice cover is more mobile and easier to deform. (c) There is a decrease in surface net longwave radiation in fall and winter because of an increase in surface temperature that promotes upward longwave radiation.

Open water creation is closely correlated with ice growth in much of the Arctic, particularly in the Canada Basin where the correlation is often above R = 0.8. By increasing the area of open water, the open water creation process helps to boost ice growth in fall and winter in 2007–2020. The behavior in ice growth at a time of SIV decline, aided by increased open water creation induced by increased ice motion and deformation, serves as a negative feedback to retard the decline, and therefore plays a role in the slowdown of the Arctic SIV decline during 2007–2020 under increasingly warm atmospheric and oceanic conditions.

Here, 2007 is selected as a starting year to examine the slowdown of the Arctic SIV in recent years. This is based on the consideration that 2007 saw a record low summer ice extent at that time, before a new record set in 2012. Nevertheless, the selection is somewhat arbitrary, and one can certainly select a different starting year for analysis. However, moderately shifting the starting year away from 2007 (e.g., 2005, 2006, 2008, and 2009) would not fundamentally change this model study's conclusions that a slowdown of the Arctic SIV decline has occurred in recent years.

Note that the model simulated Arctic SIV drops from 1979 to a local minimum in 1982 and then peaks in 1987 (Figure 2a). There is no significant trend in SIV during the period 1979–1987. While there is no significant trend in SIV either during the period 2007–2020, the later period differs from the early period 1979–1987 in two key aspects: (a) SIV in 2007–2020 is much lower than in 1979–1987, and (b) SAT and UOT are climbing increasingly higher in 2007–2020, while dropping in 1979–1987. The thinner ice cover during 2007–2020 leads the ice export and growth processes to play a role in serving as a negative feedback to slow down the SIV decline, which is not seen in 1979–1987. It is expected that such a role may become more prominent in the future. In other words, the slowdown of the Arctic SIV decline may continue for some time in the future unless a stronger Arctic warming than the present would occur. Whether it is true remains to be seen through enhanced observations and modeling.

 

[GaWx]

 After June 28th became the first day in ~45 days back up to normal, the average Arctic temperature north of 80N went right back down and is now near the coldest for July 4th going back to 1958, which is just above freezing (see image below). Variance in summer is very low. The last time it was this cold on July 4th was in 2014. If the past is a good indication, it should level off near here. The coldest on any date in midsummer back to 1958 is only down to near -0.5 C. (The blue line is at 0C.) The warmest is +2.5 C, set in 1991, 94, and 95.

 

[ORH_wxman]

Update:

On 7/1, the NSIDC SIA sootd at 7.07 million sq km. Here's how other years compared on the same date:

2021: -70k

2020: -510k

2019: -450k

2018: +320k

2017: +10k

2016: -240k

2015: +130k

2014: +260k

2013: +340k

2012: -650k

2011: -210k

2010: -380k

2009:+640k

2008: +220k

2007: -400k

 

For the minimum Sea ICe Area prediction, we can use prior years' melt out from July 1st onward to estimate how this tear will turn out because there hasn't been much of a trend in ice loss after 7/1....the big trend has been prior to 7/1. We have maybe seen a slight increase in melt from 7/1 since 2007, so I will weight those years more.

Below is a table of how 2022 would turn out if it followed previous years' melt out from 7/1 onward....

 

 

For example, if we used 1989's ice loss from here on out, we'd finish at 2.66 million sq km. 2016 had the highest meltout beyond 7/1 on record, and would produce a final SIA min of 2.55 million sq km if we followed that path in 2022. You'll note that the record year of 2012 is only like 4th or 5th most ice loss from 7/1 onward which means most of the damage was done prior to 7/1. We can pretty much rule out a new record this season based on this data. We'd need to obliterate the post 7/1 loss record set in 2016 to achieve it. A top 3 lowest min is probably out too.....likewise, a top 3 highest min in the post-2007 context is likely out as well....though another 2010 from here on out would achieve it. All other post-2007 years would fail, however.

 

All that said, the average ice loss from 7/1 onward in the post-2007 era is 4.09 million sq km which would produce a 2022 min of 2.98 million sq km (7.07 million minus 4.09 million). I'm going to stick very close to this number and go with a final minimum of 3 million sq km +/- 300k. I might hedge a little higher if the forecast was colder on the Beaufort/CAA side over the next week but they will continue to see mild weather so I will stick with 3 million.

 

Final extent prediction is a lot harder than area because extent relies a lot upon compaction/dispersion which is really hard to forecast. But that said, usually something in the 3 million sq km range for area will produce an extent min on NSIDC of around 4.5-4.8 million sq km. So I will go with 4.6 million +/- 500k....I have larger error bars on the extent. (side note: JAXA extent usually comes in around 200k lower than NSIDC extent after their algorithm update post-2013....so I will not be using Jaxa to verify this prediction. Only NSIDC daily extent)

[GaWx]

2 hours ago, ORH_wxman said:

Update:

On 7/1, the NSIDC SIA sootd at 7.07 million sq km. Here's how other years compared on the same date:

2021: -70k

2020: -510k

2019: -450k

2018: +320k

2017: +10k

2016: -240k

2015: +130k

2014: +260k

2013: +340k

2012: -650k

2011: -210k

2010: -380k

2009:+640k

2008: +220k

2007: -400k

 

For the minimum Sea ICe Area prediction, we can use prior years' melt out from July 1st onward to estimate how this tear will turn out because there hasn't been much of a trend in ice loss after 7/1....the big trend has been prior to 7/1. We have maybe seen a slight increase in melt from 7/1 since 2007, so I will weight those years more.

Below is a table of how 2022 would turn out if it followed previous years' melt out from 7/1 onward....

 

 

For example, if we used 1989's ice loss from here on out, we'd finish at 2.66 million sq km. 2016 had the highest meltout beyond 7/1 on record, and would produce a final SIA min of 2.55 million sq km if we followed that path in 2022. You'll note that the record year of 2012 is only like 4th or 5th most ice loss from 7/1 onward which means most of the damage was done prior to 7/1. We can pretty much rule out a new record this season based on this data. We'd need to obliterate the post 7/1 loss record set in 2016 to achieve it. A top 3 lowest min is probably out too.....likewise, a top 3 highest min in the post-2007 context is likely out as well....though another 2010 from here on out would achieve it. All other post-2007 years would fail, however.

 

All that said, the average ice loss from 7/1 onward in the post-2007 era is 4.09 million sq km which would produce a 2022 min of 2.98 million sq km (7.07 million minus 4.09 million). I'm going to stick very close to this number and go with a final minimum of 3 million sq km +/- 300k. I might hedge a little higher if the forecast was colder on the Beaufort/CAA side over the next week but they will continue to see mild weather so I will stick with 3 million.

 

Final extent prediction is a lot harder than area because extent relies a lot upon compaction/dispersion which is really hard to forecast. But that said, usually something in the 3 million sq km range for area will produce an extent min on NSIDC of around 4.5-4.8 million sq km. So I will go with 4.6 million +/- 500k....I have larger error bars on the extent. (side note: JAXA extent usually comes in around 200k lower than NSIDC extent after their algorithm update post-2013....so I will not be using Jaxa to verify this prediction. Only NSIDC daily extent)

This 7/1/22 update translates to a nice recovery vs just two days prior of +210K vs the average of the last 15 years from -140K to +70K. That's quite a reversal after the 380K loss (+240K to -140K) vs this average over the nine day period preceding these last two days. Is this merely a temporary recovery or the start of a new trend? We'll see.
 

Recap of current vs avg. of last 15 years as of:

6/15/22: +70K

6/16/22: +120K

6/20/22: +240K

6/22/22: +140K

6/29/22: -140K

7/1/22:     +70K

[forkyfork]

1 hour ago, GaWx said:

This 7/1/22 update translates to a nice recovery vs just two days prior of +210K vs the average of the last 15 years from -140K to +70K. That's quite a reversal after the 380K loss (+240K to -140K) vs this average over the nine day period preceding these last two days. Is this merely a temporary recovery or the start of a new trend? We'll see.
 

Recap of current vs avg. of last 15 years as of:

6/15/22: +70K

6/16/22: +120K

6/20/22: +240K

6/22/22: +140K

6/29/22: -140K

7/1/22:     +70K

recovery to what? pre 2007 ice levels are not coming back

[GaWx]

47 minutes ago, forkyfork said:

recovery to what? pre 2007 ice levels are not coming back

 I'm referring to recovering vs the average of 2007-2021, which are the years that @ORH_wxman has been comparing to ITT since June 15th. I'm not referring at all to pre-2007. I agree that pre-2007 levels aren't coming back.

[ORH_wxman]

Update: 

Now having data 10 days into July, we’ve seen 2022’s melt slow even further relative to the post-2007 field. On 7/10, NSIDC area stood at 6.35 million sq km. Here are where other years stood compared to 2022 on the same date:

2021: -480k

2020: -620k

2019: -820k

2018: -190k

2017: -20k

2016: -540k

2015: -90k

2014: +200k

2013: +20k

2012: -680k

2011: -280k

2010: -270k

2009: +230k

2008: +130k

2007: -350k

 

Theres currently a deep, cold low pressure over most of the CAB right now and that is not forecasted to move much over the next 5-6 days….so I’m skeptical we play much catch-up in the next week. There are signs that it may try and drift more to the ESS/Chukchi beyond that time frame, so perhaps the melt will pick up again if that happens. But we’re wasting what’s left of the prime sun angle up there in the meantime. 

[GaWx]

27 minutes ago, ORH_wxman said:

Update: 

Now having data 10 days into July, we’ve seen 2022’s melt slow even further relative to the post-2007 field. On 7/10, NSIDC area stood at 6.35 million sq km. Here are where other years stood compared to 2022 on the same date:

2021: -480k

2020: -620k

2019: -820k

2018: -190k

2017: -20k

2016: -540k

2015: -90k

2014: +200k

2013: +20k

2012: -680k

2011: -280k

2010: -270k

2009: +230k

2008: +130k

2007: -350k

 

Theres currently a deep, cold low pressure over most of the CAB right now and that is not forecasted to move much over the next 5-6 days….so I’m skeptical we play much catch-up in the next week. There are signs that it may try and drift more to the ESS/Chukchi beyond that time frame, so perhaps the melt will pick up again if that happens. But we’re wasting what’s left of the prime sun angle up there in the meantime. 

 So, this is a continuation of the sudden reversal toward recovery vs the 2007-21 average that started just after 6/29. It has actually reversed to the best position (+250K) vs that 15 year period since that period started being followed here on June 15th, barely bettering the +240K of 6/20/22. This is a 390K recovery vs that 15 year average over just the last 11 days:

 

Recap of current vs avg. of last 15 years as of:

6/15/22: +70K

6/16/22: +120K

6/20/22: +240K

6/22/22: +140K

6/29/22: -140K

7/1/22:     +70K

7/10/22:  +250K

[bluewave]

The slower melting over the last few weeks caused the model to come in a little higher at 4.68 million sq km for the September average extent.

 

 

[forkyfork]

it's not a recovery unless we get back to 1980s levels 

[bluewave]

While the rate of extent loss has slowed since 2007-2012, the multiyear ice levels continue to run near all-time lows. So the 2007-2012 era shifted the ice to a much thinner state. It’s  easier for the winds to push the mostly first year ice  around and cause big yearly extent variations like 2012-2013 and 2020-2021. 
 

https://www.jpl.nasa.gov/news/with-thick-ice-gone-arctic-sea-ice-changes-more-slowly
The Arctic Ocean's sea ice blanket has already lost most of its old ice and two-thirds of its thickness. The younger ice is thinning more slowly and variably.

The Arctic Ocean's blanket of sea ice has changed since 1958 from predominantly older, thicker ice to mostly younger, thinner ice, according to new research published by NASA scientist Ron Kwok of the Jet Propulsion Laboratory, Pasadena, California. With so little thick, old ice left, the rate of decrease in ice thickness has slowed. New ice grows faster but is more vulnerable to weather and wind, so ice thickness is now more variable, rather than dominated by the effect of global warming.

 

Working from a combination of satellite records and declassified submarine sonar data, NASA scientists have constructed a 60-year record of Arctic sea ice thickness. Right now, Arctic sea ice is the youngest and thinnest its been since we started keeping records. More than 70 percent of Arctic sea ice is now seasonal, which means it grows in the winter and melts in the summer, but doesn't last from year to year. This seasonal ice melts faster and breaks up easier, making it much more susceptible to wind and atmospheric conditions.

Kwok's research, published today in the journal Environmental Research Letters, combined decades of declassified U.S. Navy submarine measurements with more recent data from four satellites to create the 60-year record of changes in Arctic sea ice thickness. He found that since 1958, Arctic ice cover has lost about two-thirds of its thickness, as averaged across the Arctic at the end of summer. Older ice has shrunk in area by almost 800,000 square miles (more than 2 million square kilometers). Today, 70 percent of the ice cover consists of ice that forms and melts within a single year, which scientists call seasonal ice.

Sea ice of any age is frozen ocean water. However, as sea ice survives through several melt seasons, its characteristics change. Multiyear ice is thicker, stronger and rougher than seasonal ice. It is much less salty than seasonal ice; Arctic explorers used it as drinking water. Satellite sensors observe enough of these differences that scientists can use spaceborne data to distinguish between the two types of ice.

Thinner, weaker seasonal ice is innately more vulnerable to weather than thick, multiyear ice. It can be pushed around more easily by wind, as happened in the summer of 2013. During that time, prevailing winds piled up the ice cover against coastlines, which made the ice cover thicker for months.

The ice's vulnerability may also be demonstrated by the increased variation in Arctic sea ice thickness and extent from year to year over the last decade. In the past, sea ice rarely melted in the Arctic Ocean. Each year, some multiyear ice flowed out of the ocean into the East Greenland Sea and melted there, and some ice grew thick enough to survive the melt season and become multiyear ice. As air temperatures in the polar regions have warmed in recent decades, however, large amounts of multiyear ice now melt within the Arctic Ocean itself. Far less seasonal ice now thickens enough over the winter to survive the summer. As a result, not only is there less ice overall, but the proportions of multiyear ice to seasonal ice have also changed in favor of the young ice.

Seasonal ice now grows to a depth of about six feet (two meters) in winter, and most of it melts in summer. That basic pattern is likely to continue, Kwok said. "The thickness and coverage in the Arctic are now dominated by the growth, melting and deformation of seasonal ice."

The increase in seasonal ice also means record-breaking changes in ice cover such as those of the 1990s and 2000s are likely to be less common, Kwok noted. In fact, there has not been a new record sea ice minimum since 2012, despite years of warm weather in the Arctic. "We've lost so much of the thick ice that changes in thickness are going to be slower due to the different behavior of this ice type," Kwok said.

Kwok used data from U.S. Navy submarine sonars from 1958 to 2000; satellite altimeters on NASA's ICESat and the European CryoSat-2, which span from 2003 to 2018; and scatterometer measurements from NASA's QuikSCAT and the European ASCAT from 1999 to 2017.

[so_whats_happening]

I know this tends to not be everyone's favorite but this is just abysmal. Mind you we are only about mid July right now too. Still have a solid month of decent melt that can occur. GFS has been trying to push the idea of another ridging episode in the last week or so of July too which may just be the final push needed to melt out Beaufort, ESS, and parts of CAA. Somehow the Atlantic front is still sustaining itself even though it was the weakest portion of the ice starting into the melt season.

[so_whats_happening]

Is it possible that with extent numbers still relatively higher than the descending average and choppy ice that we see a quicker cooldown scenario going into fall allowing for a potential of thicker ice as we move into the winter? I assume this would not allow for heat to escape in anyway shape or form from the ocean though.

[ORH_wxman]

Update:

As of 7/12, NSIDC area was 6.26 million sq km. The cold cyclone over the CAB right now continues to slow the area loss. 
 

Here are where other years compared on the same date:

2021: -730k

2020: -750k

2019: -930k

2018: -190k

2017: -70k

2016: -620k

2015: -210k

2014: +30k

2013: -120k

2012: -940k

2011: -460k

2010: -330k

2009: +230k

2008: +40k

2007: -490k

[GaWx]

5 hours ago, ORH_wxman said:

Update:

As of 7/12, NSIDC area was 6.26 million sq km. The cold cyclone over the CAB right now continues to slow the area loss. 
 

Here are where other years compared on the same date:

2021: -730k

2020: -750k

2019: -930k

2018: -190k

2017: -70k

2016: -620k

2015: -210k

2014: +30k

2013: -120k

2012: -940k

2011: -460k

2010: -330k

2009: +230k

2008: +40k

2007: -490k

 Based on these numbers, the last two days have continued the significant increase based on the date of 2022 area vs the 2007-21 average area that started ~6/29. As of 7/12, the 2022 area was +370K vs the 2007-21 average. This is a 510K increase vs the 2007-21 average just since June 29th (see table below).
 
 Based on the graph at the bottom of this post of the minimum area for 2007-21, I roughly estimate an average minimum area for 2007-21 of 3.9 msk. The range in msk is roughly 2.9 to 4.6. 
 

 If the +370K vs 2007-21 average were to still be the case when the 2022 minimum is hit, that would mean a 2022 minimum area of ~4.3 msk. If that were to occur, that would mean a minimum higher than all years 2007-21 except for 2009, 2013, and 2014. However, having a +370K vs 2007-21 at the minimum of the year is very optimistic, especially considering the near record low multiyear ice levels as of last September that @bluewaveposted about yesterday. So, average ice thickness was then at near record lows. Thus, I think that the chance of this year's minimum being significantly lower than 4.3 msk is high. 
 

 But it does appear that the chance of it being at or above the 2007-21 average minimum of 3.9 msk has increased quite a bit vs how it looked on June 29th. Along with this, it appears that the chance for a September SIE at or above 4.8 msk has increased. That compares to the rough average Sept. extent of 4.7 msk for 2007-21 based on the Sept. SIE graph that @bluewaveshowed Monday.

 

Recap of current area vs avg. of last 15 years as of:

6/15/22: +70K

6/16/22: +120K

6/20/22: +240K

6/22/22: +140K

6/29/22: -140K

7/1/22:     +70K

7/10/22:  +250K

7/12/22:  +370K

 

Graph of Arctic annual ice area minimum since late 1970s from https://svs.gsfc.nasa.gov/5002

[ORH_wxman]

4 hours ago, GaWx said:

 Based on these numbers, the last two days have continued the significant increase based on the date of 2022 area vs the 2007-21 average area that started ~6/29. As of 7/12, the 2022 area was +370K vs the 2007-21 average. This is a 510K increase vs the 2007-21 average just since June 29th (see table below).
 
 Based on the graph at the bottom of this post of the minimum area for 2007-21, I roughly estimate an average minimum area for 2007-21 of 3.9 msk. The range in msk is roughly 2.9 to 4.6. 
 

 If the +370K vs 2007-21 average were to still be the case when the 2022 minimum is hit, that would mean a 2022 minimum area of ~4.3 msk. If that were to occur, that would mean a minimum higher than all years 2007-21 except for 2009, 2013, and 2014. However, having a +370K vs 2007-21 at the minimum of the year is very optimistic, especially considering the near record low multiyear ice levels as of last September that @bluewaveposted about yesterday. So, average ice thickness was then at near record lows. Thus, I think that the chance of this year's minimum being significantly lower than 4.3 msk is high. 
 

 But it does appear that the chance of it being at or above the 2007-21 average minimum of 3.9 msk has increased quite a bit vs how it looked on June 29th. Along with this, it appears that the chance for a September SIE at or above 4.8 msk has increased. That compares to the rough average Sept. extent of 4.7 msk for 2007-21 based on the Sept. SIE graph that @bluewaveshowed Monday.

 

Recap of current area vs avg. of last 15 years as of:

6/15/22: +70K

6/16/22: +120K

6/20/22: +240K

6/22/22: +140K

6/29/22: -140K

7/1/22:     +70K

7/10/22:  +250K

7/12/22:  +370K

 

Graph of Arctic annual ice area minimum since late 1970s from https://svs.gsfc.nasa.gov/5002

Bluewave’s numbers were extent while I’m posting area. Area is a little different. 

[GaWx]

1 hour ago, ORH_wxman said:

Bluewave’s numbers were extent while I’m posting area. Area is a little different. 

 Yeah, I realized that when I posted though it certainly can get confusing. 

 I said: "But it does appear that the chance of it being at or above the 2007-21 average minimum of 3.9 msk has increased quite a bit vs how it looked on June 29th. Along with this, it appears that the chance for a September SIE at or above 4.8 msk has increased. That compares to the rough average Sept. extent of 4.7 msk for 2007-21 based on the Sept. SIE graph that @bluewaveshowed Monday."

 So, I first referred to an increased chance for the minimum area to be 3.9+ msk (at or above the 2007-21 average minimum area). Then I referred to an increased chance for the September extent (SIE) to be 4.8+ msk (near or above the 2007-21 average minimum extent).

[ORH_wxman]

49 minutes ago, GaWx said:

 Yeah, I realized that when I posted though it certainly can get confusing. 

 I said: "But it does appear that the chance of it being at or above the 2007-21 average minimum of 3.9 msk has increased quite a bit vs how it looked on June 29th. Along with this, it appears that the chance for a September SIE at or above 4.8 msk has increased. That compares to the rough average Sept. extent of 4.7 msk for 2007-21 based on the Sept. SIE graph that @bluewaveshowed Monday."

 So, I first referred to an increased chance for the minimum area to be 3.9+ msk (at or above the 2007-21 average minimum area). Then I referred to an increased chance for the September extent (SIE) to be 4.8+ msk (near or above the 2007-21 average minimum extent).

Where are you getting the area numbers from? There hasn’t been a minimum above 4 million sqkm since 2006. 

[GaWx]

7 hours ago, ORH_wxman said:

Where are you getting the area numbers from? There hasn’t been a minimum above 4 million sqkm since 2006. 

 Hmmm, I must have done something wrong considering how knowledgeable and experienced you are on this, but I can't figure it out yet after looking at it again carefully. Am I doing an apples to oranges comparison? I thought you were looking at "Arctic sea ice area" in sq km. Is the source I used not looking at the same thing? Does it use a different method to measure area? Fwiw, its graph (see below) shows 2009, 2010, 2013 (highest since 2007 with 4.6 msk), 2014, and 2021 all above 4 million square km. It even says this about 2021: "In 2021, the Arctic minimum sea ice covered an area of 4.13 million square kilometers (1.6 million square miles)."

 Would you please provide a link to your source for annual Arctic minimum area as well as how current sea ice area compares to prior years for the same date?

 What do you have for 2012's minimum? This has ~2.9 million sq km:

https://svs.gsfc.nasa.gov/5002