1

**Arctic sea ice / Re: IJIS**

« **on:**April 25, 2017, 10:51:51 PM »

(this time with graph!)

Read All About Arctic on ASI Blog –
Watch all Graphs on ASI Graphs

March for Science & Please share this Forum!

Please support this Forum and Neven's Blog

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Pages: [**1**]

2

One more - this time, 10 year averages.

3

An update of my chart showing cumulative days in the bottom 3 for each date. Also a grid view showing the numbers.

2017 has now overtaken 2006, 2007, 2011 and 2015. Unless there are a couple of extended plateaux, I'd be surprised if 2017 spends much time outside the bottom 3 to the end of May.

Finally, an update of my 2017 vs 1990-16 daily range chart, now shifted to April & May.

2017 has now overtaken 2006, 2007, 2011 and 2015. Unless there are a couple of extended plateaux, I'd be surprised if 2017 spends much time outside the bottom 3 to the end of May.

Finally, an update of my 2017 vs 1990-16 daily range chart, now shifted to April & May.

4

My concluding chart for the max season.

5

Very interesting Deeennggee

Really simple yet frightening graph there.

Soon wel just be talking about the sea.

Or for anyone on the other side of the pond, 1.4m sqkm = California + Oregon + Washington State + Idaho + Nevada

6

"Let it go, let it go!!" Kind words indeed. I can only assume you've been surrounded by some very average graphs in the past.

Somethin' else here - IJIS & NSIDC extent anomalies, 2017 vs two different reference periods.

As my sense of scale often disappears when looking at maps of the Arctic: 1.4m sq km (the current variance from the 1990s average) is the same as the combined area of France, Switzerland, Germany, Luxembourg, the Netherlands, Belgium, Denmark and the UK.

Somethin' else here - IJIS & NSIDC extent anomalies, 2017 vs two different reference periods.

As my sense of scale often disappears when looking at maps of the Arctic: 1.4m sq km (the current variance from the 1990s average) is the same as the combined area of France, Switzerland, Germany, Luxembourg, the Netherlands, Belgium, Denmark and the UK.

7

Here's my corrected plot of days in the bottom 3 for each year (lesson: more haste, less speed).

For each date I've calculated the actual ranking to be in the bottom 3; this is needed because of data gaps (eg late July-early August 2002), which throws all the other years' rankings by one place. So I've definitely identified the bottom 3 years for each date now. This gives the correct cumulative plot, attached. I've excluded years that by Dec 31st had accumulated fewer than 20 bottom 3 days for any date.

Thanks to all who responded to my first, rather excitable post on this one - you were sceptical in the proper sense of the word.

Looking at the next 10 days: 2005, 2006, 2007, 2015 and 2016 all have days that are currently 25th out of 27 (see excerpt of rankings attached from my .xls). In the likely event that 2017 extent remains in the bottom 3, its cumulative score will gain one each day, and one of those other years will lose one each time.

The inevitable question is "so what?" Does this (in March) really matter when it comes to the showdown - the summer minimum? I don't really know the answer to that, but my hunch is that it*might* say something about the overall condition of the ice.

And don't worry, I won't post overly regular updates of the chart! It could get a bit tedious. But maybe monthly...

For each date I've calculated the actual ranking to be in the bottom 3; this is needed because of data gaps (eg late July-early August 2002), which throws all the other years' rankings by one place. So I've definitely identified the bottom 3 years for each date now. This gives the correct cumulative plot, attached. I've excluded years that by Dec 31st had accumulated fewer than 20 bottom 3 days for any date.

Thanks to all who responded to my first, rather excitable post on this one - you were sceptical in the proper sense of the word.

Looking at the next 10 days: 2005, 2006, 2007, 2015 and 2016 all have days that are currently 25th out of 27 (see excerpt of rankings attached from my .xls). In the likely event that 2017 extent remains in the bottom 3, its cumulative score will gain one each day, and one of those other years will lose one each time.

The inevitable question is "so what?" Does this (in March) really matter when it comes to the showdown - the summer minimum? I don't really know the answer to that, but my hunch is that it

And don't worry, I won't post overly regular updates of the chart! It could get a bit tedious. But maybe monthly...

8

Ah. Methodology flaw alert - thanks Oren. The COUNTIF was looking for days ranked 28th, 27th and 26th. And for dates we haven't had this year yet, they actually only go from 1 to 27.

I've redone it so it counts if a day is ranked 25th, 26th and 27th i.e. the previous 'bottom 3'. That means that 2006, 2007, 2011 and 2015 gain some bottom 3 days - so yes, by the end of their years they have more bottom 3s than 2017 at the moment. But not by much. As things stand, 2017 could well overtake those year by the end of April.

This needs some more work! Good to have friendly peer-reviewers.

I don't have an online graph repository - any suggestions? Google Drive?

I've redone it so it counts if a day is ranked 25th, 26th and 27th i.e. the previous 'bottom 3'. That means that 2006, 2007, 2011 and 2015 gain some bottom 3 days - so yes, by the end of their years they have more bottom 3s than 2017 at the moment. But not by much. As things stand, 2017 could well overtake those year by the end of April.

This needs some more work! Good to have friendly peer-reviewers.

I don't have an online graph repository - any suggestions? Google Drive?

9

Again, but with an essential clarification in the title!

(The graph doesn't include years that don't feature a day in the bottom 3 for the date)

(The graph doesn't include years that don't feature a day in the bottom 3 for the date)

11

Sorry, my explanation of daily rankings was a bit convoluted. To confirm, I'm ranking each year for each date of the year. So, all the January 1sts ranked, then all the January 2nds and so on.

The simplest expression is maybe 'bottom 3 days for each date'.

2017 has already had 77 of those. And indeed what i was saying is that only 2012 and 2016 had more bottom 3 days for each date over their entire year than 2017 has already had. 2012 racked them up during the summer, 2016 had them more spread out.

I'll do a cumulative plot later which should show it reasonably clearly.

The simplest expression is maybe 'bottom 3 days for each date'.

2017 has already had 77 of those. And indeed what i was saying is that only 2012 and 2016 had more bottom 3 days for each date over their entire year than 2017 has already had. 2012 racked them up during the summer, 2016 had them more spread out.

I'll do a cumulative plot later which should show it reasonably clearly.

12

So far, 2017 has had 77 days ranked in the bottom 3 for each day's comparative ice extent over the 1990-2017 period, and 2 days not in the bottom 3.

This means 2017 has already overtaken*every single year *for days in the bottom 3 for that day, apart from 2012 and 2016.

By day 79, 2016 had 58 days in the bottom 3 for 1990-2017

By day 79, 2012 had 2 days in the bottom 3 for 1990-2017.

I don't usually do emoticons, but anyway:

This means 2017 has already overtaken

By day 79, 2016 had 58 days in the bottom 3 for 1990-2017

By day 79, 2012 had 2 days in the bottom 3 for 1990-2017.

I don't usually do emoticons, but anyway:

13

Thanks Meirion and Bill F. It would have helped if I'd labelled the X axis with 'day number' or something like that!

One can also have a version that shows the trajectory of each year - rather than the grey range - but it's very noisy. Although, as indicated in the blog comments, I've embraced the noise for artistic effect (inspired by Joy Division) over at the 'cafe' thread.

One can also have a version that shows the trajectory of each year - rather than the grey range - but it's very noisy. Although, as indicated in the blog comments, I've embraced the noise for artistic effect (inspired by Joy Division) over at the 'cafe' thread.

14

(Because a) I'd feel sheepish putting this in the main IJIS thread, b) I wanted somewhere to link to from the main blog, and c) this is a home for random offtopicness...)

The first in a likely short lived series, 'Arctic charts that resemble album covers', I give you my IJIS / Unknown Pleasures mashup.

The first in a likely short lived series, 'Arctic charts that resemble album covers', I give you my IJIS / Unknown Pleasures mashup.

15

Weekly update of my 2017 vs 1990-2016 maxima chart.

16

Here's my update, also with the late season rallying years that Jim P mentions. From top 2003, 2012, 2014, 2015.

17

But it's kind of fun to have another moment in the year to speculate about. The max is in polar (haha) opposition to the minimum, and it marks the passing of one phase of the year to the other. Plus, there is a downward trend in the annual max, which is interesting in itself and part of the bigger story...

18

As per before, but now going back to 1990. 14.0 looks to me like a coin flip, albeit based on eyeballing and absolutely nothing scientific, meteorological etc.

19

An update of my '2017 vs the others' maximums' chart. I've extended it to go back to 1990.

As most of you probably know the 1980s dataset only has every-other-day readings, so there's a lot less confidence during that decade that each year's max was indeed the max.

As most of you probably know the 1980s dataset only has every-other-day readings, so there's a lot less confidence during that decade that each year's max was indeed the max.

20

Bill

Thanks for that, and I enjoyed that post of yours that you linked to there. One of the first things I thought of when I produced the heat chart above was exactly that, about how it shows the paucity of the contractions argument when they find a recent day in April, usually, that is superficially similar to another April day in the dim and distant.

Incidentally, my use of the word 'concocted' earlier wasn't meant to imply some originality on my part ref my heat chart. 'Assembled' would probably have been a better verb to use.

Thanks for that, and I enjoyed that post of yours that you linked to there. One of the first things I thought of when I produced the heat chart above was exactly that, about how it shows the paucity of the contractions argument when they find a recent day in April, usually, that is superficially similar to another April day in the dim and distant.

Incidentally, my use of the word 'concocted' earlier wasn't meant to imply some originality on my part ref my heat chart. 'Assembled' would probably have been a better verb to use.

21

Another cheeky chart of IJIS extent data if I may.

Having ranked each day from 1990-2017 against its respective days (where 1 is the most ice for that day and 28 is the least), I counted the number of days that each year has in each rank (1 to 28). I then grouped them: top 5, 6-10, 19-23 and bottom 5 (24-28).

I shouldn't be surprised at the result, but plotting it out like this is still pretty staggering.

Between 1990-2003 there were only*17 days* in total that are ranked in the bottom 10 (as in, least ice) for their respective days over the 1990-2017 period. And from 2011 to today, only 11 days ranked in their respective top 10 - and those were all in April 2012!

Having ranked each day from 1990-2017 against its respective days (where 1 is the most ice for that day and 28 is the least), I counted the number of days that each year has in each rank (1 to 28). I then grouped them: top 5, 6-10, 19-23 and bottom 5 (24-28).

I shouldn't be surprised at the result, but plotting it out like this is still pretty staggering.

Between 1990-2003 there were only

22

As I'm in the mood for trying different visualisations, here's a thing I concocted as another way to look at IJIS data. The procedure:

Downloaded the daily extent data (1990-2017).

Used the RANK equation to rank each year - from 1 to 28 where 1 is the largest extent for any year on that day, and 28 is the least.

Then conditional formatting so that 1 is blue and 28 is red, and everything graded in between.

Then made the numbers invisible, pasted into powerpoint and squeezed it vertically.

Quite a pleasing effect, and moderately informative!

No idea how this will look when I press Post, but here goes....

Downloaded the daily extent data (1990-2017).

Used the RANK equation to rank each year - from 1 to 28 where 1 is the largest extent for any year on that day, and 28 is the least.

Then conditional formatting so that 1 is blue and 28 is red, and everything graded in between.

Then made the numbers invisible, pasted into powerpoint and squeezed it vertically.

Quite a pleasing effect, and moderately informative!

No idea how this will look when I press Post, but here goes....

23

Looking at the 2003-2016 February and March IJIS data there are quite a lot of runs of 3 or 4 consecutive days of extent going down *before* the maximum has been reached.

There were runs of 5 consecutive days of decrease prior to the eventual max in 2009 and 2010.

2003 even had 8 consecutive days of decrease, losing 258k sq km in the process before it then went up again, gaining 277k sq km and hitting its maximum.

There were runs of 5 consecutive days of decrease prior to the eventual max in 2009 and 2010.

2003 even had 8 consecutive days of decrease, losing 258k sq km in the process before it then went up again, gaining 277k sq km and hitting its maximum.

24

Another update here.

Will extent follow the typical path for the 2010s, upwards until mid-March.

Or is it a case of 'Averages? Pah!' ?

Will extent follow the typical path for the 2010s, upwards until mid-March.

Or is it a case of 'Averages? Pah!' ?

25

My weekly update, this time with the 'average for the 2010s' trajectory superimposed, which on its own suggests 14.0 might still be passed. Not that the Arctic seems bothered about averages this decade; certainly not he last 12 months. Anyway, make of it as you will...

26

As mentioned a weekly update of my 2017 vs the rest maximum chart.

As we had a sharp uptick last week I've picked out 2005 for its impressive rollercoaster profile around the max.

As we had a sharp uptick last week I've picked out 2005 for its impressive rollercoaster profile around the max.

27

An update of my 'when and how big was the max' chart. Visually simplified - now a bit less like a random display of M&Ms / Smarties.

Given the upwards excursion this week just gone I've also picked out 2005, which showed impressive rollercoaster style in late Feb - mid March.

Given the upwards excursion this week just gone I've also picked out 2005, which showed impressive rollercoaster style in late Feb - mid March.

28

Cue David Rose article: "Arctic sea ice passes 2016 levels, why are we spending billions on climate change?"

29

Thanks for the feedback. I'll do a weekly update on this one until we're definitely past the max. Then an equivalent for the minimum.

30

I agree Crandles, pretty sure it will. Interesting times either way.

31

Given this thread's title, here's a graph I made that I also put up on the general IJIS thread yesterday. The legend is provocatively located - we'll see if 2017 can find its way around!

32

Thought I'd add this to the mix: when maxima/ums occurred since 2003, and the extent for each one. Plus the daily range for Feb-Mar and 2017.

33

Here's an update of my despaghettified graph - more like a double bass drawn by Salvador Dali.

For the 2010s average, I've included 2010,2011,2012,2013 & 2014. Does anyone know if that's the same convention that JAXA use, given that people often disagree about the year in which a decade starts?

Anyway, on this basis the average daily drop in week 1 of August this year (68,314), was very similar to the August week 1 average daily drop in 2010-2014 (69,355). The August week 2 average daily drop in the 2010-2014 is 77,080.

For the 2010s average, I've included 2010,2011,2012,2013 & 2014. Does anyone know if that's the same convention that JAXA use, given that people often disagree about the year in which a decade starts?

Anyway, on this basis the average daily drop in week 1 of August this year (68,314), was very similar to the August week 1 average daily drop in 2010-2014 (69,355). The August week 2 average daily drop in the 2010-2014 is 77,080.

34

Espen

For your consideration: a graph (attached) that adds an average for the 2010s, the range for the 2010s, and removes all the individual years apart from 2015. Visually I like this approach as you see the current year in the context of all the recent years, i.e. the shaded range for the 2010s. I know some like to see all the recent years plotted in order to compare with current year, but this method gives a bit more of an overview.

Anyway, this type could complement your daily charts, perhaps as a weekly update.

(I haven't worked out how to insert the image within this text, so it's attached)

For your consideration: a graph (attached) that adds an average for the 2010s, the range for the 2010s, and removes all the individual years apart from 2015. Visually I like this approach as you see the current year in the context of all the recent years, i.e. the shaded range for the 2010s. I know some like to see all the recent years plotted in order to compare with current year, but this method gives a bit more of an overview.

Anyway, this type could complement your daily charts, perhaps as a weekly update.

(I haven't worked out how to insert the image within this text, so it's attached)

Pages: [**1**]