2015 sea ice area and extent data

[DavidR]

Well the June figures for NSIDC extent are in, and at just 1.864 M km^2  its the smallest drop since 2004 which fell just over 1.2 M km^2.  At the time 2004 was the 9th lowest on record but by the end of the season it had dropped to 3rd lowest on record.

We didn't even get near the statistical 260 K anomalous drop that we normally see on July 1st.  The only  recent time we haven't seem this drop on July 1st it showed up on July 2nd so it will be interesting to see what  tomorrow brings.

All in all still an interesting season where anything can happen if the el Nino starts to  bite.

[Wipneus]

From yesterday's NSIDC data update I calculate the following CT updates (the days in this list are the days of data released, three days behind ice dates):

Thu       7.419765
Fri -138.9  7.280890
Sat -177.8  7.103101


That is including the "first of the month" effect. This effect is much less than extent but still perhepa -40k. Even without that effect, there is still a big drop within the Basin. It was visible on ADS-NIPR Jaxa's thickness/melting map as well, so the phrase "melting" does not seem out of place.
Biggest declines within the basin: ESS (-44k 1), Laptev (-29k7), Beaufort (-21k4) Chukchi (-13k7). Outside the basin there is Baffin with -23k8, but this region will be affected by the 1st of the month effect.

In the attached NSIDC delta map, pixels with larger concentration changes than 7% or colored pinkish (down) or light blue (up). Solid red and blue are where the pixel concentration crosses the 15% limit (for extent).

[Nightvid Cole]

Well the June figures for NSIDC extent are in, and at just 1.864 M km^2  its the smallest drop since 2004 which fell just over 1.2 M km^2.  At the time 2004 was the 9th lowest on record but by the end of the season it had dropped to 3rd lowest on record.

We didn't even get near the statistical 260 K anomalous drop that we normally see on July 1st.  The only  recent time we haven't seem this drop on July 1st it showed up on July 2nd so it will be interesting to see what  tomorrow brings.

All in all still an interesting season where anything can happen if the el Nino starts to  bite.

El nino is probably totally irrelevant. In the satellite record, el ninos beginning in summer of a given year appear to have no discernible impact on September extent. The two strongest ones began in 1982 and 1997 - and the September extent even in those extreme years was well within the range of the years just before and just after it. I should also add that one of those years was above the linear trendline and the other was below it, but not significantly so.

I think it is safe to conclude that even a monster El nino starting this summer should have no detectable effect on September 2015 ice extent. It is safe to ignore it.

[ChrisReynolds]

Well the June figures for NSIDC extent are in, and at just 1.864 M km^2  its the smallest drop since 2004 which fell just over 1.2 M km^2.  At the time 2004 was the 9th lowest on record but by the end of the season it had dropped to 3rd lowest on record.

We didn't even get near the statistical 260 K anomalous drop that we normally see on July 1st.  The only  recent time we haven't seem this drop on July 1st it showed up on July 2nd so it will be interesting to see what  tomorrow brings.

All in all still an interesting season where anything can happen if the el Nino starts to  bite.

El nino is probably totally irrelevant. In the satellite record, el ninos beginning in summer of a given year appear to have no discernible impact on September extent. The two strongest ones began in 1982 and 1997 - and the September extent even in those extreme years was well within the range of the years just before and just after it. I should also add that one of those years was above the linear trendline and the other was below it, but not significantly so.

I think it is safe to conclude that even a monster El nino starting this summer should have no detectable effect on September 2015 ice extent. It is safe to ignore it.

It is rather a nice feeling to agree with you Nightvid, we seem to disagree on key issues rather a lot...

Take Curry 2004.
http://onlinelibrary.wiley.com/doi/10.1029/2004GL019858/full
ENSO has an effect, but it is not dominant.

Take NINO 3.4
http://www.esrl.noaa.gov/psd/gcos_wgsp/Timeseries/Nino34/

Take NSIDC September extent.
ftp://sidads.colorado.edu/DATASETS/NOAA/G02135/Sep/N_09_area.txt

NSIDC extent has a strong trend, so I detrend using interannual difference of the September extent. This transforms the question to 'does June to August average NINO 3.4 have an relationship to the interannual difference of NSIDC September extent'. Doing a scatter plot and using a linear fit, the slope is -0.047, the R2 is 0.0015 (An R2 of 0 means no relationship, an R2 of 1 means a 1:1 correlation. So the slope is very slight, and the relationship is non existent.

If people think this isn't the right test they have the data, in the links above, but using such tests I have been unable to find any evidence of a relationship.

Tray another angle: Take NCEP/NCAR correlations from here:
http://www.esrl.noaa.gov/psd/data/correlation/

NINO 3.4 is positive recently
https://www.ncdc.noaa.gov/teleconnections/enso/indicators/sst.php
Only got up to May right now, but I presume the +ve phase has continued.

AO is positive recently.
http://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/ao_index.html

What does regression tell us?

See the attachments. Regression of ENSO is much smaller than for AO. AO suggets a negative regression over the Arctic Ocean (+ve AO index -> low pressure), ENSO suggests a positive regression over the Arctic Ocean (+ve NINO 3.4 -> high pressure).

Thus the NINO 3.4 index does not account for June's low pressure dominance over the Arctic Ocean.

[wili]

" el ninos beginning in summer of a given year appear to have no discernible impact on September extent."

That's true, but hasn't this El Nino been building for a lot longer than this summer?

[Tor Bejnar]

I think I remember reading somewhere that a strong El Niño during the northern hemisphere (NH) autumn or winter affects the following year's Arctic sea ice.  (Sometimes a La Niña is in place during the affected melting season.)  According to NOAA, we had a week El Niño last NH autumn/winter.  The effects of this now moderate (forecast by many to become strong) El Niño, if I remember correctly, wouldn't be felt until 2016.

[ChrisReynolds]

" el ninos beginning in summer of a given year appear to have no discernible impact on September extent."

That's true, but hasn't this El Nino been building for a lot longer than this summer?

You've got links to data, feel free to demonstrate it.

[wili]

I'm not sure there would be enough data to produce anything robust. How many week El Nino's have there been in August over the years?

Do you know if there's any support for Tor's claim?

ETA: This is what it says in the NSIDC FAQ page:

Lastly, the El Nino and La Nina cycle also appear to influence sea ice in the Pacific sector. El Nino patterns (a warm eastern tropical Pacific) are associated with warmer winds and less ice; the opposite is true for La Nina.

That is just talking about the 'Pacific sector' rather than the whole ice pack, and it doesn't say when in the year the El Nino occurs when it has even this limited effect. And it doesn't, unfortunately, link to any supporting study. More to search for...

http://nsidc.org/arcticseaicenews/faq/

[LRC1962]

I do think there will be problems with any past studies of the enso effects on the Arctic as this will be the first time a change of El Nino/La Nina has occurred since the Terrible Troughs/Ridged Ridges has become a major influence in the North American weather pattern cycle.
Although things can change in a hurry, Southern Ontario has still seen generally cooler than normal temps given a warming world since January. This is indicative of troughs much more prevalent than ridges over that time span.

[Tor Bejnar]

What I remember having read might be this 2010 ASI blog post: Arctic sea ice melt: a correlation with ENSO?

[ChrisReynolds]

Tor,

Unfortunately things didn't pan out as suggested in that post. Wayne is reported as saying "If El-Nino persists till the spring, and La-Nina follows, ships at the Pole will wander unobstructed in August 2010." What happened was that the El Nino persisted through until the summer, with a La Nina in the autumn.
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml
So we will never know whether Wayne's anecdotal prediction was sound or not.

The major problem with that post though is that despite being entitled 'A correlation' there is a singular lack of mathematical correlation analysis using past data.

Wili,

I'm not sure there would be enough data to produce anything robust. How many week El Nino's have there been in August over the years?

Then ignore the neutral/weak states! Order in terms of most postive and most negative and see how much of the expected order of detrended September extent carries through that ordering process.

Or you could just look at Curry et al - They conclude that the ENSO and AO impact the ice 'greatly'. Yet with respect to the ENSO the impact found by Curry is really rather limited, the detail is the paragraph under figure 3.

Figure 4a shows the situation in Bering Sea and Alaska, with northward flow in El Nino years towards Beaufort & Chukchi. One could argue that this plays a role in the current Chukchi decline. True, extent in Chukchi is tracking close to 2007 and 2011. But in Beaufort extent continues to fall, after initial losses decline has arrested as the ice edge has hit the boundary of MYI (see HYCOM). And it is hard to seperate the impacts of the 10 June storm from the possible impact of El Nino in Chukchi. I should add that (from part of my upcoming June blog posts - the first half of June does show the expected northwards flow from the Pacific, but the second half does not).

[DavidR]

" el ninos beginning in summer of a given year appear to have no discernible impact on September extent."

That's true, but hasn't this El Nino been building for a lot longer than this summer?

You've got links to data, feel free to demonstrate it.

There are a number of indicators that the impact  of this El Nino may  be different  than previous ones.

1:  This year  the SOI has been reasonably  negative since August last  year:
http://www.bom.gov.au/climate/current/soihtm1.shtml
compared to  1982 and 1997  where it  was neutral or positive until the northern spring. 

2: the 1982 El Nino was impacted by the El Chichon volcanic eruption in April 1982 which would have reduced global temperature increases, however a record global temperature, and Sea Ice extent minimum occurred the following year.
https://en.wikipedia.org/wiki/El_Chichon

3. The Pacific Decadal Oscillation Index sat  above +2 for four months from December to  March this year:
http://research.jisao.washington.edu/pdo/PDO.latest
The last time this happened was June - Sept  1997; and every time it  has happened we have seen global temperature records in that year or the subsequent year.

So, given that all the precursors of significant warming were in  place from late last year, I  believe there is the possibility of significant impact this year.

[Wipneus]

From yesterday's NSIDC data update I calculate the following CT updates (the days in this list are the days of data released, three days behind ice dates):

Fri       7.282727
Sat -177.7  7.105018
Sun -11.7  7.093290

Bad melt day (you know what I mean). About the only region with a decent decline is Laptev (-29k1). A few regions with small (~+10k) increases: CAB, ESS, Hudson and "Lakes".

In the attached NSIDC delta map, pixels with larger concentration changes than 7% or colored pinkish (down) or light blue (up). Solid red and blue are where the pixel concentration crosses the 15% limit (for extent).

[Tor Bejnar]

Thanks, Chris, for putting my remembered reading into perspective. 
Now, how do I erase that memory I've been holding on to for 4 years?! (I long for the interrobang!)

[ChrisReynolds]

Thanks, Chris, for putting my remembered reading into perspective. 
Now, how do I erase that memory I've been holding on to for 4 years?! (I long for the interrobang!)

Often when I answer even the simplest of questions it takes ages of digging to make sure I remembered it right.

David R,

Let's leave the thread to its proper purpose now. If you want to make a testable prediction based on your reasoning (sorry I don't find it at all persuasive) then you can evaluate it in September.

[Jim Pettit]

ADS-NIPR Extent:

9,341,259 km2 (03 July)

Down 4,600,801 km2 (33.%) from 2015 maximum of 13,942,060 km2 on 15 February.

6,163,804 km2 above record minimum extent of 3,177,455 km2 (16 September 2012).

Down 46,395 km2 from previous day.

Down 476,460 km2 over past seven days (daily average: -68,066 km2).

Down 141,057 km2 for the month of July (daily average: -47,019 km2).

331,325 km2 below 2000s average for this date.

420,123 km2 above 2010s average for this date.

563,453 km2 above 2014 value for this date.

595,324 km2 above 2012 value for this date.

8th lowest July to-date average.

8th lowest value for the date.

64 days this year (34.78% year-to-date) have recorded the lowest daily extent.

32 days (17.39%) have recorded the second lowest.

32 days (17.39%) have recorded the third lowest.

128 days (69.57%) in total have been among the three lowest on record.

CT Area:

7,103,773 km2 (03 July [Day 0.5013])

Down 6,170,782 km2 (46.49%) from 2015 maximum of 13,274,555 km2 on 17 February [Day 0.1288].

4,869,764 km2 above record minimum area of 2,234,010 km2 (14 September 2012).

Down 178,954 km2 from previous day.

Down 729,672 km2 over past seven days (daily average: -104,239 km2).

Down 3,032,807 km2 for the month of July (daily average: -1,010,936 km2).

317,648 km2 below 2000s average for this date.

228,125 km2 above 2010s average for this date.

95,877 km2 below 2014 value for this date.

789,966 km2 above 2012 value for this date.

6th lowest July to-date average.

5th lowest value for the date.

7 days this year (3.8% year-to-date) have recorded the lowest daily area.

19 days (10.33%) have recorded the second lowest.

29 days (15.76%) have recorded the third lowest.

55 days in total (29.89%) have been among the lowest three on record.

[Wipneus]

From yesterday's NSIDC data update I calculate the following CT updates (the days in this list are the days of data released, three days behind ice dates):

Sat       7.103773
Sun -11.7  7.092045
Mon -212.6  6.879408

Strong melt day follows a weak one reported on Sunday. Biggest decline in the CAB (-92k4), then Hudson (-61k9), ESS (-29k) and Laptev (-25k5).

In the attached NSIDC delta map, pixels with larger concentration changes than 7% or colored pinkish (down) or light blue (up). Solid red and blue are where the pixel concentration crosses the 15% limit (for extent).

[oren]

ADS-NIPR Extent:

9,341,259 km2 (03 July)

Down 4,600,801 km2 (33.%) from 2015 maximum of 13,942,060 km2 on 15 February.

6,163,804 km2 above record minimum extent of 3,177,455 km2 (16 September 2012).

Down 46,395 km2 from previous day.

Down 476,460 km2 over past seven days (daily average: -68,066 km2).

Down 141,057 km2 for the month of July (daily average: -47,019 km2).

331,325 km2 below 2000s average for this date.

420,123 km2 above 2010s average for this date.

563,453 km2 above 2014 value for this date.

595,324 km2 above 2012 value for this date.

8th lowest July to-date average.

8th lowest value for the date.

64 days this year (34.78% year-to-date) have recorded the lowest daily extent.

32 days (17.39%) have recorded the second lowest.

32 days (17.39%) have recorded the third lowest.

128 days (69.57%) in total have been among the three lowest on record.

CT Area:

7,103,773 km2 (03 July [Day 0.5013])

Down 6,170,782 km2 (46.49%) from 2015 maximum of 13,274,555 km2 on 17 February [Day 0.1288].

4,869,764 km2 above record minimum area of 2,234,010 km2 (14 September 2012).

Down 178,954 km2 from previous day.

Down 729,672 km2 over past seven days (daily average: -104,239 km2).

Down 3,032,807 km2 for the month of July (daily average: -1,010,936 km2).

317,648 km2 below 2000s average for this date.

228,125 km2 above 2010s average for this date.

95,877 km2 below 2014 value for this date.

789,966 km2 above 2012 value for this date.

6th lowest July to-date average.

5th lowest value for the date.

7 days this year (3.8% year-to-date) have recorded the lowest daily area.

19 days (10.33%) have recorded the second lowest.

29 days (15.76%) have recorded the third lowest.

55 days in total (29.89%) have been among the lowest three on record.

There's an error in the area numbers July to date.

[Neven]

From yesterday's NSIDC data update I calculate the following CT updates (the days in this list are the days of data released, three days behind ice dates):

Sat       7.103773
Sun -11.7  7.092045
Mon -212.6  6.879408

Well well, a double century. That's going to make the CAPIE/compactness numbers even more interesting than they already are.

[Buddy]

Two months of interesting weather left....and July should prove to be the most interesting of all...

"The ice that remains, has been ravaged through the years,
As politicians spout convincing lies about how she is in fine shape,
But they cannot save the final crystals of the Arctic with their deception,
The truth lay in wait....for the final crystals to pass....and aye, they WILL pass."

"author unknown"

[Jim Pettit]

There's an error in the area numbers July to date.

Yes, thanks.

If you spot such errors in the future, please send a message directly to me. That will ensure the issue comes to my attention more quickly--and it will also ensure the error doesn't get copied and thus perpetuated in places I can't edit.

Thanks...

[Nightvid Cole]

" el ninos beginning in summer of a given year appear to have no discernible impact on September extent."

That's true, but hasn't this El Nino been building for a lot longer than this summer?

You've got links to data, feel free to demonstrate it.

There are a number of indicators that the impact  of this El Nino may  be different  than previous ones.

1:  This year  the SOI has been reasonably  negative since August last  year:
http://www.bom.gov.au/climate/current/soihtm1.shtml
compared to  1982 and 1997  where it  was neutral or positive until the northern spring. 

2: the 1982 El Nino was impacted by the El Chichon volcanic eruption in April 1982 which would have reduced global temperature increases, however a record global temperature, and Sea Ice extent minimum occurred the following year.
https://en.wikipedia.org/wiki/El_Chichon

3. The Pacific Decadal Oscillation Index sat  above +2 for four months from December to  March this year:
http://research.jisao.washington.edu/pdo/PDO.latest
The last time this happened was June - Sept  1997; and every time it  has happened we have seen global temperature records in that year or the subsequent year.

So, given that all the precursors of significant warming were in  place from late last year, I  believe there is the possibility of significant impact this year.

Arctic sea ice is only impacted by local forcing, not global temperatures. This forcing can include temperature, dew point, shortwave and longwave radiation (upwelling and downwelling), wind, sea surface temperature and salinity, waves, and other factors.

At the level of changes in climate, which take place over multidecadal and longer time scales, local temperatures and global temperatures are very tightly correlated, for both natural climate changes observable from paleoclimate data, and also from human-induced climate change which is easiest to observe in the records of the most recent decades (i.e. the 1980s and later).

However, when it comes to weather changes that cause year-to-year fluctuations in Arctic ice extent, area, and volume, there is no reason to presume a correlation with year-to-year global weather fluctuations.

 

[BornFromTheVoid]

Update for the week to July 4th

The current 5 day mean is on 9,860,000km2 while the 1 day extent is at 9,726,000km2.
The daily anomaly (compared to 81-10) is at -788,180km2, a decrease from -831,290km2 last week. The anomaly compared to the 07, 11 and 12 average is at +618,200km2, an increase from +299,000km2 last week. We're currently 8th lowest on record, down from 7th last week.



The average daily change over the last 7 days was -79.0k/day, compared to the long term average of -85.2k/day, and the 07, 11 and 12 average of -124.6k/day.
The average long term change over the next week is -80.5k/day, with the 07, 11, and 12 average being -115.8k/day.



The loss so far this July is the 8th smallest on record. To achieve the largest monthly drop, a daily loss of at least 121.2k/day is required, while the smallest drop requires a loss of less than 65.4k/day and an average drop requires a loss of 89.7k/day.



The loss in June was the 18th smallest on record, while the average extent was the 4th smallest on record.

[Wipneus]

From yesterday's NSIDC data update I calculate the following CT updates (the days in this list are the days of data released, three days behind ice dates):

Sun       7.091670
Mon -212.6  6.879042
Tue -102.6  6.776474

Tuesday's century is supported by the following regions: CAB (-25k7), Greenland Sea (-20k3), ESS (-18k7), Laptev (-16k1) and Chukchi (-14k4). Opposition is by Beaufort (+23k3).

In the attached NSIDC delta map, pixels with larger concentration changes than 7% or colored pinkish (down) or light blue (up). Solid red and blue are where the pixel concentration crosses the 15% limit (for extent).

[DavidR]

Arctic sea ice is only impacted by local forcing, not global temperatures. This forcing can include temperature, dew point, shortwave and longwave radiation (upwelling and downwelling), wind, sea surface temperature and salinity, waves, and other factors.

NightVid,
My point was that there is insufficient evidence on the impact of major EL Ninos to rule an impact in or out this year.  The circumstances of the only two we have seen since 1979 were quite different from each other, and from this year, so any inferences of an impact or otherwise is purely  speculative.

SST and Air temperatures across the Arctic rebounded in June from the anomalously  low temperatures in May and if this continues for the next couple months we should return to a more dramatic melt than we have seen in the last month. An EL Nino could assist in raising these temperatures.

[LRC1962]

Arctic sea ice is only impacted by local forcing, not global temperatures. This forcing can include temperature, dew point, shortwave and longwave radiation (upwelling and downwelling), wind, sea surface temperature and salinity, waves, and other factors.

At the level of changes in climate, which take place over multidecadal and longer time scales, local temperatures and global temperatures are very tightly correlated, for both natural climate changes observable from paleoclimate data, and also from human-induced climate change which is easiest to observe in the records of the most recent decades (i.e. the 1980s and later).

However, when it comes to weather changes that cause year-to-year fluctuations in Arctic ice extent, area, and volume, there is no reason to presume a correlation with year-to-year global weather fluctuations.

I have no clue where you get that idea from. Example: Jan - Mar there were a lot of cyclones that hit the Atlantic side of the Arctic. They did not start in the Arctic they got their start from the equatorial region of the Atlantic. To say they had no influence on the area/extent of the ice is nuts. To also say those storms and the frequency of them were not influenced by global temps is also nuts.
The Arctic is not an isolated insulated environment not affected by what is going on on a daily global scale is absolutely unbelievable otherwise we could weather forecast out to 2-3 months very accurately instead of 4-5 days.
Ice area and extent IMO are directly related to what happens on a global scale. The more local it becomes the greater the influence, but if the influence is big enough the effect on the ice even if the event is in Australia will change the ice in the Arctic.

[plinius]

I'd say a lot of your discrepancies appears to come from neglect of the non-linearities/feedbacks in the system. It is certainly true that most of the forcing/melt energy is coming from insolation and not atmospheric or oceanic heat advection. However, that's a bit of a short leap, isn't it? If you do not remove the snow cover first, albedo over the pack is so large that you keep your equilibrium temperature below 0 over most of the pack, so there will never be any serious melt if you do not initiate it. I think that's also one of the main reasons why ice climate models have had so many failures in correct predictions. Just don't cover the feedbacks on the ice properly.

[Jim Hunt]

Jan - Mar there were a lot of cyclones that hit the Atlantic side of the Arctic. They did not start in the Arctic they got their start from the equatorial region of the Atlantic. To say they had no influence on the area/extent of the ice is nuts.

I even made a video of all that pushing and shoving:

[Wipneus]

From yesterday's NSIDC data update I calculate the following CT updates (the days in this list are the days of data released, three days behind ice dates):

Mon       6.878442
Tue -102.5  6.775924
Wed +39.2  6.815120

The largest increase since May 13, thanks to the CAB (+45k2) , Hudson (+27k2) and many regions with smaller increases. Declines in ESS (-29k3) and Chukchi (-25k1).

In the attached NSIDC delta map, pixels with larger concentration changes than 7% or colored pinkish (down) or light blue (up). Solid red and blue are where the pixel concentration crosses the 15% limit (for extent).

[ghoti]

Oh look! Toronto harbour is icing up again in time for the Climate Summit of the Americas

I really have trouble trusting all the data when the Great Lakes data keeps showing them icing up and melting while the temperatures hit 30C in July.

Oh well we don't seem to have a better source.

[Nightvid Cole]

LRC1962 and DavidR,

There is a lag of several months between El Nino region SST's and global temps; also, there's no evidence of correlation of detrended global temps with detrended Arctic temps over a relevant period (say, 1980-present or 1981-2010).

Yes, some mixing does occur, but so what? That could lead to a positive correlation, or it could lead to a negative correlation (since warm air going into the Arctic would be expected to occur simultaneously with cold air flowing out of the Arctic and thus reducing temperatures elsewhere), or, it could lead to a zero correlation.

You simply don't have any justification at all for inferring from year-to-year fluctuations of El Nino, PDO, or global temps to temps in the Arctic, except on a multidecadal (climate) basis.

The mere existence of atmospheric mixing does nothing to change that.

[Wipneus]

Oh look! Toronto harbour is icing up again in time for the Climate Summit of the Americas

I really have trouble trusting all the data when the Great Lakes data keeps showing them icing up and melting while the temperatures hit 30C in July.

Oh well we don't seem to have a better source.

A repeat seems in order about lake ice.

CT-area and NSIDC extent are based on the same data, available as NSIDC gridded sea ice concentration data.

NSIDC does not include in its calculation ice detected in lakes, so there is no need to filter that says "No lake ice in July". In the ocean it _is_ filtered, so there will be "No ice in St. Lawrence" for example.

CT-area does include lake ice and also does not filter anything unless NSIDC has filtered it already.

This is only one of the deficiencies (my judgement) of the CT-area figures. But is a popular measure so I add some additional information about it. That is report it timely, summarize the regional deviations and note obvious "false" changes, like (dis)appearance of "lake ice".

Today's lake ice (and to be reported by CT on Wednesday) bumped +7k, which is under my radar.

[oren]

This melting season is wierd 

[LRC1962]

On the other hand is it at all possible to detrend what goes on in the Arctic with what goes on elsewhere? As I stated before and you have in no way denied, it is not an isolated environment.
You can not detrend the weather in England from what occurs in the Caribbean or what is happening in Greenland, they are all linked.
The extent in the Arctic is linked to the amount of snowfall on the ice, the SST of the Arctic, lows and highs, rainfall, wind direction, fires ..... all of which are governed far more by outside Arctic influences then by the Arctic itself especially in its current state. The problem science has is that the more chaotic things get the more trouble science has at defining what is important or how much each influence has importance. The less ASI you have the more chaotic everything becomes. The more chaotic thins become the less you can detrend anything because everything starts impacting everything else more and more.
Example: up to about 2000ish when you had a very high min extent, the jet stream was very fast smooth and very predictable. With that cam fairly predictable overall seasonal weather patterns. Now the jet stream wanders all over the place, seasonal weather patterns in any given geographical location have become more and more unusual, ocean currents are now starting to noticeably change etc.
BUT the greatest influences are no longer IMO based on multidecadal impacts, it is much more on the impacts of systems being sent into the Arctic that get their start far further away.
When a blocking high gets setup over Greenland systems get stuck over North America and systems in the Pacific then are tending to end up going into the Arctic. At the same time systems coming up the East coast of NA hit that high and either go west and push things north of Greenland into the Arctic or go east and because of a vacuum type effect end up going into the Arctic. These are not decadal systems these are just regular systems that always move around. The problem with that Greenland blocking high is that if the ice extent had been high it would have been located over the north pole. All this trying to say that you can no longer decouple the Arctic from the globe because chaos reigns supreme.

As Dr Barber and many others have stated. When they started studying the ASI is was boring because nothing happened, now things are changing so fast they have no clue what is going on and why.

[seaicesailor]

This melting season is wierd 

I hear you

[Nightvid Cole]

On the other hand is it at all possible to detrend what goes on in the Arctic with what goes on elsewhere? As I stated before and you have in no way denied, it is not an isolated environment.
You can not detrend the weather in England from what occurs in the Caribbean or what is happening in Greenland, they are all linked.
The extent in the Arctic is linked to the amount of snowfall on the ice, the SST of the Arctic, lows and highs, rainfall, wind direction, fires ..... all of which are governed far more by outside Arctic influences then by the Arctic itself especially in its current state. The problem science has is that the more chaotic things get the more trouble science has at defining what is important or how much each influence has importance. The less ASI you have the more chaotic everything becomes. The more chaotic thins become the less you can detrend anything because everything starts impacting everything else more and more.
Example: up to about 2000ish when you had a very high min extent, the jet stream was very fast smooth and very predictable. With that cam fairly predictable overall seasonal weather patterns. Now the jet stream wanders all over the place, seasonal weather patterns in any given geographical location have become more and more unusual, ocean currents are now starting to noticeably change etc.
BUT the greatest influences are no longer IMO based on multidecadal impacts, it is much more on the impacts of systems being sent into the Arctic that get their start far further away.
When a blocking high gets setup over Greenland systems get stuck over North America and systems in the Pacific then are tending to end up going into the Arctic. At the same time systems coming up the East coast of NA hit that high and either go west and push things north of Greenland into the Arctic or go east and because of a vacuum type effect end up going into the Arctic. These are not decadal systems these are just regular systems that always move around. The problem with that Greenland blocking high is that if the ice extent had been high it would have been located over the north pole. All this trying to say that you can no longer decouple the Arctic from the globe because chaos reigns supreme.

I'll show you one example. Here is a plot of July-August temperatures averaged over the tropics, year by year:

http://www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries.pl?ntype=1&var=Air+Temperature&level=2000&lat1=23.5&lat2=-23.5&lon1=0&lon2=360&iseas=1&mon1=6&mon2=7&iarea=1&typeout=2&Submit=Create+Timeseries

and here is the same for the Arctic:

http://www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries.pl?ntype=1&var=Air+Temperature&level=2000&lat1=90&lat2=66.5&lon1=0&lon2=360&iseas=1&mon1=6&mon2=7&iarea=1&typeout=2&Submit=Create+Timeseries

You can open them side by side in two browser windows and see that, apart from the overall warming trend, the individual warm/cool years in one series don't line up with those in the other series.

[jdallen]

From yesterday's NSIDC data update I calculate the following CT updates (the days in this list are the days of data released, three days behind ice dates):

Mon       6.878442
Tue -102.5  6.775924
Wed +39.2  6.815120

The largest increase since May 13, thanks to the CAB (+45k2) , Hudson (+27k2) and many regions with smaller increases. Declines in ESS (-29k3) and Chukchi (-25k1).

In the attached NSIDC delta map, pixels with larger concentration changes than 7% or colored pinkish (down) or light blue (up). Solid red and blue are where the pixel concentration crosses the 15% limit (for extent).

What's the granularity on CT again?  12.5KM?

It would be simply absurd to imagine that Wednesday increase is due to anything except for (1) sensors being fooled by surface conditions, (2) eccentricities of the algorithm or (3 - my favorite) ice breaking up at very small scales and spreading out.

(Edit: to my point - quick shot of detail around Wrangel island from yesterday.)

[Wipneus]

What's the granularity on CT again?  12.5KM?

Grid cell size (25x25km for CT and NSIDC) affects mostly extent. On area not so much and even less for the CT version (reason is the 0% cut-off applied in the algorithm, no ice is discarded when it just falls outside a particular grid cell).

[Rubikscube]

What's the granularity on CT again?  12.5KM?

It would be simply absurd to imagine that Wednesday increase is due to anything except for (1) sensors being fooled by surface conditions, (2) eccentricities of the algorithm or (3 - my favorite) ice breaking up at very small scales and spreading out.

(Edit: to my point - quick shot of detail around Wrangel island from yesterday.)

Considering the large drops during the last several days, it doesn't really seem that strange to have an uptick again, the maps Wipneus have been posting show that small upticks in concentration are quite usual, even on days with massive overall drops. It would be much stranger if ESS or Chukchi was gaining area again when these are the regions which are currently being pounded with heat, and it would be vary strange if this uptick is followed up by more slow days during this week.

[Greenbelt]

I wonder if strong winds from the north are expanding the ice pack toward Franz Josef land and Novaya Zemlya.  This could maybe have a similar impact as Fram export, expanding the ice extent at first, but reducing it over longer period of time.

[Wipneus]

From yesterday's NSIDC data update I calculate the following CT updates (the days in this list are the days of data released, three days behind ice dates):

Tue       6.775785
Wed +39.4  6.815138
Thu -164.3  6.650875

Back to the big declines on Thursday. The big three are ESS (-33k3), Kara (-33k1) and CAB (-30k4), followed at some distance by Hudson (-20k8), Chukchi (-11k9) and CAA (-11k2).

In the attached NSIDC delta map, pixels with larger concentration changes than 7% or colored pinkish (down) or light blue (up). Solid red and blue are where the pixel concentration crosses the 15% limit (for extent).

[seaicesailor]

From yesterday's NSIDC data update I calculate the following CT updates (the days in this list are the days of data released, three days behind ice dates):

Tue       6.775785
Wed +39.4  6.815138
Thu -164.3  6.650875

Back to the big declines on Thursday. The big three are ESS (-33k3), Kara (-33k1) and CAB (-30k4), followed at some distance by Hudson (-20k8), Chukchi (-11k9) and CAA (-11k2).

In the attached NSIDC delta map, pixels with larger concentration changes than 7% or colored pinkish (down) or light blue (up). Solid red and blue are where the pixel concentration crosses the 15% limit (for extent).

Wipneus, allow me one question: the NSIDC map labeled with "July 6, 2015" reflects data of same time as the Uni-Bremen map labeled with "July 6, 2015" or is there some time offset? Thx.

[Wipneus]

Wipneus, allow me one question: the NSIDC map labeled with "July 6, 2015" reflects data of same time as the Uni-Bremen map labeled with "July 6, 2015" or is there some time offset? Thx.

They are both of July 6. I assume that is from 0h00 UTC to 23h59 UTC, although that information is seldom mentioned.

[seaicesailor]

Wipneus, allow me one question: the NSIDC map labeled with "July 6, 2015" reflects data of same time as the Uni-Bremen map labeled with "July 6, 2015" or is there some time offset? Thx.

They are both of July 6. I assume that is from 0h00 UTC to 23h59 UTC, although that information is seldom mentioned.

Thank you.

[Wipneus]

From yesterday's NSIDC data update I calculate the following CT updates (the days in this list are the days of data released, three days behind ice dates):

Wed       6.815259
Thu -164.0  6.651229
Fri -193.1  6.458162

The second century of how many? Actually without lake ice Friday (+11k8) would show a double century. Big area declines in Hudson (-55k2),  the CAB (-44k4), ESS (-43k6). Smaller declines (-13k - -18k) in Baffin, Beaufort, Kara and Chukchi. 

In the attached NSIDC delta map, pixels with larger concentration changes than 7% or colored pinkish (down) or light blue (up). Solid red and blue are where the pixel concentration crosses the 15% limit (for extent).

[F.Tnioli]

...
What's the granularity on CT again?  12.5KM?

It would be simply absurd to imagine that Wednesday increase is due to anything except for (1) sensors being fooled by surface conditions, (2) eccentricities of the algorithm or (3 - my favorite) ice breaking up at very small scales and spreading out.

(Edit: to my point - quick shot of detail around Wrangel island from yesterday.)

How about (4) - melt pond draining? Is my guess on it - of any much significance, you think?

[Peter Ellis]

Melt ponds come under (1) - surface conditions - surely?

[F.Tnioli]

Melt ponds come under (1) - surface conditions - surely?

Nope, not if draining is complete, at least. Imagine a melt pond on a large solid thick (1m+) plate of ice. Surface melt. Let's say it gets to some, say, 20cm depth at some point. Ok, sensors fooled, that's (1) alright - but fooled "back then", not "now". Imagine that during following melt - i.e. "now" - water "drills a hole" in the middle of the pond, right through the ice, possibly through some imperfections - this kind of thing we see all the time on top of Greenland ice sheet summer-time, nowadays. And then _whole_ pond - to very last gramm of water, - goes down through the hole and into the ocean. Then we still have that large solid not-so-thick-now, but still quite thick plate of ice, with a small hole in the middle which "collects" any further melt water immediatelly. End result: where was water surface before ("fooled sensors"), we now have ice surface, real physical change - increased ice area at this specific place.

Now how many of such plates are out there? Hundreds thousands square kilometers? I guess alot. And sometimes lots of them get "drained" during a single day, hence real thing, affecting albedo, melt mode, etc. Partial drains give partial effect, etc.

This way, it's "previous larger drops" which were a result of "surface conditions", - but not the recent +40k gain, which then would be a result of "disappearance of surface conditions" a.k.a. melt ponds. To me, that matters...

[Peter Ellis]

I think we're talking at cross purposes.  In your example, the apparent increase in area was not a real increase in area, precisely because the sensors were initially fooled by the presence of a melt pond.  That's clearly a subset of (1) in my view.

[LRC1962]

Interesting article about melt ponds. http://www.pri.org/stories/2014-12-11/scientist-unlocks-one-mysteries-arctic-ice-melt .

Here's how: The second bucket was full of fresh water, the kind you get when snow melts, not salty sea water. Fresh water freezes at a higher temperature than salt water, so it solidifies inside those tiny cracks in the ice when sea water won’t.

It's Polashenski’s eureka moment: That’s why the ponds stick around.

Once the snow is gone and the melt water from it is gone, you would be then left for the most part, based on the fact most ice in the Arctic now young ice, with salty melt which can not collect into ponds. Also puts a lie to the fact that Arctic ice is solid.