2016 sea ice area and extent data

[Rob Dekker]

That is true. The drop in "compactness" is mostly caused by the effects of liquid water: from dry ice/snow to wet ice/snow to melt ponds. NSIDC sea ice concentration, calculated by the NASA Team algorithm, is more sensitive than the other two methods used in that graph.

The effects of resolving leads and open water are there and similar, but not that big.

Thank you Wipneus. That makes total sense.
It explains two observations in the concentration graphs :

- NASA team algorithm on SSMIS drops far lower than Bootstrap AMSR2 in the graphs during the melting season, accounting for the higher sensitivity to liquid water on the ice.
- In 2016, with more ice fragmentation, ASI AMSR2 shows 2016 in the lead (lowest concentration), but since there was less melt-onset (due to these persistent lows in June) there was less wet snow, and fewer melting ponds, so NASA Team SSMIS for 2016 was in the middle of the pack.
- There is still a (relative) difference between ASI and Bootstrap on AMSR2 data. ASI showing 2016 in the lead for ice concentration, while Bootstrap AMSR2 showing 2016 at a more moderate ranking.

It may be interesting to run NASA Team's algorithm on AMSR2 to see if this explains the differences in ice concentration, and how much (if any) of the difference is caused by difference in resolution.

[edit] From these findings and the concentration maps that you publish, can we cautiously conclude that 2016 is in the lead when it comes to lead/polynia/open water next to ice, while when it comes to overall water on the ice (including wet snow and melting ponds) that it is in the middle of the pack, and not setting records ?

[Richard Rathbone]

From these findings and the concentration maps that you publish, can we cautiously conclude that 2016 is in the lead when it comes to lead/polynia/open water next to ice, while when it comes to overall water on the ice (including wet snow and melting ponds) that it is in the middle of the pack, and not setting records ?

I'm not sure about the leads, but that's how I get an idea of melt pond behavior prior to Schroeder's analysis coming out with the SIPN prediction.

[Neven]

Richard, I have some info on that in my latest blog post: 2016 Melting momentum, part 3

[Richard Rathbone]

Richard, I have some info on that in my latest blog post: 2016 Melting momentum, part 3

That looks like "stick" rather than "twist" to me but for the August poll I will be taking a lot of notice of what the Slater projection looks like then.

[bbr2314]

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:

day  CT-date       NH               SH                Global
Sun 2016.5178 -118.1  6.036550 +173.2 12.619207   +55.1 18.655757
Mon 2016.5205  -56.1  5.980457 +232.2 12.851426  +176.1 18.831883
Tue 2016.5233 -156.1  5.824326 +140.1 12.991531   -16.0 18.815857
Wed 2016.5260 -108.8  5.715556  +14.6 13.006105   -94.2 18.721661

The century is supported by the CAB (-58k) and Hudson (-31k).

Shadow NSIDC extent is 8.4181 dropping  -94.4k. Hudson is star: -76k. Chukchi increased +24k, Beaufort dropped -20k.

Attached delta map shows what is going on.

Would you be able to provide 2012's area benchmarks for 8/1 and 8/16? Just curious!

On current trajectory, I think sub-4MKM2 is very doable by 8/1. How far below determines whether we end up merely somewhat below 2012 or nearly ice-free under 1M KM2 @ minimum. If we are near 3.5M KM2 by 8/1, going to be in even bigger trouble.

[Rob Dekker]

On current trajectory, I think sub-4MKM2 is very doable by 8/1.

On current trajectory, I don't see that there is any way to get sub-4Mkm^2 by 8/1.
https://ads.nipr.ac.jp/vishop/vishop-extent.html?N

[Bill Fothergill]

On current trajectory, I think sub-4MKM2 is very doable by 8/1.

On current trajectory, I don't see that there is any way to get sub-4Mkm^2 by 8/1.
https://ads.nipr.ac.jp/vishop/vishop-extent.html?N

Rob,
When bbr mentioned the possibility of a sub-4 by the beginning of August, I'm pretty sure this was referring to the shadow-CT value kindly provided by Wipneus, rather than the ADS extent value to which you linked . As the shadow-CT area now (as of Wed 13th July) stands at 5.716 million sq kms, it is certainly within the bounds of possibility for the drop rate over the next 18 days to average about 95k/day.

I'm not sticking my neck out and saying that this will happen, but I don't think it can be ruled out.

...
Would you be able to provide 2012's area benchmarks for 8/1 and 8/16? Just curious!
...

Links to the actual CT numbers (NH, SH & Global) are contained within the Sea Ice Graphs link on the ASIB.
https://sites.google.com/site/arcticseaicegraphs/

The values for the dates you requested were 3.787 & 2.942 respectively.

[bbr2314]

On current trajectory, I think sub-4MKM2 is very doable by 8/1.

On current trajectory, I don't see that there is any way to get sub-4Mkm^2 by 8/1.
https://ads.nipr.ac.jp/vishop/vishop-extent.html?N

Rob,
When bbr mentioned the possibility of a sub-4 by the beginning of August, I'm pretty sure this was referring to the shadow-CT value kindly provided by Wipneus, rather than the ADS extent value to which you linked . As the shadow-CT area now (as of Wed 13th July) stands at 5.716 million sq kms, it is certainly within the bounds of possibility for the drop rate over the next 18 days to average about 95k/day.

I'm not sticking my neck out and saying that this will happen, but I don't think it can be ruled out.

...
Would you be able to provide 2012's area benchmarks for 8/1 and 8/16? Just curious!
...

Links to the actual CT numbers (NH, SH & Global) are contained within the Sea Ice Graphs link on the ASIB.
https://sites.google.com/site/arcticseaicegraphs/

The values for the dates you requested were 3.787 & 2.942 respectively.

Thanks Bill!

I think 2016 will be about even or slightly lower than 2012 come the first with a much larger lead by the 16th. I will say 3.6MKM2 and 2.5MKM2 respectively.

[Bill Fothergill]

@ Rob,

I'm afraid the old grey matter went out for a stroll there. In order to substantiate what I had said in my above post, I meant (but failed lamentably in the intention) to include the following from Wipneus' running shadow-CT numbers...

On 2016.4986, the shadow-CT value was 6.946 million sq kms, but, by 2016.5260 this had dropped to 5.716 million sq kms. Averaged over that 10-day period, the resultant daily loss works out at about 123k/day.

Although you were correct in pointing out that the ADS drop rate is/was less impressive (about 89k/day over the same 10 day period) this could be another example of the two sets of figures showing apparently divergent behaviour as the concentration ratio alters.


@ bbr2314

I should have posted a slight "health warning" regarding date formats. Using the standard CT dating conventions, the date YYYY.5260 should actually relate to the 11th of July, not the 13th.

There is a trap for the unwary in that, for example, 2016.0000 means the 31st Dec 2015, and NOT the 1st of Jan 2016, as one might otherwise expect. (The reasons are historic, but primarily relate to the limited computing power available to the Polar Research team at UIUC in the dim and distant past.)

The values I gave (3.787 and 2.942) were for the 1st and 16th of August in 2012. The date stamp for each of these dates should read approximately 2016.5836 and 2016.6246 when we finally get there next month. (NB the daily increment value ranges from 0.0026 - 0.0029, but since 2012 was also a Leap Year, I would expect the values to match more or less exactly.)

Therefore, although Wipneus' latest figure (2016.5260 <> 5.716 million sq kms) appears to relate to the 13th July, it actually equates historically to values for the 11th. The 1st of August is 19 days on from the 13th July, but the same offset taken from the 11th July only takes us up to July 30th. Similarly, a 34-day offset from 13th July gives your 16th August date, but, from the 11th July, this would equate to the 14th August.

The values for July 30 2012 (2012.5781) and Aug 14 2012 (2012.6191) were 4.085 and 2.986 million sq kms respectively.

[bbr2314]

Thanks Bill!

Today's numbers should be particularly interesting since we didn't get an update yesterday. If the benchmark for 2012 is 4.1M by 8/1 I think that is very doable indeed, and by that point 2016 should be opening quite a lead. Again, I will call for 3.6M km2 on that date this year for anyone who wants to rake me across the coals if I'm wrong.

[Wipneus]

NSIDC servers are back. This is the report from sea ice concentration dated 2016-07-12.

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:


day  CT-date       NH               SH                Global
Tue 2016.5233 -156.1  5.824326 +140.1 12.991531   -16.0 18.815857
Wed 2016.5260 -108.8  5.715507  +14.5 13.006019   -94.3 18.721526
Thu 2016.5288 -127.6  5.587861   -4.2 13.001869  -131.8 18.589730
Fri 2016.5315  -82.5  5.505383  -64.0 12.937885  -146.5 18.443268


Shadow NSIDC is now 8.0671 that is a drop of -170.8 (after a drop of -1801.1 the day before).

Half of the declines (area and extent) are in the Hudson region, where the phrase "falling of a cliff" is only a small exaggeration. Chukchi, ESS and CAB (just area) follow at distance.

The delta map is back too.

[Jim Pettit]

NSIDC servers are back. This is the report from sea ice concentration dated 2016-07-12.

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:


day  CT-date       NH               SH                Global
Tue 2016.5233 -156.1  5.824326 +140.1 12.991531   -16.0 18.815857
Wed 2016.5260 -108.8  5.715507  +14.5 13.006019   -94.3 18.721526
Thu 2016.5288 -127.6  5.587861   -4.2 13.001869  -131.8 18.589730
Fri 2016.5315  -82.5  5.505383  -64.0 12.937885  -146.5 18.443268

That shadow SIA reading for Day 0.5315 is lower than the SIA annual minimum recorded in 1980. One down, 36 to go...

[Bill Fothergill]

That shadow SIA reading for Day 0.5315 is lower than the SIA annual minimum recorded in 1980. One down, 36 to go...

Well spotted, that man!

[Wipneus]

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:

day  CT-date       NH               SH                Global
Wed 2016.5260 -108.8  5.715507  +14.5 13.006019   -94.3 18.721526
Thu 2016.5288 -127.6  5.587861   -4.2 13.001869  -131.8 18.589730
Fri 2016.5315  -82.6  5.505247  -64.0 12.937879  -146.6 18.443126
Sat 2016.5342  -93.5  5.411760  -51.9 12.885998  -145.4 18.297758


Most of that decline is in the ESS: -55k.

Shadow NSIDC is 7.9235 dropping -143.7. Chukchi went down -55k, followed by CAA, CAB and Hudson (all around -20k).

The delta map shows the details.

[seaicesailor]

Correct me if I am wrong but that is about 600k of extent gone in 4 days

[BornFromTheVoid]

Correct me if I am wrong but that is about 600k of extent gone in 4 days

You're right. About 589k in 4 days, 495k in the last 3 (with daily NSIDC extent). Still above 2012 and 2011 though.

[magnamentis]

ok, so can anyone provide the number that is officially considered a "cliff" so that i can learn to use terms correctly or refer to such a statement. yesterday there was a discussions where some guys said the cliff does/did not happen and is this a one. so if i'm wrong here i  at least wanna know for the future and the other way
would of course be even better :-)

[BornFromTheVoid]

ok, so can anyone provide the number that is officially considered a "cliff" so that i can learn to use terms correctly or refer to such a statement. yesterday there was a discussions where some guys said the cliff does/did not happen and is this a one. so if i'm wrong here i  at least wanna know for the future and the other way
would of course be even better :-)

I don't know about "official", but the mega melt weeks (>1m in 7 days) are usually rare enough and long lasting enough to be considered a cliff, rather than a step.

Using the single day NSIDC data, which is influenced somewhat by the correction at the start of July,  we didn't have the first mega melt week until 1991 (which had 2 with overlapping dates). The next occurrence wasn't until 2005, which had 1.
2007 changed things with 6 (overlapping dates again). Of the years from 2008 to 2015, 5/8 saw mega melt weeks.
2012 was the first to have 2 completely separate mega melt weeks, both the earliest and latest on record in June and in August.

The largest 7 day melt period so far this year is the latest one, at -733k, which is provisionally the smallest since 2006 and 9th smallest on record. This will probably change over the next few days though.

So yeah. I go with defining a cliff as being a loss of 1 million km2 or greater over 7 days, the mega melt week.

[magnamentis]

ok, very well, that makes sense, i took it more optical, some kind of steepening from the previous main
direction of the curve but i think what you elaborated can be well used as a definition for now. thanks, very much appreciated.

[Tensor]

ok, very well, that makes sense, i took it more optical, some kind of steepening from the previous main
direction of the curve but i think what you elaborated can be well used as a definition for now. thanks, very much appreciated.

If the last three days would have occurred prior to our comments, I wouldn't have argued with your use of the word cliff.  Although I think Born's definition is a pretty good one.   

[Jim Pettit]

ADS-NIPR Extent:

7,611,399 km2 (14 July)

Down 6,331,108 km2 (45.41%) from 2016 maximum of 13,942,507 km2 on 29 February.

4,433,944 km2 above record minimum extent of 3,177,455 km2 (16 September 2012).

Down 145,972 km2 (-1.88%) from previous day.

Down 735,737 km2  (-8.81%) over past seven days (daily average: -105,105 km2).

Down 1,362,309 km2  (-11.13%) for July (daily average: -97,308 km2).

1,154,419 km2 below 2000s average for this date.

211,858 km2 below 2010s average for this date.

644,187 km2 below 2015 value for this date.

65,254 km2 below 2012 value for this date.

Lowest year-to-date (01 January - 14 July) average.

Lowest July to-date average.

2nd lowest value for the date.

128 days this year (65.64% year-to-date) have recorded the lowest daily extent.

34 days (17.44%) have recorded the second lowest.

19 days (9.74%) have recorded the third lowest.

181 days in total (92.82%) have been among the lowest three on record.

CT Area:

5,411,760 km2 (15 July [Day 0.5343])

Down 7,509,598 km2 (58.12%) from 2016 maximum of 12,921,358 km2 on 29 March [Day 0.2384].

3,177,751 km2 above record minimum area of 2,234,010 km2 (14 September 2012).

Down 93,487 km2 (-1.7%) from previous day.

Down 742,871 km2 (-12.31%) over past seven days (daily average: -106,124 km2).

Down 1,682,508 km2 (-15.11%) for July (daily average: -112,167 km2).

971,742 km2 below 2000s average for this date.

283,265 km2 below 2010s average for this date.

457,741 km2 below 2015 value for this date.

308,751 km2 above 2012 value for this date.

2nd lowest value for the date.

* - NOTE: due to the prolonged absence of official CT sea ice area data, I've incorporated Wipneus' shadow area numbers as calculated from NSIDC data. The official numbers will be inserted if/when available. In the meantime, thanks, Wipneus!

[Wipneus]

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:

day  CT-date       NH               SH                Global
Thu 2016.5288 -127.6  5.587861   -4.2 13.001869  -131.8 18.589730
Fri 2016.5315  -82.6  5.505247  -64.0 12.937879  -146.6 18.443126
Sat 2016.5342  -93.5  5.411760  -51.9 12.886004  -145.4 18.297764
Sun 2016.5370  -66.7  5.345110  -66.6 12.819450  -133.2 18.164560

Biggest decline in the CAA (-24k), followed by Baffin (-19k) and Chukchi (-17k).

Shadow NSIDC is now 7.8817 dropping -41.8k. Strong declines in Chukchi (-40k), Beaufort (-37k) and Baffin (-23k). Upticks in Hudson where the hoovering around 15% seems to have started (+27k) and ESS (+25k).

As always, there is the delta map:

[Jim Pettit]

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:

day  CT-date       NH               SH                Global
Thu 2016.5288 -127.6  5.587861   -4.2 13.001869  -131.8 18.589730
Fri 2016.5315  -82.6  5.505247  -64.0 12.937879  -146.6 18.443126
Sat 2016.5342  -93.5  5.411760  -51.9 12.886004  -145.4 18.297764
Sun 2016.5370  -66.7  5.345110  -66.6 12.819450  -133.2 18.164560

FWIW, 2016 area is now lower than the annual minimums set in 1980, 1983, and 1986.

As always, thanks Wipneus!

[Wipneus]

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:

day  CT-date       NH               SH                Global
Sat 2016.5342  -93.5  5.411760  -51.9 12.886004  -145.4 18.297764
Sun 2016.5370  -66.5  5.345257  -66.2 12.819805  -132.7 18.165062
Mon 2016.5397  -12.8  5.332489  +42.1 12.861951   +29.4 18.194440
Tue 2016.5425  -91.8  5.240669   +5.3 12.867237   -86.5 18.107906


Big drop of area in the CAB (-79k), Beaufort increased by +26k.

Shadow NSIDC extent is now 7.8312 an increase of +23.3k. Big increases in the Beaufort (+107k) and Chukchi (+34k). Extent dropped in the ESS (-36k), Hudson (-27k) and CAA (-26k).

Of course, there is a delta map as well.

[Steven]

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:

Tue 2016.5425    ...   5.240669

Third lowest, behind 2012 and 2007:

[BornFromTheVoid]

Update for the week to July 16th

The current 5 day trailing average is on 7,902,000km2 while the 1 day extent is at 7,831,000km2.

(All the following data is based on a trailing 5 day average)
The daily anomaly (compared to 81-10) is at -1,760,000km2, an increase from -1,576,000km2 last week. The anomaly compared to the 07, 11 and 12 average is at +33,000km2, an increase from +19,000km2 last week. We're currently 3rd lowest on record, the same as last week.



The average daily change over the last 7 days was -109.3k/day, compared to the long term average of -82.9k/day, and the 07, 11 and 12 average of -111.3k/day.
The average long term change over the next week is -89.4k/day, with the 07, 11, and 12 average being -85.6k/day.



The extent loss so far this July is the 10th largest on record. To achieve the largest monthly loss, a drop of at least 129.7k/day is required (requiring >144.1k/day with with single day values), while the smallest loss requires a drop of less than 29.3k/day (<28.3k/day with single day values) and an average loss requires a drop of 73.0k/day (~78.7k/day with single day values).

[Jim Pettit]

To accompany BFTV's charts:

[BornFromTheVoid]

Daily NSIDC extent has hit a wall and is now 4th lowest on record.

Average drop with the single day extent values last 4 days = -19.3k/day (81-10 avg is -87k/day).

[Wipneus]

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:

day  CT-date       NH               SH                Global
Sun 2016.5370  -66.5  5.345257  -66.2 12.819805  -132.7 18.165062
Mon 2016.5397  -12.8  5.332489  +42.1 12.861951   +29.4 18.194440
Tue 2016.5425  -91.8  5.240649   +5.7 12.867640   -86.2 18.108289
Wed 2016.5452 -129.5  5.111185  -20.6 12.847038  -150.1 17.958223


Regionally CAB and Laptev are the main sponsors: -53k and -44k.

Shadow NSIDC extent is 7.8463  an uptick of +15.1k. Most regions change little, Hudson most (+19k).

The delta map is attached as usual.

[Jim Pettit]

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:

day  CT-date       NH               SH                Global
Sun 2016.5370  -66.5  5.345257  -66.2 12.819805  -132.7 18.165062
Mon 2016.5397  -12.8  5.332489  +42.1 12.861951   +29.4 18.194440
Tue 2016.5425  -91.8  5.240649   +5.7 12.867640   -86.2 18.108289
Wed 2016.5452 -129.5  5.111185  -20.6 12.847038  -150.1 17.958223

Thanks, Wipneus. For those keeping score, 2016 SIA is now lower than the September minimums measured in 1979-1980, 1982-1983, 1986-1988, and 1996.

[DoomInTheUK]

Thanks, Wipneus. For those keeping score, 2016 SIA is now lower than the September minimums measured in 1979-1980, 1982-1983, 1986-1988, and 1996.

Thanks Jim - I love this time of year when you start ticking off the previous minima. It helps to keep the current state in perspective.

[Neven]

Neven, I think there's something wrong with those compactness graphs.  Calculating CAJAX, I obtain the same values as in your graph above, for all the years 2007-2015, but not for 2016.  It doesn't seem to be a leap-year problem, since I got the same values as you for 2012 and 2008 (but not for 2016).  I guess the problem is that the CT-area values for the year 2016 in your spreadsheet are not aligned correctly relative to the 2007-2015 data?

Steven, the latest SIA number as reported by Wipneus is 5111185 km2. I have entered this number for July 17th on my spreadsheet. You probably have it for July 16th, right?

[Steven]

Neven, I think there's something wrong with those compactness graphs.  Calculating CAJAX, I obtain the same values as in your graph above, for all the years 2007-2015, but not for 2016.  It doesn't seem to be a leap-year problem, since I got the same values as you for 2012 and 2008 (but not for 2016).  I guess the problem is that the CT-area values for the year 2016 in your spreadsheet are not aligned correctly relative to the 2007-2015 data?

Steven, the latest SIA number as reported by Wipneus is 5111185 km2. I have entered this number for July 17th on my spreadsheet. You probably have it for July 16th, right?

Neven,  Here is a piece of my CT-area spreadsheet (for the last 4 years), with the last column including the latest values reported by Wipneus:

Date            2012      2013      2014      2015      2016
0.5370         5.047     5.905     5.961     5.735     5.345
0.5397         5.037     5.784     5.864     5.616     5.332
0.5425         4.862     5.512     5.735     5.492     5.241
0.5452         4.800     5.351     5.606     5.399     5.111

Could you check whether your CT-area values are aligned in the same way?  I think your CT-area data in the last column (for 2016) are aligned in a different way compared to the other years.

[Neven]

That last row for date 0.5452 is July 18th in my spreadsheet, so the value for 2016 moves up one row and you get:

Date            2012      2013      2014      2015      2016
0.5370         5.047     5.905     5.961     5.735     5.332
0.5397         5.037     5.784     5.864     5.616     5.241
0.5425         4.862     5.512     5.735     5.492     5.111
0.5452         4.800     5.351     5.606     5.399   tomorrow

This has something to do with 0.000 not being the first day of the year, but the last day. Or something like that. I believe that as Wipneus is taking the latest NSIDC data to produce the CT SIA number, the 5.111 value must be for July 17th.

I'm so fed up with this! Every year!   

CT is no longer updated. I think I'm dropping it next year.

[Steven]

Neven,  according to Wipneus' comment today, the latest value (5,111,185 km2) is for CT-date 2016.5452, which corresponds to the last row in my table above.  I assume Wipneus is using the same conventions as in this file, from which I got the corresponding values for the other years:

Date              CT-area     
2015.5452    5.3986096
2014.5452    5.6062360
2013.5452    5.3513508
2012.5452    4.8000555
...

which correspond to the numbers on the last row of the table in my previous comment. 

I don't see any reason to move the values for 2016 in the table up by one row compared to the other years.  I agree that the exact day that corresponds to "date 0.5452" is a bit ambiguous  (it could be interpreted as 17 July, or 18 July or whatever), but as long as the same date 0.5452 is used for all the years there is no risk of confusion.  In any case I'm pretty sure that your CAJAX numbers for 2016 are not consistent with the other years (regardless of how leap years are treated).

[Neven]

I must have made a mistake when switching from the CT SIA data file to Wipneus' estimations. I'll try to remedy that, but it's a PITA.

edit: I'm checking my spreadsheet and 0.5452 for all other years is July 18th, but we can't know 0.5452 for this year yet as it's the 18th today. I've been over and over this with Wipneus (almost every year it seems), but I thought my spreadsheet was in agreement with his. Before, when CT was lagging, and Wipneus gave us the numbers two days ahead, the latest date would correspond with that of IJIS/JAXA.

And that's how it goes for me now too: JAXA SIE number for previous day comes in in the morning, Wipneus reports in the afternoon, et voilà. It would make sense if CT was lagging, not for it to be ahead one day.

Anyway, I'm confused. Wip, please advise, if you have the time.

And like I said, I'm dropping CT SIA next year, because of this annual waste of time and because the people behind CT seem to have switched to other stuff, which is fine, of course, CT being a legendary Arctic - and Antarctic - sea ice data provider.

[Steven]

edit: I'm checking my spreadsheet and 0.5452 for all other years is July 18th, but we can't know 0.5452 for this year yet as it's the 18th today.

Neven, here is the calculation that was posted by Wipneus exactly 1 year ago, on 18 July 2015:

http://forum.arctic-sea-ice.net/index.php/topic,1112.msg57098.html#msg57098

The latest CT-value calculated in that comment was 5.398 million km2.  That is the value for date 2015.5452.

So, it seems that for every year (both leap years and non-leap years), the value for CT-date  0.5452 is the one that is calculated by Wipneus on 18 July, and hence corresponds to satellite observations that were made on 17 July.

[Bill Fothergill]

@ Steven & Neven    RE: CT date formats

A couple of years ago, I exchanged emails with Bill Walsh of the Polar Research team at UIUC. He indeed confirmed that, using their numbering conventions, the date stamp for, say, 2016.0000 would relate to the 31st December 2015.

I don't remember the exact reason(s), but it was something of a historic "accident". In the dim and distant past, that was simply the way some early coding had been set up - I think it was then subsequently felt that introducing a change could cause more confusion than it would eliminate.

(My email system is playing up at the moment, otherwise I would cut&paste the words Bill Walsh actually used - rather than employing my "less than 100% reliable memory".)

As regards yyyy.5452, I've only imported the CT data from 2000 onward, and that date stamp certainly relates to July 18 each time. (Which is Day 200 for leap years, and Day 199 the rest of the time.)

[Wipneus]

@ Steven & Neven    RE: CT date formats

As regards yyyy.5452, I've only imported the CT data from 2000 onward, and that date stamp certainly relates to July 18 each time. (Which is Day 200 for leap years, and Day 199 the rest of the time.)

So far so good. July 18 would be the tag that CT uses and there would be no reason to doubt that. Unless of course you try to replicate it like I have been doing for a couple of years with increasing accuracy. Then you find, as I did, that that requires sea ice concentration data from NSIDC that was tagged 17th of July.

It is really up to you, 17th or 18th. I am trying to be safe reporting 2016.5452

[Neven]

So, basically, if I get this year to be in line with all the other years in my spreadsheet, and then transfer that to my CAJAX sheet, CT will basically be one day ahead of JAXA?

[Tealight]

I had the same problem with CT area in my calculations. Thankfully Wipneus provides 'NSIDC Area' on his website which has the same date as NSIDC extent. lt also excludes lake ice and uses correct grid cell sizes. My forecast model improved by 10% thanks to him.

If CT won't switch to the new F18 data this year than maybe we should switch completly to 'NSIDC Area'.

[Wipneus]

So, basically, if I get this year to be in line with all the other years in my spreadsheet, and then transfer that to my CAJAX sheet, CT will basically be one day ahead of JAXA?

If CT reports data from the 17th as the 18th, I would call that "be one day behind".

[Wipneus]

Here is the shadow CT-area report based on calibrated F18 NSIDC sea ice concentration data:
day  CT-date       NH               SH                Global
Mon 2016.5397  -12.8  5.332489  +42.1 12.861951   +29.4 18.194440
Tue 2016.5425  -91.8  5.240649   +5.7 12.867640   -86.2 18.108289
Wed 2016.5452 -129.8  5.110889  -20.5 12.847105  -150.3 17.957994
Thu 2016.5479  -83.6  5.027240  -14.1 12.833030   -97.7 17.860270

Serious declines in CAB (-27k), ESS (-24k) and Laptev (-21k).

Shadow NSIDC is now 7.7462 dropping -100.2k. The way to the bottom is led todayby ESS (-57k) and Baffin (-35k).

Of course a delta map is attached to the daily shadow report.

[Neven]

Okay, I've slept only 3 hours last night, and been working and designing some stuff since 6 AM. And now I've spent a whole hour to see where the mistake is.

Neven,  Here is a piece of my CT-area spreadsheet (for the last 4 years), with the last column including the latest values reported by Wipneus:

Date            2012      2013      2014      2015      2016
0.5370         5.047     5.905     5.961     5.735     5.345
0.5397         5.037     5.784     5.864     5.616     5.332
0.5425         4.862     5.512     5.735     5.492     5.241
0.5452         4.800     5.351     5.606     5.399     5.111

Could you check whether your CT-area values are aligned in the same way?  I think your CT-area data in the last column (for 2016) are aligned in a different way compared to the other years.

Steven, I've just checked the CT SIA data file and the numbers correspond with the data up to 2015. However, if I look at Wipneus' file, I get this:

Date        Steven    Wip
0.5370         5.345     5.332
0.5397         5.332     5.241     
0.5425         5.241     5.111     
0.5452         5.111     5.027     

Could you please check your spreadsheet and compare to Wipneus' AreaCalculatedLikeCryosphereToday.txt file?

[Rob Dekker]

It seems that the ice dispersion due to persistent lows over the Arctic is finally taking a toll on the Central Arctic Basin.

Over the past couple of days, Sea Ice Area in the CAB has taken an unprecedented nose dive on Wipneus' AMSR2 graph :



When 'area' drops, 'extent' drops are typically close to follow.
So expect some dramatic drops in the major indexes in the next couple of days.

[Pmt111500]

This has something to do with 0.000 not being the first day of the year, but the last day. Or something like that. I believe that as Wipneus is taking the latest NSIDC data to produce the CT SIA number, the 5.111 value must be for July 17th.

Ah, there it is again. This explained the day or two error between my and some other' calculations. Too bad the CT SIA old values from 1979-2000? are cannot be scientifically correlated to the other area values. I've almost stopped following the progression of the melt this year, but I guess I need to come back to check the situation in nov-dec.

[dnem]

Wipneus's CAB graph posted above strikes me as the most relevant and important graph with respect to where this season might be headed that I've seen so far this year.  The drop is "unprecedented" more in how early it is than how steep or long.

[Jim Pettit]

Cyrosphere Today has *always* had a bit of trouble with dates, as we've seen and discussed time and again.

For starters, they've handled leap years/leap days in an inconsistent fashion:

• For 1980, a Day 0.1631 was added between Days 0.1616 and 0.1644
• For 1984, a Day 0.1658 was added between Days 0.1644 and 0.1671
• For 1988, 1992, 1996, 2000, and 2004, they inserted an extra Day 0.1644 (meaning those years have two)
• For 2008, they inserted an extra Day 0.4438 (meaning there are two of them).
• For 2012 and 2016, they again inserted an extra Day 0.1644, so there are two of them per year,

On top of that, there's the "which day is the actual start of the year?" question. Obviously, some use Day 0.0000 as the December 31, which doesn't really make a lot of sense, given that, for instance, December 31 this year would be "2017.0000". Others--such as myself--consider the day as belonging to the year represented by the four digits to the left of the decimal.

Meh.

The thing for me has always been consistency. That is, we're not so much concerned with daily values as we are about trends; it's less important to know on what calendar day the annual minimum fell than it is to know just what the minimum was.

Having said all that: given that CT's SIA file has shown the same value for 74 consecutive days now, and given the vagaries of area measurements, I too would be happy to drop it as a metric here. It's simply too much work to illustrate a sometimes nearly valueless number...

[Steven]

Steven, I've just checked the CT SIA data file and the numbers correspond with the data up to 2015. However, if I look at Wipneus' file, I get this:

Date        Steven    Wip
0.5370         5.345     5.332
0.5397         5.332     5.241     
0.5425         5.241     5.111     
0.5452         5.111     5.027     

Could you please check your spreadsheet and compare to Wipneus' AreaCalculatedLikeCryosphereToday.txt file?

Neven, you're not aligning the "Wip" values correctly.  Compare with the values in this comment by Wipneus upthread:

    CT-date              NH         ...                     
    2016.5370  ...  5.345257  ...
    2016.5397  ...  5.332489  ...
    2016.5425  ...  5.240649   
    2016.5452  ...  5.111185 

[Neven]

We're getting to the crux of the matter.

On July 18th and 19th Wipneus wrote this:

2016.5452 - 5.111185

But in the AreaCalculatedLikeCryosphereToday.txt data file it says (see attachment):

2016.5452 - 5.027240

So, what is it? I couldn't be more confused.