Support the Arctic Sea Ice Forum and Blog

Show Posts

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.

Messages - danp

Pages: [1] 2
Arctic sea ice / Re: Home brew AMSR2 extent & area calculation
« on: August 23, 2013, 04:58:20 PM »
Thanks Artful Dodger!

Couldn't you use the 4 km 3-6-7 daily data to select the 250 m data to download? Then you'd only be downloading large files that you will actually use.

I've thought of doing something like that, although my clear-swath algorithm requires a lot more data than just the 3-6-7.  In general the best data series for my purposes has turned out to be the level-2 MOD09 swaths.  These include *all* MODIS channels at their top resolutions, and there's no subsampled version.  The reason I chose this series was that the extra processing that the NASA land team does beyond the 1B files to remove haze, correct for sun angle and atmospheric reflection, and calculate absolute surface reflectance, was quite effective and beyond my ability to reproduce without tremendous work.  Even though those efforts aren't all well-optimized for the Arctic, they're a major help since the sun angle is always so low.

I've needed to use all of this good-quality data to assess my algorithm's selection abilities, as well for as a couple of other so-far unsuccessful projects.  Once everything about the method of producing this data is settled, I can reassess the ongoing downloads required, but for now it's an easy if inconvenient task.  I suppose I should have said the true bottleneck is my programming time :)

Arctic sea ice / Re: Home brew AMSR2 extent & area calculation
« on: August 23, 2013, 01:04:42 AM »
Hi Werther,

I was hoping a long-time ice-watcher like you would find this presentation useful.  I have considered processing all the way to the top MODIS resolutions (250m, compared with my 500m), so I will take that as a positive vote.  It would take a good deal longer and processing time is a bit long right now, but I haven't optimized much for that yet.  As far as my main constraint (download bandwidth) I already am downloading the appropriate data to do so; it is only channels 1 & 2 that have the top resolution, and the top resolution data (for all 37 channels!) is there in the swath downloads, so I could do what NASA does and resample the lower-resolution channels to match. 

I agree the blocked nature of the presentation means neighboring tiles can have discontinuities in ice position.  The typical dispersion between tiles in an 8-day composite might be about 4 days, and when I'm manually clear-image hunting I often wind up comparing MODIS images from that far apart anyway, so I'm already resigned to that resolving those kinds of time discontinuities when they are presented obviously.  In any case it is that is far less objectionable than the pixel-by-pixel smearing that NASA's composites have.  Of course I could easily go the other direction and process with larger tile sizes for a more coherent picture, but at a sacrifice of clearness, since I wouldn't be able to remove as many clouds.  NASA's ordinary rapid-response mosaics aren't trying to remove clouds, so they have a lot fewer objectionable seams (except at the pole) since their effective block size is usually pretty large.

It may well be that there are different optimal choices for different cloud conditions and compositing time lengths, but I do want to settle on something so I can start back-processing data from this year and previous years.

In any case this product is very much in an "alpha" or "beta" version, so any comments you have are appreciated! 

Arctic sea ice / Re: Arctic Image of the Day
« on: August 22, 2013, 10:17:37 PM »
Very likely it was a small section of a linear seam between two adjacent swaths that were stitched to make the mosaic.  Then the jpg compression produced the extra noise as Espen said.

When I produce mosaics there are usually small sections of 1-2 pixel seams that result from choices the reprojection software makes about where to place border pixels, and I have to intentionally remove them by filling with average neighbor values.  You can still see remnants of them on my 8-day composites because I use a light hand on the filling procedure so that larger data dropouts aren't obscured.

Arctic background / Re: USCGC Healy: scientific mission to the Arctic
« on: August 22, 2013, 09:43:17 PM »
Healy is heading east again near the coast (and encountering some unpleasant-looking rainy weather).  It looks as though they're about in position to do the 4th most easterly of their transects according to the map at

In the northernmost two of the remaining segments of the cruise they should encounter sea ice that's been pretty consistently under cloud, so it'll be nice to see it.  The opening of McClure strait in particular should still have plenty of ice whenever they get there.

Arctic sea ice / Re: Home brew AMSR2 extent & area calculation
« on: August 22, 2013, 05:16:59 PM »
I mentioned over at
that I'm at a beta version on a new method for compositing any time length of MODIS images to select cloud-free scenes.  Here's a crop from the most recent 8 days, which have been cloudy enough to still challenge my algorithm:

(click to see full size, about 1700 pix across)

Arctic sea ice / Re: When will we see open ocean at the north pole?
« on: August 21, 2013, 11:26:20 AM »
Nice Pmt!  Here's a grayscale ch. 1 grab again from swath day 232 (Aug 20), 0500 UT:

5 km (20 pixel) diameter circle this time.

Santa has got to be pretty uncomfortable on that rubble.

Arctic sea ice / Re: The Cause of the Muted Melt of 2013
« on: August 18, 2013, 12:41:36 AM »

Please note the low density and particle size. The particles are 1/10th to 1/100th the size of dust particles and 1/10 the density. The subsequent 50-600m^2/g surface to mass ratio is huge. 10,000,000,000g would have a surface area in 500,000 to 6,000,000 km^2 range.

Surely the total scattering cross section depends on the 2-dimensional projected area, though, not the total surface area.  The main question is whether the light scatters 0 or 1 times on the way down, not whether it intersects some facet of the grain while bouncing around in the atmosphere. 

At such low densities you could ignore particle overlap and the projected area would be the area you'd get if you put all the grains on the ground and swept them into a 1-grain deep solid configuration.  That area is proportional to the total volume/thickness, or equivalently the mass / (density * radius).  So while using your size estimates might indeed raise the scattering power 100 to 1000-fold over typical dust, that still doesn't bring them near a typical volcano.  Also, I'm ignorant of volcanoes but I wouldn't be surprised if their ash spectrum already included these tiny particles in some amounts as well.


Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: August 18, 2013, 12:26:05 AM »
Captions are easy and helpful, and I've been lazy not to have included them on all my animations; I just need to add that consistently to the snippets of code to spit them out.

A simple viewer: yeah, it would be nice.  Animated gifs are ancient, bulky, and silly.  FWIW, you can always open the animated gifs in most image editors and page forward/back for a similar effect. But that's clunky and the the captions would be a help, wouldn't they!

Arctic sea ice / Re: When will we see open ocean at the north pole?
« on: August 17, 2013, 09:56:07 PM »
Thanks Peter!

Since my processing of this summer has obviously been cherry-picked towards exciting events, I've been hoping to find time to go back and find analogous events in the past for comparison.  I will start digging through swaths around that date to find the most open water right then and see what the n pole looks like, at least to have something to compare this year to.

Ideally I could use those 8-day composites (once I fill in past years) for fair comparisons of things like this but to be honest I find the small-scale detail over ice to be pretty bad.  I believe it comes from a combination of reprojection distortion and that the algorithms are optimized for land where features are not changing on a daily basis.


Arctic sea ice / Re: The Cause of the Muted Melt of 2013
« on: August 17, 2013, 09:46:23 PM »
Yeah, I'm pretty skeptical too.  Hundreds of tons just isn't that much, and the original press release at

also indicates what's impressive here was detecting it at all.  Compared to e.g. the 2010 eruption of Eyjafjallajökull, I can't imagine it made much of a local climate blip at all.  Was there a big Arctic impact from those eruptions?  Certainly it didn't show up in an obvious way in that summer's melt.

Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: August 17, 2013, 04:07:08 PM »
Is it ok to share the image?


Sure!  Spread them around, etc.  It would probably be useful to credit NASA/MODIS along with me and/or the forum for people who want to see where they come from and keep up to date.  I suppose I could just add a small caption to the mosaics.


Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: August 17, 2013, 08:31:21 AM »
Hot off the presses, our octodaily images return with the day 217 mosaic.  Huge changes are visible since the previous mosaic, with the wide stretches of low concentration straight through the pole finally showing up.  Here's the channel 7-2-1 blink comparison with day 209, and the 1-4-3 and 7-2-1 current thumbnails.  As always full-size versions at

Arctic sea ice / Re: When will we see open ocean at the north pole?
« on: August 17, 2013, 05:18:09 AM »

That is a powerful image, thanks for posting.
To give the size scale, how many metres or km across does the green circle correspond to?

Thanks ... it is kind of growing in impact as I look at it.  The resolution is 250m/pixel and the circle has an 8 pixel radius, so 2km radius / 4km diameter.

I'll have to add my thanks Espen.  I've quickly gained a love of Greenland's geography despite my recent near ignorance, thanks to following along this forum, and especially owing to your wonderful satellite-based mini-travelogues.

Arctic background / Re: Icebreaker Oden
« on: August 17, 2013, 01:44:04 AM »
Nice find Vergent.  And it sounds like some very important research on ice quality, even if we don't manage to get real-time updates from the trip:

"Among the most interesting data from the moorings is that concerning the frequency and thickness of ice ridges. ...

Ground-truthing remote sensing

The researchers will also employ a variety of technologies to quantify the ice around the ship and icebergs that are in the vicinity of the ship. In addition to using electromagnetic sensors to continuously measure the ice thickness around the ship itself, Lubbad says the researchers will use near real-time satellite images to look for icebergs and then ground-truth the icebergs that they identify.

In this case, “ground-truthing” an iceberg means visiting it by helicopter and using a remotely piloted aircraft system (RPAS) to make a three-dimensional image of the iceberg, including its size and thickness. At the same time, researchers will use a remotely operated underwater vehicle (ROV) to collect the same information about the underside of the iceberg."

Arctic sea ice / Re: When will we see open ocean at the north pole?
« on: August 17, 2013, 01:24:09 AM »
Usually I don't process down to 250m resolution; it's only channels 1 & 2 that have that resolution, so the other bands are upsampled to match in the composite images, and the pictures are big enough even at 500m resolution.  But for fun, here's the 250m ch. 1 grayscale crop from day 227 (yesterday), 1930 UT, which had the clearest image:

Arctic sea ice / Re: When will we see open ocean at the north pole?
« on: August 17, 2013, 12:36:47 AM »
Here's an animation covering the pole itself (marked w/ a circle) over the last 36 hours.  As you can see there's a stubborn ice floe I'm calling "Santa's last stand" that's stayed put over the pole during that time, even while polynyas open around and the neighboring ice has been mobile.


Arctic sea ice / Re: When will we see open ocean at the north pole?
« on: August 16, 2013, 05:50:23 AM »
Thanks for the heads-up Juan! 

By the way, those animations come about from downloading really huge amounts of data in .hdf format ... when I'm up-to-date on my downloads, I can just spit them out with a single script (and then a little manual removing of cloudy frames), but it requires devoting enough bandwidth for a while to suck down the swath data.  I'll go ahead and look again when I've gotten my data up to date.

Technically one could get the analogous pictures from the jpgs at

but I am not generally patient enough to search through these.  And then aligning the different images would be an unpleasant task, and you'd have more artifacts, etc.


Arctic background / Re: USCGC Healy: scientific mission to the Arctic
« on: August 15, 2013, 05:01:41 PM »
And Healy turned south again, back nearly to the shore.

I think I saw their summer route plans somewhere on their website. Looked like the plan was a series of short transects perpendicular to the northern Alaska coast. Didn't look like they were going to go north of 74N this year.

You're right - although it's a little ambiguous I believe the map here:

refers only to this mission, Healy 1302, running through Sept. 2.  Googling the PI's name gave this form:

which decribes the purpose of the mission as "Survey seafloor locations between Barrow Canyon, Mackenzie Delta, Amundsen Gulf, shelf and slope of Banks Island, and M'Clure Strait using a towed, high-resolution CHIRP sonar.  ...   Major cruise objective is to find geological evidence of a massive flood that came north to the Arctic via the Mackenzie River about 13,000 years ago."  And the map matches that itinerary.

Greenland and Arctic Circle / Re: The Nares Strait thread
« on: August 14, 2013, 05:18:32 PM »
It's actually moved quite a lot over the past week or so after being stuck for so long.  Here's an animation from day 216-224, with all clear-ish swaths picked out.  500m resolution, MODIS Terra:

And here's a sub-animation with 7 clear frames showing distinct positions of the island:

Arctic sea ice / Re: When will we see open ocean at the north pole?
« on: August 14, 2013, 02:38:44 AM »
Still pretty cloudy over the pole but there was some clearing this weekend so we can see the broken-up ice around the pole.  Definitely some occasional swimming possible with holes as big as a few km opening between floes immediately around the pole:

36-hour animation, processed from MODIS/TERRA swaths days 222-223, 500m resolution.  Orientation is Greenland-down, and N pole is marked with a circle.

Arctic background / Re: USCGC Healy: scientific mission to the Arctic
« on: August 12, 2013, 10:53:18 PM »
There's a big arm of the ice pack sticking out to the south around longitude 162-165 west, so if the Healy continues wandering NW it should intercept that.  Perhaps the Healy's heading that way ... there's been some discussion about whether that ice will manage to detach from the main pack, so it would be interesting to see what it looks like.

Healy's position currently marked with arrow:

Arctic sea ice / Re: Arctic Image of the Day
« on: August 11, 2013, 05:11:45 PM »
Nice!  AMSR2 gives the pack edge there as 75° N.  If there aren't any more detours, they should reach it tomorrow.  It'll be nice to get a real look at the ice.


Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: August 10, 2013, 03:13:55 AM »
Day 209 (i.e., days 209-216) mosaic is up.  Here's a comparison with the previous mosaic.  Note that due to the usual time lag the most recent storm hadn't started up yet by the end date of the latest mosaic.

I also uploaded 1-4-3 and 7-2-1 mosaics going all the way back to day 145 of this year.
( )

Here's an animation including all of these mosaics, which started May 25th.

Attached are thumbnails of the most recent mosaics as usual.

BFTV: The images are from a video of the 2007 ice pack. I discussed in this comment last August
I think that must be synthetic aperture radar, probably RADARSAT or ENVISAT.  It's such a shame that the data is generally unavailable even to researchers without paying and going through lots of red tape.  Makes me appreciate the great volumes of unwieldy data we do get from e.g. MODIS.

Arctic sea ice / Re: Arctic Image of the Day
« on: August 05, 2013, 07:48:14 AM »
Hah!  well spotted!

Arctic background / Re: USCGC Healy: scientific mission to the Arctic
« on: August 04, 2013, 09:13:55 PM »
Took a sharp turn southeast today and found some more varied ice conditions, with plenty of open water but a few isolated dirty chunks of thicker ridged ice like this:

Arctic sea ice / Re: Predictions
« on: August 03, 2013, 09:46:44 PM »
What are the dates of those two Topaz plots from 2012 (i.e., are they the model's current best guess for what the actual thickness was at the time?)  I am not familiar with Topaz and wondering if it has a tendency e.g. to be aggressive about melt predictions, in which case it would be interesting to see how the modeled predictions for those 2012 dates changed as the actual dates approached.

Basically I find that comparison as shocking as Dromicosuchus and so am looking about for some reason for skepticism.

Arctic background / Re: USCGC Healy: scientific mission to the Arctic
« on: August 02, 2013, 07:19:19 PM »
Last year had really unusual conditions in the late spring/early summer for the Bering, though ... I wonder if that ice was just the fast-melting detritus left over from all that.  Naturally its presence would have been keeping down the surface water temperature too.  But clearly the Beaufort to the east was in terrible shape last year.  Still, looks like there was a heat sink in this area last year we don't have this year.

Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: August 02, 2013, 06:14:59 AM »

Thanks for that. Eastern CAB not ice free on Aug 1(as I predicted(my bad)). But it will be soon.


Cheer up!  The mosaic only covers July 20-27 :)  (there is always a lag before the 8-day tiles are uploaded, and it was a little longer this time).  And since the algorithm picks the clearest tile with no preference for recentness, the average date will be in the middle of that period. So really the next mosaic will have an effective center date around Aug. 1st.

But it's true that the up-to-date MODIS images also show that ice sticking around stubbornly.  Heck, being wrong (and remembering) is good - I've mostly been too confused this year even to get that far.

Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: August 02, 2013, 04:18:25 AM »
The latest mosaics are processed, day 201-208.  This one was the cloudiest one yet, so if you're hole-hunting the 721 image is particularly useful for avoiding being fooled by persistent clouds.

Here's an animation of the three 7-2-1 mosaics:

You can see some big changes in the East siberian sea and holes opening up in the west central Arctic.

I also attached overviews of the latest 143 and 721 mosaics.  As always you can find the full-size versions at

and they will be up in the split zoom at arcticio soon.

Arctic sea ice / Re: Arctic Image of the Day
« on: August 01, 2013, 11:04:11 PM »
The take-home message I get from that is just how heterogeneous sea ice is.  Each one of those crazy white lines represents a ridge, which can be up to ten times the thickness of the flat grey/brown areas in between.  ~50% of the entire Arctic ice volume is in the ridges.  Eyeballing that picture, however good the resolution, isn't going to tell us that much about how much ice is actually there.  I guess this is why the ice models used in PIPS (and PIOMAS etc) model ice as a statistical distribution within each pixel, rather than a single value per pixel.

As for your question, it would help if we had even one more picture from the same satellite to compare!  Danp, where did you get it?  Are there any pictures available for previous years, where the subsequent melt behaviour is known?

I totally agree.  It was eye-opening to see the crazy map of features at this detail level.  It may be asking a little much of this data source to find truly comparable images, but because of the built-in geographic restriction, if we can find any, they will at least be from nearly the same area of the southern Beaufort.  I only just decided to go digging.

The public data is available from REVERB (also one of my main sources for Terra/Aqua data), via geographic search.  Here's the search I did to find this data:

The dataset is "L1B Registered Radiance at the sensor" (not the expedited equivalent).   The visible/near IR bands within the hdf file are 1, 2, 3N, 3B.  I used 3N, 2, 1 as RGB, but you don't gain much over just a grayscale image from ch. 1, which is in the green.  (3B is an offset backward-looking channel for stereoscopic images). 

I just repeated the search (July 1-31) as a repeated annual search since 2000 and came up with 358 results, including the 5 from  this year (2 of which are right at the shoreline or just inland).  So it does look like it might be a plausible project to compare ice at the southern edge of the Beaufort year-to-year in this fashion.

Arctic sea ice / Re: Arctic Image of the Day
« on: August 01, 2013, 09:02:45 PM »
MODIS/ASTER image from July 12th, from a frame in the Beaufort sea maybe 200 km north of the Alaskan coast.  Cropped image is full-resolution, 15m/pixel; full image is an overview of the 60x90km frame. 

ASTER has a narrow field of view and thus doesn't have full coverage of the Earth like Terra/Aqua.  In addition, free data access is pretty restrictive except for the US & territories, where level 1B data is fully open.  [EDIT:  I realized that was confusing.  Data taken over the US & territories (apparently extending out at least a little into the waters around Alaska) is fully open-access to anyone in the world.  Other data you have to pay for, or else you have to jump through a bunch of hoops before you get it.]

This is one of a swath of 5 frames, the only set since July 1st in open access and within the Arctic.  Converted from hdf using GDAL and then some gimp to combine, etc.

Arctic sea ice / Re: W.Meier on Conservative Scientists
« on: July 31, 2013, 09:24:25 PM »

How much cloud coverage there’s actually been compared to other years?
What makes the difference between DMI +80dN and NCEP/NCAR data on 2m+SL Temperature?
Are SST’s really lower all through the Arctic Ocean?
Is the set-up more like the climatologic trend this year?
Is the supposed change in NH atmospheric pattern creating a positive feedback for the ice, while creating extremities in the mid-latitudes?

So many questions have risen this season.

Need time to analyse in the aftermath.
It’s not that obvious, you know… or maybe it is when the final numbers come in and some will say “…I told you”…, others “…nothing to see, just go on…”

I agree.  To me nothing looks obvious at all about analyzing this season, and while I am no Arctic expert I find it exhilarating as that's always the most scientifically exciting place to be.  Peter's observation about the Beaufort cracks may well be a highly important factor this season and it is not at all what I expected to happen.  It's wonderful to be surprised, particularly by something that may be positive news.  But as I understand it, regrowth around cracking involves much more detailed ice dynamics than the thinness->regrowth negative feedback the models the models have pushed, even though it's fair to put it in the same category.  After all, there were perfectly plausible physical reasons that many folks on the forum thought it would go the other way.

I disagree that there is an "anti-scientist" bent here.  There is disagreement about some findings by some scientists.  Science is a slow process - hypotheses, data gathering, analysis, review by peers, not to mention fundraising, all take time, and each step forward is only one piece of the puzzle, and one subject to further review at that.  I have nothing but the highest regard for scientists in general and those who are expanding our understanding of how AGW works in particular.

Unfortunately, global climate is changing faster than science can keep up.  I am sure it makes it an exciting time to be a climatologist.  But it may be frustrating for those seeking to nudge the policy debate not to have clearer answers now. It also raises the temptation (I am pointing the finger at myself!) to play armchair scientist.  There are many here, though, who have a lot to add to the debate, not all of whom hold a doctorate or have published work in peer-reviewed journals.

Exactly my feelings.  On these boards there is certainly the kind of arrogance that comes from stepping into someone else's field without having quite appreciated how much effort the trained scientists have devoted to the work they've produced so far.  As a scientist in an unrelated field I'm frankly humbled at how generous and open individual scientists have been with their time and energies when folks here have asked them for information about their work.

I also appreciate the scientists' resource limitations as well as the natural pressures that lead to conservatism when it comes to places where their best models are doing poorly.  The kind of messy, phenomenological, observational-based open science that keeps me coming to these forums is time-intensive and may never lead easily to well-defined models predictive models, but nonetheless I constantly find amazing insights from everyone here that complement my erratic learning from the published literature. 

There's no way to avoid people getting grumpy when it comes to prediction time:  there's a reason we all are here as a community, and it's not because the Arctic is just some abstract scientific system we don't understand.  Since Arctic ice is the vanguard of visible climate change and the poster child for possible excess conservatism in past scientific predictions, any predictions come along with a host of policy and political baggage that gains weight with the stature of the person delivering it.  I certainly have my moments of being upset with excess scientific reticence, and Meier's comments irked me for some of the reasons others commented on.

But in the end it's just a matter of time; the ice will do what it will do over the next few years, and we'll be learning quickly about our new Arctic.  Global warming is vastly depressing, but the Arctic is a subsystem in which I've found some scientific excitement as consolation.  So for the moment I'm just sticking to the old arctic sea ice blog mantra:  keep our eyes on the ice.

Arctic sea ice / Re: Melt Ponds!
« on: July 30, 2013, 04:38:58 AM »
The recent history of melt ponds seen by the pole-cams:


That's a nice summary.  Thanks!


Arctic sea ice / Re: Melt Ponds!
« on: July 30, 2013, 12:42:07 AM »
Can anyone outline or link to some analysis how this year compares to last year and the years before? How unusual is this melt there?  Is this happening every year, maybe this year just earlier or was this previously considered stable? Although are there projections which saw this development coming? Thanks.

The climate models have a poor record with regard to arctic ice.


I think prokaryotes was asking more specifically about the pattern of melt ponding, and how much we know about its long term trends.  I'm also interested if anyone knows; for example here's James Morison of the North Pole Environmental Observatory quoted as saying that he had seen much more extensive melt ponds (in an article that still fails to point out the webcam isn't at the pole):

Your point about any predictions drawn from the large-scale climate models is of course well-taken.


Arctic background / Re: USCGC Healy: scientific mission to the Arctic
« on: July 29, 2013, 10:43:59 PM »
Well, whatever trial icebreaking the Polar Star was doing is over for now, as it'll be back in Juneau for tours on Aug. 3.

"Coast Guard crewmembers will be standing by to answer questions about the 399-foot icebreaker and their most recent operation in the Arctic Ocean."  Now's your chance to get the inside scoop if you happen to be there :)

Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: July 29, 2013, 10:26:22 AM »
Funky interface but I think I'm getting the hang of it!  Great job arcticio and thank you so much for your work.  Now it's definitely my turn to get my scripts working to start adding in some more mosaics from the past for comparison.

Where should I start?  I was thinking of continuing back through the season to say June, and then starting to add in date-comparable mosaics from past years.  Any particular requests?  I'll probably stick to bands 143 and 721 as my current method only has me downloading the first 7 bands (the downloads are heavy enough as it is), and within that range I didn't find a band combination particularly superior to 721 for the ice right now.

P.S.  stay tuned ... there may be daily mosaics on the horizon!

Arctic sea ice / Re: Home brew AMSR2 extent & area calculation
« on: July 28, 2013, 01:06:48 AM »
4-day animation (through end of day 207, July 26) of Fram strait.  You can see the crumbling ice in the north.  In the full-size version (500m resolution, 9MB, linked below) you can see cracks forming and perhaps more imminent detachments of large sections of fast ice along the Greenland coast.

(overview of the whole image, along with a subset at full resolution:)

Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: July 25, 2013, 12:50:45 AM »
By the way, anyone who wants to download full-size single-images of these mosaics and the ones I will be producing in the future can find them in the folder at

They are about 4900x4900 pixels and 9 MB JPGs.  They are named by day # of first day in the mosaic, and band combination.

Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: July 25, 2013, 12:18:20 AM »
Thanks Neven! 

To back up what I said above about estimating the amount of open water in the east central arctic, here are snapshots of the band 7-2-1 mosaics for the same time periods.  You can see the mosaics are not as cloud-free as the visible 1-4-3 combination make them look, since most of the clouds are over white ice.

Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: July 24, 2013, 08:22:38 PM »
Apparently animated gifs don't play in the previews here (click to open the attachment to play it).  If you just want to see an overview of the latest image, here it is:

Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: July 24, 2013, 08:10:46 PM »
NASA has new 8-day averaged tiles out, so I produced a new mosaic!  I've attached a preview comparison with the earlier one.  Note that while NASA's averaging algorithm is making an attempt to pick out clear pixels, consistently cloudy areas will still wind up with smeared clouds in the result, which makes it a bit tough to judge the ice condition in the center of the pack from the overview (it's easier in the full size).  For example I doubt the amount of open water in the east Arctic basin has reduced; it's just that there was more consistent cloud there in the second period.  I will be producing mosaics in other band combinations that should clarify this kind of issue. has graciously offered host the full-sized images using the excellent zooming interface already in place there, so one of us will post a link to that shortly.

1st frame: July 4-11
2nd frame: July 12-19          (open attachment to play animation)

Greenland and Arctic Circle / Re: The Nares Strait thread
« on: July 20, 2013, 05:57:59 AM »
48 hours of motion, from processing ch. 1-4-3 of MODIS swaths.  Note the time stamps as the delay between frames is pretty irregular (it depends on the vagaries of satellite overpasses).  You can definitely see some currents or wind shoving ice back up against the southern edge of the jam, as well as the role of Peterman ice island forcing the ice around it.

Arctic sea ice / Re: Melt Ponds!
« on: July 19, 2013, 06:24:08 PM »

It's more like a melt lake, with an ice island

Verg ;-)

Beautiful catch there Vergent!  It's nice when nature delivers a metaphorical image all on its own.

Arctic background / Re: USCGC Healy: scientific mission to the Arctic
« on: July 19, 2013, 09:28:11 AM »
Some nice images and videos from the Coast Guard operations, including a flyover of the Polar Star on July 16th at the edge of the Chukchi sea:

The short videos are particularly fun, though without knowing exactly where they are you can't really say anything about the apparent state of the ice.

Arctic sea ice / Re: Home brew AMSR2 extent & area calculation
« on: July 19, 2013, 09:18:30 AM »
Polar Stern's sister ship Polar Star is somewhere up there out of Unalaska doing sea trials after her refit. But I haven't been able to find out where, or whether any obs will be made public.

Good find Anne!  Digging around I came across a little info here:

There are some photos from July 16th of the Polar Star at the edge of the ice in the Chuckchi sea.  There are also some other neat photos taken from various other Arctic overflights, instrument drops, etc.

Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: July 17, 2013, 06:15:18 PM »
Wow, impressive, danp!

Last year I wrote a series of posts called Peeking through the clouds. These revolved around MODIS Composites that are released every 7 days by Environment Canada. They make a True-Colour Composite image as well as a False-Colour Composite image.

One commenter called dabize would send me cleaned up - or "declouded" - versions of those MODIS False-Colour composite images.

That's right - I had completely forgotten about those posts and this data source.  It is basically the same thing as this 8-day mosaic although I don't think they've done as well with preserving the data's dynamic range.  I should probably focus on the daily mosaics then.  Unfortunately I had miscalculated the amount of the data download for the daily composite tiles to produce the mosaics: at 12 GB or so a day it seems to be more than downloading all the swaths that produced the composite, which I don't understand.  Perhaps I'll go back to my original DIY declouding technique since downloading all the swaths also allows animations. 

My own technique for declouding is more primitive than NASA's.  Somehow they manage to average all data, weighted by the estimated cloud-free quality per pixel.  I instead just pick out the most cloud-free pixel from among all swaths that covered that pixel.  I use NASA's level 2 cloud mask product, which has 48 bits corresponding to various tests based on IR channels, temperature measurements, etc.  I looked through all of these tests and identified 5 that seemed useful discriminators in the Arctic, ranked them, and used the combination to produce a numerical score for every pixel.  Then whichever swath has the highest pixel score in that calculation gets used.  Unfortunately this method still requires a fair bit of manual oversight in the end, as some swaths turn out to add a lot of scattered noise to mess up the picture but no real improvement.

Arctic sea ice / Re: Clear-sky 8-day Arctic mosaic
« on: July 17, 2013, 05:58:18 PM »
The preview looks great, unfortunately the map interface doesn't load openlayers.js from http. Might be a Chrome issue.

Same here, a little shield in the url bar appears. Clicking that gives an option to proceed at your own risk.

Hmm, it might be a hosting issue with http vs https as I haven't tried directly hosting something on google drive before.  It was the easiest free option I could think of for a premade folder structure as google sites makes it very painful to recreate that.

Arctic sea ice / Clear-sky 8-day Arctic mosaic
« on: July 17, 2013, 02:47:47 AM »
MODIS Terra & Aqua produce a variety of processed higher-level data products, almost all of which are then projected onto a sinusoidal tiled grid.  I had mostly been ignoring these products since the sinusoidal grid is almost unusably distorted through most of the Arctic, until I realized that the projection is area-preserving and ideally invertible.  In practice it still took some work to get reasonable results, but here's one.

The data are MOD09A1, Terra 8-day "clear-sky" tiles.  (See for the full list of MODIS products). NASA takes all of its observations every 8 days and produces a best-view mosaic that has unsurprisingly exceeded my best amateur efforts.  They also do daily versions that of course have more clouds but are more up-to-date.   I reprojected to polar stereographic and produced the mosaic using the MODIS reprojection tool, and then combined bands 1, 4, 3 for RGB.  I could produce mosaics of any other band combination as well.

This particular mosaic covers days 185-193. (July 5-12).  Attached is a thumbnail image; the link is to a full-resolution zoomable version, 500m resolution.  NASA's projection from rectangular swath to sinusoidal followed by my undoing of that projection does result in some degradation of resolution, but it is not too terrible, and the procedure is much simpler than direct processing of the swath data has been.  (It also requires far less data downloading).

Pages: [1] 2