Please support this Forum and Neven's Blog

Author Topic: Home brew AMSR2 extent & area calculation  (Read 819786 times)

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #100 on: July 05, 2013, 08:49:28 AM »
Regions of the day are Laptev and ESS. Net decline has been steady here lately, but the detail shows in Laptev it is still a balance between red and blue.
If the levels are going to follow the 2012 curves, some heavy melting will have to start soon.

(image attached, log in to see)

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #101 on: July 05, 2013, 09:23:58 AM »
And here is Beaufort, Chukchi. I have marked with ellipses where Jaxa/IJIS AMSR2 ice monitor shows polynya from July 1-4. They are clearly not present here.

(imaged attached, log in to see)




Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #102 on: July 06, 2013, 08:19:09 AM »
I cannot update now :-(

A new "feature" was added to my source data. From a new README:

We apply monthly ocean masks to the data according to the masks used by NSIDC
(http://nsidc.org/data/smmr_ssmi_ancillary/ocean_masks.html). These masks are interpolated
to the finer resolution used here and manually corrected for coarse coastlines at the original
25 km resolution.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!! APPLICATION OF OCEAN MASKS STARTS FROM 6 JULY 2013 !!
!! RE-PROCESSING OF HITHERTO PRODUCED DATA IS ONGOING !!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!


Unfortunately, this ocean mask seems to be applied upside down making the data quite unusable. I have notified the people from the U-Hamburg about the problem.

ftp://ftp-projects.zmaw.de/seaice/AMSR2/README.txt

wanderer

  • ASIF Citizen
  • Posts: 156
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #103 on: July 06, 2013, 08:21:44 AM »
Oh no! I was waiting all morning for your update and if we come closer to 2012...

I cannot update now :-(

A new "feature" was added to my source data. From a new README:

We apply monthly ocean masks to the data according to the masks used by NSIDC
(http://nsidc.org/data/smmr_ssmi_ancillary/ocean_masks.html). These masks are interpolated
to the finer resolution used here and manually corrected for coarse coastlines at the original
25 km resolution.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!! APPLICATION OF OCEAN MASKS STARTS FROM 6 JULY 2013 !!
!! RE-PROCESSING OF HITHERTO PRODUCED DATA IS ONGOING !!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!


Unfortunately, this ocean mask seems to be applied upside down making the data quite unusable. I have notified the people from the U-Hamburg about the problem.

ftp://ftp-projects.zmaw.de/seaice/AMSR2/README.txt

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #104 on: July 06, 2013, 09:00:33 AM »
Little choice today, the region of the day is Central Arctic.
Little pixels blue and red mostly cancel over a few days.

Below is the filled in "pole hole".

(image attached, log in to see)

 

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #105 on: July 07, 2013, 09:32:46 AM »
Update 20130706:

The error in the upstream data corrected now for July 5 and 6. Unfortunately the data file for July 4 is still unusable. I have tried to correct this using a manually made backup, but am only 99% confident that everything is OK now ( For a particular date I compare with previous dates to remove "ghost ice"; the misapplied mask does affect the 5/6 July).

Extent:  +96k5 (+325k vs 2012)
Area: +50.7 (+404k9 vs 2012)

So we have an uptick here, bit unusual after the extremely smooth ride we had for weeks. Also the the SSMIS data have a small uptick for extent (+5k) and minimal decline in area (-3k), so it is probably real, measuring with different satellites

What is noticeable is that the uptick is shown over many regions, Hudson the biggest followed by CAA and Baffin. Only the CAB shows a decline in extent (but increase in area).

The details:
Extent:
           Arctic Basin       East Siberian Sea              Laptev Sea
                   -4.7                     2.0                     4.1
               Kara Sea             Barents Sea           Greenland Sea
                    6.4                     2.9                    13.6
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   13.5                     0.1                    37.9
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   23.7                     0.8                    -5.9
             Bering Sea          Sea of Okhotsk            Total Extent
                    0.4                     1.8                    96.5

Area:
           Arctic Basin       East Siberian Sea              Laptev Sea
                   16.4                    -3.4                    10.4
               Kara Sea             Barents Sea           Greenland Sea
                   -0.9                     2.2                    17.6
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   13.2                     0.0                    -1.4
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   20.1                    -9.0                   -15.5
             Bering Sea          Sea of Okhotsk              Total Area
                    0.2                     0.8                    50.7
>


Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #106 on: July 07, 2013, 09:43:40 AM »
Region of the day is Hudson. Large areas are now covered with low concentration ice, near the cutoff of 15%. Small changes have large impacts on the extent measure.

(log in to see the attached image)

Neven

  • Administrator
  • ASIF Governor
  • *****
  • Posts: 3272
    • View Profile
    • Arctic Sea Ice Blog
Re: Home brew AMSR2 extent & area calculation
« Reply #107 on: July 07, 2013, 09:48:44 AM »
There's some flashing and unflashing in Hudson and Baffin on the UB SIC maps, so maybe that explains the uptick.
Il faut cultiver notre jardin

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #108 on: July 07, 2013, 11:38:56 AM »
Attached (log in to see) update of the comparative AMSR2-3.125km SSMIS-12.5km SIE and SIA graphs.
They are not identical, but quite similar.


AartBluestoke

  • ASIF Lurker
  • Posts: 15
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #109 on: July 07, 2013, 02:52:52 PM »
Every area which is going to melt will melt out at some stage will have a maximum contribution of exactly the same as last year -- 100% of the current ice. Estimating what happens this year is determined by the ice left over.

Last year the following areas melted out: (from the area graph in the first post)
East Siberian Sea
Laptev Sea
Chukchi Sea
- Kara Sea
- Beaufort Sea
= Baffin/Newfoundland Bay
= Hudson Bay
* Barerits Sea
* Bearing Sea
* St Lawrence
* Sea of Okhotsk
 ( - = melting slower than last year)
 ( = = major melt pattern underway, complete melt highly probable)
 ( * = already melted out this year)
---
Other Regions
Arctic Basin
Greenland Sea
Canadian Archipelago

---
In estimating the final minimum extent i think 2 questions should be asked:
a) will the areas which melted out last year melt out this year?
  a1) the East Siberian Sea and the Beaufort sea have the furthest to go, but melting is only just starting.
b) what will be the extent of the surviving ice areas be?
  b1) the majority of this is the arctic basin, and that is running about the same as last year
  b2) could the greenland sea almost completely melt out, as we are below last year by more than the minimum ice area from last year.

if every area that was 100% melted for more than a month last year melts out this year, and the greenland Sea area is worse, the final "do we set a record extent" is almost completely determined by the arctic basin.

I don't know enough ice dynamics/ have enough forecasting ability/... to even make a guess at this, but i can see that keeping exact track of areas which we already know will have 0 contribution at the final minimum could just be adding noise to the system.

dree12

  • ASIF Lurker
  • Posts: 24
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #110 on: July 07, 2013, 04:30:14 PM »
  b2) could the greenland sea almost completely melt out, as we are below last year by more than the minimum ice area from last year.

AFAIK the Greenland Sea cannot almost melt out because any open water created will be replaced by ice exiting the Arctic. Greenland is actually likely to be the last sea with summer ice, as in the next few years it is possible the Central Arctic can become a killing ground with only the ice escaping through Greenland surviving.

BornFromTheVoid

  • ASIF Middle Class
  • Posts: 930
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #111 on: July 07, 2013, 04:47:29 PM »
I'd just like to extend my thanks to Wipneus for his work here. It's become part of my morning sea ice routine!

AFAIK the Greenland Sea cannot almost melt out because any open water created will be replaced by ice exiting the Arctic. Greenland is actually likely to be the last sea with summer ice, as in the next few years it is possible the Central Arctic can become a killing ground with only the ice escaping through Greenland surviving.


Hi dree12
The sea ice can melt out on the Greenland sea, though it doesn't happen too often. The end of summer/early Autumns of 2002 and 2003 had just a few thousand km2 of ice left in the Greenland sea at most.



Bob Wallace

  • ASIF Upper Class
  • Posts: 1123
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #112 on: July 07, 2013, 07:12:37 PM »
Recently we've had little flow of ice through the Fram.  The Greenland Sea is melting out and most likely warming.  This means that ice coming though later in the season will meet warmer water and melt quicker.  A melt accelerating factor.

Extent/area numbers are being held high by the slow start of melting in Kara and Beaufort.  In all other regions melt has been as fast or faster than 2012.  Both Kara and Beaufort are now melting with curves resembling 2012, only with a delayed onset.  Even with a one month delay there is more than enough time to melt out those regions.  2012 hit zero almost two months before the end of the season.

There's no way that 2013 is out of the running for a new record.  It all comes back to the big hole which seems to be developing in the CAB.  If that opens up some serious area loss is going to occur.

Wipneus has provided a most excellent tool for comparing 2013 and 2012. 

(Is regional volume possible?  Even modeled would be a help.  We need to focus our attention on the CAB, this could be the year we realize its weakness.)

jonthed

  • ASIF Lurker
  • Posts: 65
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #113 on: July 07, 2013, 07:27:19 PM »
Melting in the CAB is surely a bit of an unknown quantity given the unusual distribution and composition of the ice this year.

If the hole opens up then there will be a lot more edge to be eaten away at, if the drift of the ice makes the hole 'close by compressing the outside inward, then the hole won;t open but the area would shrink slightly faster than normal as beside edge melt there would be this inward movement and concentration. Both of these seem to frustrate any attempt to compare with previous years.

To me it seems it is the drift direction in the CAB (that will determine the fate of the hole) that is critical to the total melt this season.

jdallen

  • ASIF Upper Class
  • Posts: 2333
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #114 on: July 07, 2013, 08:56:20 PM »
Melting in the CAB is surely a bit of an unknown quantity given the unusual distribution and composition of the ice this year.

Absolutely.  What we do know is, there is a lot more low-albedo surface there to collect energy. The amount and type of cloud cover I think is the key factor there right now.

If the hole opens up then there will be a lot more edge to be eaten away at, if the drift of the ice makes the hole 'close by compressing the outside inward, then the hole won;t open but the area would shrink slightly faster than normal as beside edge melt there would be this inward movement and concentration. Both of these seem to frustrate any attempt to compare with previous years.

It's not really a hole so much as its a vast expanse of ice cubes of varying size and composition (FYI, MYI, & "slush").  The area outside of the hole isn't so much solid "surface" as much as it is simply higher concentration of The Same.  The hole is an illusion of concentration.  Compression won't fully close up the leads, nor will lateral/edge melt be any significant factor in melting except for "slush" and the smallest floes ( < 100M diameter or so, from what I've read).  That said, comparison to previous years is very problematic.  We are uncertain of water temperature, ice thickness, and just how much insolation is actually being captured.

To me it seems it is the drift direction in the CAB (that will determine the fate of the hole) that is critical to the total melt this season.

This I'm not so sure of.  The fact we got a "hole" is the key fact.  That hole closing or opening more simply reflects areas of concentration changing; the open water available to capture heat will remain the same or increase, right up until the refreeze starts.  That there *is* drift carries its own peril; ekman pumping will bring heat to the ice even if the albedo increases. Hope for high, thick clouds, over as wide a region as possible....
This space for Rent.

Artful Dodger

  • ASIF Citizen
  • Posts: 436
  • The era of procrastination is coming to a close.
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #115 on: July 08, 2013, 08:17:59 AM »
To me it seems it is the drift direction in the CAB (that will determine the fate of the hole) that is critical to the total melt this season.

This I'm not so sure of.  The fact we got a "hole" is the key fact.  That hole closing or opening more simply reflects areas of concentration changing; the open water available to capture heat will remain the same or increase, right up until the refreeze starts.  That there *is* drift carries its own peril; ekman pumping will bring heat to the ice even if the albedo increases. Hope for high, thick clouds, over as wide a region as possible....

Good analysis.  8)

I think there is one more risk related to the 2012 and 2013 Spring cracking events in the CAB.

With sufficient loss of MYI, we see that the ice is now highly vulnerable to strong winds during the cold season. Couple those with the trans-Arctic drift current, and there is a real possibility that a big portion of the MYI remaining after the Summer minimum could be swept into the Atlantic over the Winter.

We wake up in Spring 2014 with >90% FYI. Then it all melts out by end of Summer 2014.  Could also happen in 2015, 16, or later. :o

Not saying that this will happen, but it is one possible trajectory for the forced disappearance of the perennial ice pack.
Cheers!
Lodger

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #116 on: July 08, 2013, 08:28:43 AM »
daily update 20130707.

Extent : -96k (+340k vs 2012)
Area: -88k5 (+423k vs 2012)

Biggest decline in Kara, Laptev follows as well. Also the -26k of Chukchi  is remarkable. Hudson recovered a bit from yesterdays increase. CAA has another big increase. Beaufort has big area increase but manages to show an extent decline.

Graphics are in the top post.

The details:

Extent:
           Arctic Basin       East Siberian Sea              Laptev Sea
                  -14.7                    -2.9                   -18.4
               Kara Sea             Barents Sea           Greenland Sea
                   -9.3                    -0.2                   -17.2
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                    3.3                     0.0                   -35.2
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                    3.9                     3.4                    -9.3
             Bering Sea          Sea of Okhotsk            Total Extent
                   -1.1                     1.7                   -96.0

Area:
           Arctic Basin       East Siberian Sea              Laptev Sea
                    3.0                    -1.1                   -14.2
               Kara Sea             Barents Sea           Greenland Sea
                  -32.0                     0.0                   -14.6
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                    9.7                     0.0                   -26.0
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   19.5                    -5.6                   -26.6
             Bering Sea          Sea of Okhotsk              Total Area
                   -0.7                     0.2                   -88.5

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #117 on: July 08, 2013, 08:42:14 AM »
Regions of the day are the Chukchi and East Siberian Seas. Chukchi has shown a steady decline at a rate slightly below 2012. The ESS should start declining about now if it wants to copy 2012.

(log in to see attached image)

 

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #118 on: July 08, 2013, 09:22:59 AM »
Comparison Arctic Roos Norsex Area with AMSR2 3.125 Area. Arctic Roos use a 25km grid:

http://arctic-roos.org/observations/satellite-data/sea-ice/ice-area-and-extent-in-arctic

All years except  2012 and 2013 were removed.

(log in to see attachment)


ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: Home brew AMSR2 extent & area calculation
« Reply #119 on: July 08, 2013, 06:33:17 PM »
Wipneus,

A technical question not directly related to this thread. I've asked at Tamino's but got no response from the commenters there.

I have a pattern in sea level pressure (SLP) which is the average SLP for summer 2007 to 2012. I want to make a numerical index of the presence of this pattern in daily/monthly/seasonal SLP data from NCEP/NCAR.

In an initial play around with monthly and JJA averages I have produced the sort of behaviour I have been expecting; from 2007 the JJA averages and almost all monthly averages show a strong index, years before 2007 show a weak index. But I've just used Excel's Correl function (Pearson correlation) applied to 2d matrices of SLP across the northern hemisphere north of 30degN.

I'm not sure correlation is really the right tool for the job. I've been digging around and EOFs seemed hopeful, but they seemed to be mainly used for finding patterns, not deducing the strength of a standard pattern in data.

In producing the AO Index the NOAA use a 'loading pattern' and project observed SLP onto it so as to get a numerical index of the strength of the AO pattern. That's the sort of thing I will be doing, and if I draw a blank I'll email them to ask how they 'project'.

Have you ever done anything like this? Can you point me in the right direction?

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #120 on: July 08, 2013, 07:58:04 PM »
I have not done anything like this, but don't mind to think along.


I have a pattern in sea level pressure (SLP) which is the average SLP for summer 2007 to 2012. I want to make a numerical index of the presence of this pattern in daily/monthly/seasonal SLP data from NCEP/NCAR.


Unfortunately I am lost here already.

By the pattern you mean you have something like a 2D grid, with every on every grid cell the average of pressures at that point during the summers 2007-2012. Correct?

Now you want to do something with "the presence of this pattern" in an actual pressure pattern. Even if I knew what that was, is it not the wrong way around looking for the presence of an average (which is a summation of a number of actual patterns) in an individual one?

I'm not sure correlation is really the right tool for the job. I've been digging around and EOFs seemed hopeful, but they seemed to be mainly used for finding patterns, not deducing the strength of a standard pattern in data.

EOFs like in PCA are used to establish the MEI index, and indicator for the El Nino strength. IIRC MEI uses a dozen or so inputs, not a full grid though. 


In producing the AO Index the NOAA use a 'loading pattern' and project observed SLP onto it so as to get a numerical index of the strength of the AO pattern. That's the sort of thing I will be doing, and if I draw a blank I'll email them to ask how they 'project'.

I thought AO was just a simple pressure difference. Perhaps not, it could mean they are doing something like the MEI.

Not too helpful,  I am afraid.


ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: Home brew AMSR2 extent & area calculation
« Reply #121 on: July 08, 2013, 08:28:30 PM »
Thanks anyway, just to fill in some gaps.

Here is the average pressure field for JJA 2007 to 2012.


The question is: To what degree is this field present in the years since 1980 (IIRC when the data I have access to starts from). I know that it's not visibly present pre 2007, but is strongly present since then. On a monthly basis just by correlating 2d matrices of sea level pressure (the numeric version of NCEP/NCAR pressure anomaly plots) it's present in all June July Aug since 2007 apart from July 2010 and June 2013.

The NAO is the difference between pressure over the Azores and Iceland (IIRC it's normalised to the SD). But the AO index is constructed by projecting the daily (00Z) 1000mb height anomalies poleward of 20°N onto the loading pattern of the AO.
http://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/ao.shtml

The reason I've suddenly veered onto this is that the pattern shown in the image above plays a strong role in the post 2007 sea ice behaviour, Overland et al 2012 find it generates a dipole: low pressure along Siberia and high over the pack to Greenland. The failure of this pattern so far this year means this is a perfect opportunity to dig around to try to find out the cause. It's an opportunity I just can't pass up.

To do that I need a timeseries (down to daily) from which to begin to establish correlations.

I'm going to email the people who do the AO index.

BornFromTheVoid

  • ASIF Middle Class
  • Posts: 930
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #122 on: July 08, 2013, 08:57:18 PM »
Chris,

I worked on something similar, but more more simplified, thing to yourself for my undergraduate dissertation last year.
Related to the dipole, I created an index around the Fram Strait using NCEP/NCAR reanalysis data SLP for two pairs of points, seen in the image below.



I figured that to enhance sea ice export, you need the air flowing towards Fram, and from the strait itself toward the Greenland sea, helping to reduce summer sea ice extent (my understanding has come along some way since then!).

Anyway, the index worked off the monthly data, and whereby if the pressure on the western points are higher (lower) than the east, you'd get enhanced (subdued) summer export and reduced (increased) extent. So basically a pressure difference index.
Here's what I got for the Summer timeseries.



The 2007-2012 period had the highest 6 year mean for the index for June also, with this year being the first below average June since 2005.

I suspect it would be closely related to the pressure pattern your working on.

EDIT: Also, I think you'd find it hard to differentiate your pattern from the NAO loading pattern http://www.cpc.ncep.noaa.gov/products/precip/CWlink/pna/nao_loading.html
« Last Edit: July 08, 2013, 09:02:41 PM by BornFromTheVoid »

Neven

  • Administrator
  • ASIF Governor
  • *****
  • Posts: 3272
    • View Profile
    • Arctic Sea Ice Blog
Re: Home brew AMSR2 extent & area calculation
« Reply #123 on: July 08, 2013, 09:49:47 PM »
Chris, when I have time later this week I want to download weekly SLP maps from the daily composite page, from May to September, and put those on a special page on the ASIG (a bit like SIC map comparison page). Depending on how much space I have, I'm planning to do all the years since 2005 (or at least since 2007).

Maybe that'll be useful to you. It will be for me, but I'm more into the weather forecast side of things. Unfortunately, I lack the skills to quantify.
Il faut cultiver notre jardin

anonymous

  • Guest
Re: Home brew AMSR2 extent & area calculation
« Reply #124 on: July 08, 2013, 10:09:56 PM »
Chris, you can access most GFS data (wind, pressure, temps) between Mar 2004 and Jan 2013 with an url like this:

:
http://nomads.ncdc.noaa.gov/dods/NCEP_GFS_ANALYSIS/analysis_complete.ascii?pressfc[30:1:33][180][0]
The scheme after the question marks is param[time][latitude][longitude]. Example gives you pressure (pa) for four points in time with a distance of 6 hours for the North pole. Excel should be able to reformat the result to your needs, GSpreadsheet can.

:
pressfc, [4][1][1]
[0][0], 100517.0

[1][0], 100512.0

[2][0], 100399.0

[3][0], 100027.0


time, [4]
731649.5, 731649.75, 731650.0, 731650.25
lat, [1]
90.0
lon, [1]
0.0

Description. The nomad servers just migrated to new hardware, expect some hiccups.

Wipneus, feel free to move this to dev.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: Home brew AMSR2 extent & area calculation
« Reply #125 on: July 08, 2013, 10:13:25 PM »
Born from the Void,

Fascinating, a good approach. What I'm trying to do is to capture the wider pattern, so I've already ditched the idea of a weighting type approach.

Neven,

Go ahead, this is really just the continuation of what started with my original Summer Daze post back in 2011.
http://dosbat.blogspot.co.uk/2011/09/summer-daze.html


Both,

For what it's worth, using the Excel Correl function for north of 30degN I get the following.

2000   -0.158
2001   -0.218
2002   -0.418
2003   -0.185
2004   +0.169
2005   +0.531
2006   -0.539
2007   +0.845
2008   +0.845
2009   +0.833
2010   +0.439
2011   +0.893
2012   +0.631

If I 'window' in on the immediate region of the pattern (around 3000km away from Greenland) I should get even greater correlations post 2007. It seems to work, but that's no guarantee it's the best method, or even a sound method. I've been messing around with artificial datasets, and have read a couple of primers on PCAs, which are the simpler cousins of EOFs. If I can't get to grips with a better technique I may just use correlation in the interim until I can do better.

If only I weren't such a math dunce, what I've read on PCAs is very daunting.

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: Home brew AMSR2 extent & area calculation
« Reply #126 on: July 08, 2013, 10:17:16 PM »
Arcticio,

I've been using the text download link on the NCEP/NCAR pages I normally access. But discovered last night that the numbers change each time I access!

Anyway all I'm doing is getting a feel for the problem at present. I can get the base data at daily resolution here:
http://www.esrl.noaa.gov/psd/data/gridded/tables/surface.html

If people want to discuss this further I can do a new thread to stop spamming Wipneus's thread with off topic stuff. But not tonight.

crandles

  • ASIF Upper Class
  • Posts: 1594
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #127 on: July 08, 2013, 10:18:01 PM »
Chris, that pattern looks quite easy to describe - a high centred on mid Greenland.

You could calculate the ring around centre of Greenland that has lowest pressure deepest blue and measure pressure difference between central Greenland and that ring.

That probably isn't all that you want. It shouldn't be hard to create a measure being the proportion of points, which compared to surrounding 8? points, obey the rule that the further from the centre, the lower the pressure should be.

I don't know if having two measures like that is a major problem compared to having 1 index or whether it would be good to have a 'how-well-placed' measure and a strength measure. You may well be better with expert advice rather than listening to me.

anonymous

  • Guest
Re: Home brew AMSR2 extent & area calculation
« Reply #128 on: July 08, 2013, 11:57:29 PM »
Chris, never made it through that link jungle. But wonder why you use pressure as proxy for wind as proxy for drift?

ChrisReynolds

  • ASIF Upper Class
  • Posts: 1711
    • View Profile
    • Dosbat
Re: Home brew AMSR2 extent & area calculation
« Reply #129 on: July 09, 2013, 07:40:20 AM »
Arcticio,

I'm not using the post 2007 pattern for anything, it just represents a post 2007 change in NH circulation and needs investigating.

Crandles,

I have pondered doing a sum of differences at specific points/regions, e.g. Siberian coastal region, US Eastern Seaboard to UK. But I'd rather use the whole pattern.

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #130 on: July 09, 2013, 08:24:40 AM »
Update 20130708.

Extent: -143k6 (+258k vs 2012)
Area: -146k5 (+329k vs 2012)

All regions are now going down without extreme fast runners. The whole Siberian side shows solid decline. The CAA that did show increases for the last days, has slowly returned to melting.

The details:

Extent:
           Arctic Basin       East Siberian Sea              Laptev Sea
                  -12.6                   -11.1                   -18.8
               Kara Sea             Barents Sea           Greenland Sea
                  -19.2                    -0.8                    -9.0
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                  -27.7                     0.0                   -19.2
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   -4.2                   -11.9                    -6.6
             Bering Sea          Sea of Okhotsk            Total Extent
                   -0.4                    -1.9                  -143.6

Area:
           Arctic Basin       East Siberian Sea              Laptev Sea
                    0.5                   -20.7                   -20.7
               Kara Sea             Barents Sea           Greenland Sea
                  -20.4                    -0.5                   -21.8
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                  -23.7                     0.0                     3.5
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                  -28.9                    -4.8                    -7.9
             Bering Sea          Sea of Okhotsk              Total Area
                   -0.2                    -0.9                  -146.5

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #131 on: July 09, 2013, 08:42:23 AM »
To days image, the Canadian Archipelago. The break offs in the Nares Strait are clearly on the move: blue in the front, red in the wake. Cracks don't show.

(log in to see attached image)

Artful Dodger

  • ASIF Citizen
  • Posts: 436
  • The era of procrastination is coming to a close.
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #132 on: July 09, 2013, 10:52:55 AM »
To days image, the Canadian Archipelago. The break offs in the Nares Strait are clearly on the move: blue in the front, red in the wake.
Hi Wipneus,

This is awesome stuff!  8)

Any chance of creating a "2-week delta" map of sea ice loss over the whole Arctic?

EDIT: BAH! I see it's already done!! Wonderful, fellows  8)


I think that'd be an awesome contribution to Neven's bi-weekly sea ice updates over at the 'blog  ;)

And it would really put the cumulative effect of daily losses into perspective!

Thanks again for your excellent work with this project
« Last Edit: July 09, 2013, 12:54:42 PM by Artful Dodger »
Cheers!
Lodger

danp

  • ASIF Lurker
  • Posts: 66
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #133 on: July 10, 2013, 02:48:27 AM »
To days image, the Canadian Archipelago. The break offs in the Nares Strait are clearly on the move: blue in the front, red in the wake. Cracks don't show.

(log in to see attached image)

As some visible light confirmation of the melt pattern the microwave maps are showing, here's the last 5 days for one of the northernmost red-marked melt spots, between Melville & Bathurst Islands if I've got my geography correct.  The images are extracted from MODIS Terra swaths, ch. 1+4+3, 500m resolution.  You can see both melting in place and small floes breaking off along the rotten edge of the hole.

You can also see how deep blue, cracked, and ready-to-melt the ice elsewhere in the vicinity has looked for quite a while now.  But the visible light also agrees with the microwave-derived maps that no significant melt ponds or holes have opened up.   It will be exciting as more of it starts to go!


Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #134 on: July 10, 2013, 07:29:42 AM »
Thanks Dan, that is fun. Here is the amsr2 3.125km image of that spot. I had to rotate 45deg to get the same orientation, causing a slight blur (North is to the right in the image now).

 

danp

  • ASIF Lurker
  • Posts: 66
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #135 on: July 10, 2013, 08:55:26 AM »
Thanks Dan, that is fun. Here is the amsr2 3.125km image of that spot. I had to rotate 45deg to get the same orientation, causing a slight blur (North is to the right in the image now).

 

Thanks for the close-up Wipneus - very helpful.  I should probably figure out how to get my land mask working properly to facilitate alignment like this.

Looking closer I realize a couple things:
1. AMSR2 is doing really well at picking out the open water boundaries and the minute changes around those edges.

2. I was overly focused on the concentration changes, rather than the actual concentrations marked.  On this AMSR2 has clearly been confused into categorizing the solid but blue ice as <100% concentration: note the large greyish-labelled areas that are still clearly frozen in visible light. 

In addition I misread one of those red patches as being a recent opening of the hole, but in fact both of the patches in the upper right of the closeup are still solid blue ice as of my latest imagery from the 9th (though my processing may be running slightly behind as I'm always waiting on downloads).  Still I suspect that in the blue ice areas the microwave may be showing some combination of noise near its discrimination threshold, and real thinning in the ice which does not yet represent opening water.

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #136 on: July 10, 2013, 09:00:48 AM »
20 minutes later than we are used to, the data from U-Hamburg came in.

Extent: -127k3 (+218k vs 2012)
Area: -126k6 (+293k  vs 2012)

All Siberian regions from Kara to Chukchi are contributing to the decline. Beaufort seems to start as well. The ice in the Greenland sea continues to disappear.
Although Hudson and Baffin still have some ice, they contribute less than a third of the net total decline.

The details:

Extent:
           Arctic Basin       East Siberian Sea              Laptev Sea
                   -3.9                    -9.7                   -11.5
               Kara Sea             Barents Sea           Greenland Sea
                  -18.8                     0.4                   -22.0
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   -6.8                     0.0                   -32.8
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   -3.2                   -10.0                    -8.1
             Bering Sea          Sea of Okhotsk            Total Extent
                   -0.2                    -0.7                  -127.3

Area:
           Arctic Basin       East Siberian Sea              Laptev Sea
                   16.6                   -36.6                   -15.7
               Kara Sea             Barents Sea           Greenland Sea
                   -5.5                     0.1                   -15.2
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   -7.9                     0.0                    -5.1
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                  -22.5                   -18.7                   -15.8
             Bering Sea          Sea of Okhotsk              Total Area
                   -0.1                    -0.2                  -126.6

 

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #137 on: July 10, 2013, 09:27:22 AM »
Beaufort is region of the day. Polynya are more extended today as not before this season.

(log in to see the attached image)

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #138 on: July 10, 2013, 09:54:14 AM »

2. I was overly focused on the concentration changes, rather than the actual concentrations marked.  On this AMSR2 has clearly been confused into categorizing the solid but blue ice as <100% concentration: note the large greyish-labelled areas that are still clearly frozen in visible light. 


Just to make sure, it is me that marks pixels with concentration going from over 15% to below 15% red, and from below to over 15% blue. Pixels that stay below 15% or above 15% are grey (black - white).

Be aware as well, that the sea ice concentration algorithms are optimized for sea and ice. Pixels near coasts will have some contamination introduced from the microwave emissions from land surface. Pixels that are surrounded by "land pixels" will be affected more. In most cases this translates into more ice than reality. I use some logic to remove such ice, this works quite well over the whole Arctic but maybe not in every little corner.

Attached, a "raw" image of that area.

« Last Edit: July 10, 2013, 10:15:52 AM by Wipneus »

danp

  • ASIF Lurker
  • Posts: 66
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #139 on: July 10, 2013, 11:46:10 AM »

Be aware as well, that the sea ice concentration algorithms are optimized for sea and ice. Pixels near coasts will have some contamination introduced from the microwave emissions from land surface. Pixels that are surrounded by "land pixels" will be affected more. In most cases this translates into more ice than reality. I use some logic to remove such ice, this works quite well over the whole Arctic but maybe not in every little corner.

Attached, a "raw" image of that area.
This is not the issue I'm noticing though.  In your raw image, there are already lots of deep grey areas that the microwave sensor algorithm believes are between 15-100% concentration.  But the visible pictures between clouds just show expanses of cracked but unbroken deep blue ice, which I'm suggesting is fooling the sensors.

Peter Ellis

  • ASIF Middle Class
  • Posts: 501
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #140 on: July 10, 2013, 12:17:14 PM »
Yup, some of the ice seems to be thin enough (or wet enough on top, or both) that the sensor sees straight through it.  It'll be gone soon though.

Neven

  • Administrator
  • ASIF Governor
  • *****
  • Posts: 3272
    • View Profile
    • Arctic Sea Ice Blog
Re: Home brew AMSR2 extent & area calculation
« Reply #141 on: July 10, 2013, 12:32:10 PM »
Wipneus, this has become my favourite corner of the Internet.
Il faut cultiver notre jardin

BornFromTheVoid

  • ASIF Middle Class
  • Posts: 930
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #142 on: July 10, 2013, 12:44:50 PM »
Wipneus,

Would it be possible to take the MODIS Arctic mosaic, then overlay your pixel change images with enough transparency to see both the changed pixels colour and the visible ice below?

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #143 on: July 11, 2013, 08:24:37 AM »
Update for 20130710.

Extent: -99k7 (+361k vs 2012)
Area: -136k6 (+332k vs 2012)

This decline is strong as most of the decline is now within the Arctic itself: the peripherals Hudson, Baffin and Greenland have, taken together, almost no change. ESS and Beaufort really seem to  take off now.


The details (in 1000 km2):

Extent:
           Arctic Basin       East Siberian Sea              Laptev Sea
                   -9.9                   -18.1                   -14.0
               Kara Sea             Barents Sea           Greenland Sea
                  -14.1                    -0.5                     7.2
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   -2.2                     0.0                    -6.2
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                  -15.3                   -14.7                   -14.1
             Bering Sea          Sea of Okhotsk            Total Extent
                    0.3                     1.8                   -99.7

Area:
           Arctic Basin       East Siberian Sea              Laptev Sea
                  -24.0                   -46.5                   -20.3
               Kara Sea             Barents Sea           Greenland Sea
                   -4.8                    -0.5                     8.8
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   -3.0                     0.0                    -4.5
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   -4.3                   -27.0                   -12.7
             Bering Sea          Sea of Okhotsk              Total Area
                    0.2                     2.0                  -136.6

Artful Dodger

  • ASIF Citizen
  • Posts: 436
  • The era of procrastination is coming to a close.
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #144 on: July 11, 2013, 08:36:03 AM »
Hi Wipneus,

It looks like sea ice transport has restarted in the Greenland sea. Dipole, anyone?
Cheers!
Lodger

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #145 on: July 11, 2013, 08:43:23 AM »
Regions of the day are Kara and Barentsz. Barentsz is now officially and completely ice free.
Kara, at this rate, will be in around 4 weeks. The decline of the CAB is almost all captured in this image.

(log in to see attached image)


Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #146 on: July 11, 2013, 08:57:40 AM »
Lodger:

Most of the decline in the CAA is in the western corner (between Banks and Victoria islands and the continent). See yesterday's image of the day.

Artful Dodger

  • ASIF Citizen
  • Posts: 436
  • The era of procrastination is coming to a close.
    • View Profile
Re: Home brew AMSR2 extent & area calculation
« Reply #147 on: July 11, 2013, 09:56:08 AM »
Lodger:

Most of the decline in the CAA is in the western corner (between Banks and Victoria islands and the continent). See yesterday's image of the day.
Hi Wipneus,

Wonderful charts. But with both Area and Extent up in the Greenland sea, and with high SSTs in the region, we can say with confidence that that sea ice arrived from the Central Basin.
Cheers!
Lodger

Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #148 on: July 11, 2013, 10:50:05 AM »
Ah, yes. I misread Greenland for Canada some how.

The Greenland sea ice is starting to move again (still melting though). If this continues a couple of days, I will dig up my animation skills again.


Wipneus

  • ASIF Governor
  • Posts: 3147
    • View Profile
    • Arctische Pinguin
Re: Home brew AMSR2 extent & area calculation
« Reply #149 on: July 12, 2013, 08:25:44 AM »
Update 20130711.

Extent:-87k3  (+449k vs 2012)
Area:-163k2  (+300k vs 2012)

The "torch" seems to be on in the Chukchi sea. Strong declines on the Siberian side, less in Beaufort and CAA. Area is outpacing extent esp. in CAB, ESS and CAA.


The details (in 1000 km2):

Extent:
           Arctic Basin       East Siberian Sea              Laptev Sea
                   -4.6                   -12.1                   -16.9
               Kara Sea             Barents Sea           Greenland Sea
                  -16.8                     0.3                     9.7
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                  -14.1                     0.0                    -2.0
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   -2.8                    -4.2                   -22.5
             Bering Sea          Sea of Okhotsk            Total Extent
                    0.5                    -1.9                   -87.3

Area:
           Arctic Basin       East Siberian Sea              Laptev Sea
                  -65.1                   -27.1                   -17.3
               Kara Sea             Barents Sea           Greenland Sea
                  -18.8                     0.2                    -2.3
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                   -9.1                     0.0                     1.6
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                  -10.5                     3.2                   -16.5
             Bering Sea          Sea of Okhotsk              Total Area
                    0.5                    -2.0                  -163.2