Home brew AMSR2 extent & area calculation

[Shared Humanity]

Isn't there a  lot of ice flowing through the Fram? Could an increase further north be causing the uptick?

[SATire]

Isn't there a  lot of ice flowing through the Fram? Could an increase further north be causing the uptick?

Hi,

the tiny "up-tick" of the black line could be explained by the increase further north as you mentioned. But the difference between the purple line and the black line Wipneus explained by different resolution (10 km grid vs. 3.125 km grid). Both measurements investigate the same area and come to different results - so the reason must be the sensor...

[Wipneus]

Isn't there a  lot of ice flowing through the Fram? Could an increase further north be causing the uptick?

No, the uptick only exists in the Jaxa and NSIDC sea ice concentration data. Such excursions are not uncommon in the Greenland Sea region, see the graph. When the ice is tightly packed against the island Jaxa and Uni Hamburg extent are very similar, when the ice spreads out Jaxa jumps upward.

[Wipneus]

Update 20150610.
Changes from 20150603.

Extent: -273.0 (-150k vs 2014, -502k vs 2013, +144k vs 2012)
Area: -466.2 (+99k vs 2014, -189k vs 2013, +490k vs 2012)

This week the decline in extent slowed down to about -40k/day, but area dropped at over -66k/day. This suggests that significant melt ponding (that affects all microwave sea ice concentarion products equally) has started. Finally, as this is later than in the three preceding years.  The slower start causes area now to be higher than in 2014 and especially 2012. In extent it is still lower than in 2013 and 2014.
Regionally ice in the Kara Sea declined most, followed by Baffin Bay and Hudson Bay. Laptev showed a remarkably drop in area, a fact discussed in a a post a few days ago. The odd region is the ESS, the only region that increased both in area and extent.
The melting area visible in the ADS-NIPR-Jaxa AMSR2 thickness/melting map shows and increase to 2013/2014 levels. To stay with 2013 and 2014, the melting has to greatly increase this week. Temperatures have risen above freezing now over most of the Arctic, so this is entirely feasible.
Taken all evidence together this seem to be the message, the next week is crucial that decides whether 2015 is to be among the great melting years, or perhaps has to content with racing in the background with years like 2013 or 2014.
 
You will find the updated graphs in the top post

The details (in 1000 km²):


Extent:
   Central Arctic Basin       East Siberian Sea              Laptev Sea
                    4.3                    17.3                   -30.1
               Kara Sea             Barents Sea           Greenland Sea
                  -81.0                   -14.4                    20.2
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                  -77.0                     0.0                   -49.3
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                   -5.1                   -18.7                   -20.0
             Bering Sea          Sea of Okhotsk            Total Extent
                   -0.6                   -18.7                  -273.0

Area:
   Central Arctic Basin       East Siberian Sea              Laptev Sea
                  -17.9                    17.7                   -89.3
               Kara Sea             Barents Sea           Greenland Sea
                  -96.7                   -15.7                   -35.6
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                  -51.7                     0.0                   -93.6
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                  -29.5                   -17.8                   -21.5
             Bering Sea          Sea of Okhotsk              Total Area
                   -0.5                   -14.1                  -466.2

[Neven]

Taken all evidence together this seem to be the message, the next week is crucial that decides whether 2015 is to be among the great melting years, or perhaps has to content with racing in the background with years like 2013 or 2014.

Very much on the ball, Wipneus. The melt pond situation needs to change drastically for melting momentum to increase. With the CT SIA numbers you reported yesterday it seems compactness will remain high for the time of year.

[Wipneus]

The ice in the Canadian Archipelago is breaking of from the north, first in the Amundsen Gulf followed by the M'Clure Strait and now the fast ice between Prince Patrick and Ellesmere is mobilized (within the "Great Crack"). Serious torching in the south.

[seaicesailor]

Further speculation, fun weather for the next few days. Hot in CAA Hudson and Laptev, over zero in much of the Arctic, and that storm ...

My 2c extent falls 300 k and area 700 k in one week, compactness starts to drop.

Definitely weather forecasts can be so misleading, especially for those who don't read them well. Such a short drop in area with the expectations the reports were creating, and actually I thought to put 800k due to that general increase of temperature.

I observe two things:
1-For some reasons, I find a tendency of models to overestimate temperatures on the 5-10 day range. Particularly GFS.
2-In this forum some of us tend to get over-excited if forecast look bad for ice, and expect drops that then don't come.

If next week there is not a drop of AT LEAST 1000k in area, I eat all those crow left-overs that people hadn't to eat in March. Same with extent AT LEAST 600k.
This means I don't expect to keep pace with 2012 but to stay close.

[plinius]

Think it is more about one of your basic assumptions that "bad weather = area/extent drop".
Particularly at this time of the year, the question if the weather is benign for the CAB or not will have only very weak effects on CAB extent and rather weak effects on the area (by melt ponding): I do not think that the ice area/extent cares a lot if you eat away the snow/upper 10 cm. You have the same area/extent, but a lot worse albedo that then shows up weeks later, when those regions finally weaken.

[Wipneus]

We have noticed before that there is far less fast ice in the East Siberian Sea than last year, it is still there as the animation shows. The shores a blackened by the "torching" effect, also seen in the Laptev and Kara regions. There is a curious white line in the fast ice near the New Siberian Islands that resists the blackness and persists through several frames.

[Wipneus]

Similar feature seen by Sentinel-1A

(quick-look from S1A_EW_GRDM_1SDH_20150610T204054_20150610T204155_006316_0084BA_D537)

[Wipneus]

And in MODIS (Aqua 20150611 via WorldView) as well.

[Shared Humanity]

That line makes no sense to my limited brain.

[slow wing]

Weird! Could it be real ice? Like a huge pressure ridge that resists melting?
Or is it too stable and stationary for that? Maybe instead being a software artifact?
Isn't the sea shallow near the Siberian Coast? So could it be something anchored or attached to the sea floor? A vein of methyl clathrates??
In summary then: weird!

[plinius]

Software artifacts do not get seen in several images from different sources. I'd have a guess that this is an old arc of a fracture in the fast ice that happened I think in February (there was at least a very good candidate) of this year. Created then a number of ridges/crevasses that have a less even surface and probably accumulated far more snow. Now that most of the other snow is gone, this area sticks out like a sore thumb.

[Peter Ellis]

It's the remnant of this cracking event in February:
https://earthdata.nasa.gov/labs/worldview/?p=arctic&l=MODIS_Aqua_CorrectedReflectance_TrueColor(hidden),MODIS_Terra_CorrectedReflectance_TrueColor,Reference_Labels(hidden),Reference_Features(hidden),Coastlines&t=2015-02-21&v=-1548904.1463688046,1078842.529624604,-11880.14636880462,1930810.529624604

It happened shortly before visibility returned to the area, so in the linked photo you can just see the newly-iced-over lead.  Scrolling through the months thereafter, it stays visible as a dark line of thinner ice until the snow covers it.  Now, as the snow melts, it's come back as a whiter line.  I would suspect it's now an area of ridged ice from continual "flexing" of the lead.

[johnm33]

In both the animation and Peters link the line has a similar curve to where the ice break occurs, just slightly rotated, perhaps during the refreeze a large body of ice was driven shorewards and accumulated all the thin ice between those positions until it was stopped where it remains?

[Carex]

This feature is crystal clear on today's MODIS image.  The analysis here also seems spot on except that post cracking snow accumulation seems more likely than pressure ridging.  The feature should also make tracking the fate of this raft of ice through the melt quite easy.  It may remain as clear the infamous goat's head.

[Wipneus]

Thanks Peter. I have looked through the AMSR2 images, the last data that the current ridge was still  a recognizable lead was on Feb 9. Here is a sequence with some dates left out on which nothing was to be seen.

[plinius]

This feature is crystal clear on today's MODIS image.  The analysis here also seems spot on except that post cracking snow accumulation seems more likely than pressure ridging.

Didn't I point to snow accumulation? A bit puzzled.

[quanitlaq]

This feature is over a mile wide for the most part. It seems unlikely that ridging or additional snow cover would cover an area so consistently as to limit melt pond formation in an area where considerable melt is occurring. Is it possible something else may be happening? Could it be that the new ice that formed rapidly in the lead has different properties that are not conducive to melt pond formation? Specifically, I’m wondering if the young, rapidly formed ice may be more porous with brine channels that allow the melt water to drain through the ice.

[Lord M Vader]

The only thing I think is possible to explain that "bone" is that after the "crack event" there was a period with higher pressure, very low temperatures and no precipitation which produced an very thick layer of sea ice at this place. Normally, I would believe that the other parts of the Arctic had seen some precipitation earlier during the beginning of the refreeze season. Snow cover is highly efficient in slowing down or even preventing thickening of the ice even at a few centimeters layer.

No snow cover, low temps and weak winds is a great recipe for a very thick layer of ice.

Do you think that this theory is valid to explain the "bone"?

Best, LMV

[seaice.de]

The feature is clearly visible on recent 3km SMAP radar images. It has a low reflectivity which means that the surface is relatively smooth on the wavelength scale of 20cm which speaks against a ridge. It is probably the refrozen lead that survived within the fast ice.

Unfortunately, we are not yet allowed to share SMAP images until the official release 

[seaicesailor]

The only thing I think is possible to explain that "bone" is that after the "crack event" there was a period with higher pressure, very low temperatures and no precipitation which produced an very thick layer of sea ice at this place. Normally, I would believe that the other parts of the Arctic had seen some precipitation earlier during the beginning of the refreeze season. Snow cover is highly efficient in slowing down or even preventing thickening of the ice even at a few centimeters layer.

No snow cover, low temps and weak winds is a great recipe for a very thick layer of ice.

Do you think that this theory is valid to explain the "bone"?

Best, LMV

I do. fwiw

[slow wing]

Yes, Lord Vader & seaice.de, that sounds like a good explanation, thanks.

[Wipneus]

Here is a detail of the feature as seen by Landsat 8, at 15m/pix resolution. Unfortunately some light clouds and their shadows make it a little more difficult to see, but I think its is clear that it is indeed a frozen lead. I am amazed though to see how dark the ice beyond the feature is, almost all brightness is coming from the spiderweb of cracks.

(click that image for the full resolution)

[Nick_Naylor]

I am amazed though to see how dark the ice beyond the feature is, almost all brightness is coming from the spiderweb of cracks.

Yes, that is startling. What is the nature of this image? Is it visible light, and is the contrast enhanced somehow?

[Wipneus]

I am amazed though to see how dark the ice beyond the feature is, almost all brightness is coming from the spiderweb of cracks.

Yes, that is startling. What is the nature of this image? Is it visible light, and is the contrast enhanced somehow?

Sorry, it is natural visible light composed from R, G, B and sharpened with the 15m/pix pan-band.  Details are somewhere in the developers corner in this forum.
In The Gimp I brighten up the image using the Color->Automatic->White_Balance option (linear stretch of each color between darkest and brightest color).
That normally gives acceptable images composed of ice and snow.

[slow wing]

Great stuff! 

[Wipneus]

The darkening in sea ice concentration maps that are so visible on the Siberian/Russian coasts are also visible in the other side in the Canadian Archipelago. Here too, the concentration is surely grossly underestimated due to the melting in progress. Some is even dropping below 15% affecting extent.

[anthropocene]

The darkening in sea ice concentration maps that are so visible on the Siberian/Russian coasts are also visible in the other side in the Canadian Archipelago. Here too, the concentration is surely grossly underestimated due to the melting in progress. Some is even dropping below 15% affecting extent.

Wow, if the keeps going at that rate the NWP could be open extra early this year. By the way, don't you mean "the concentration is surely grossly overestimated" in the sentence above?

[Siffy]

The darkening in sea ice concentration maps that are so visible on the Siberian/Russian coasts are also visible in the other side in the Canadian Archipelago. Here too, the concentration is surely grossly underestimated due to the melting in progress. Some is even dropping below 15% affecting extent.

Wow, if the keeps going at that rate the NWP could be open extra early this year. By the way, don't you mean "the concentration is surely grossly overestimated" in the sentence above?

No he means what he said. The method used to generate concentration is treating the areas as if the ice has melted out when what has happened is that large scale melt ponds are sitting on the ice.

You can see this using EOSDIS maps from NASA. The ice is very blue showing meltponding but still definitely a solid mass of ice rather than open water with ice surrounding it.

[anthropocene]

Ah, yes of course . Thanks for the clarification.

[Wipneus]

Siffy is quite correct, but I also wrote that the effect is "surely" caused by melting in progress. Implying that this is not good for the ice at all.

[Siffy]

Siffy is quite correct, but I also wrote that the effect is "surely" caused by melting in progress. Implying that this is not good for the ice at all.

Absolutely, my apologies if I gave the impression that this wasn't a fairly big deal. I'm thinking with the way the melt is progressing the CAA and the Beaufort we may see the NWP open up in July.

[Wipneus]

At some places the ice is melting, but some ice prefer to exit the Arctic altogether. Export from the Fram is "high", unlike 2013 and 2014 when it had stalled for the summer.

(click needed)

[plinius]

I find it most impressive how the West Spitzbergen current almost instantly melts out all the ice that enters its realm. Looks nearly like a magic boundary where it simply disappears.

[Andreas T]

Keep an eye on buoy IMB2015E which is about 40km 30km from that edge at just south of 80 deg north. I expect it to move south along the ice edge while ice is melting from the bottom of the floe. How rapid that melting is we will see in the coming days as ice melts between that floe and the open water, lat / lon given here http://imb.erdc.dren.mil/2015E.htm which allows us to locate it in the Worldview images.

[Yuha]

The ESS feature that Wipneus noticed last week and Peter identified as a lead from February is clearly visible on MODIS today (the image at the bottom).

Could the reason for its whiteness be the lack of dirt (dust, algae, etc.) in the ice? This image from the buoy 2015A shows how dirty the ice can get.

[Peter Ellis]

If you go back through the images, the lead flipped very suddenly from being darker than the surrounding ice to lighter than it at the start of June.
http://1.usa.gov/1KVvyz1

Certainly between the 1st and 5th, and I think you can actually narrow it to the window from 2nd to 4th.

That's to narrow a window (and too little ice movement) to have anything to do with ridging or ice thickening, it has to be to do with the overlying snow cover.  Given that this is about the time melt ponding usually sets in, and looking at the greyish hues in the picture above, I think the lead is for some reason accumulating fewer ponds - maybe because the surface is likely flatter and so the water layer will be spread more thinly.

Once the snow melts back fully, it may well revert to being darker than the surrounding ice.

[Steven]

I think the lead is for some reason accumulating fewer ponds

Eicken et al. 2004 suggests that sea ice in (refrozen) leads usually has higher permeability and fewer melt ponds.  In the quote below they refer to the SHEBA project in Beaufort Sea in 1998:

Typically, level first-year ice exhibits the highest pond fractions of any ice type ... Yet, at SHEBA a substantial portion of the first-year ice cover, estimated at between 10 and 30%, exhibited low or zero pond coverage
...
We interpret the unponded ice to have formed on leads and open water in the winter, seeing little to no snow accumulation resulting in rapid ice growth and higher ice salinities. Such ice exhibits substantially higher permeabilities allowing for efficient drainage of surface meltwater.

...
Further evidence for linkages between snow depth and pond fraction was obtained during SHEBA at the airstrip, where snow was completely removed from first-year ice for maintenance, resulting in pond fractions close to zero compared to much higher pond fractions on undisturbed ice.

[Peter Ellis]

Nice find!

Incidentally this makes a great case study for the discussion of albedo measurement.  If the lead represents the colour of level newly-frozen first year ice with minimal snow cover and minimal melt ponding, then presumably this has stayed more or less the same colour while the rest of the pack has darkened around it as ponds begin to form - hence starting darker than the flanking regions and ending up lighter than them.

I'm willing to bet that the optical pictures don't tell that precise story, due to differential exposure time, swath angle, sun angle, etc...

[ChrisReynolds]

Thanks Steven,

I said to Nightvid over at my blog that FYI has more melt ponds, but couldn't remember where I'd read it - now you remind me.

But I had totally forgotten the bit about the drainage through new ice in leads. With a thinner pack that might increase.

[anotheramethyst]

i wonder if melt pond drainage through leads contributed to the 2013 and 2014 melt seasons.  iirc, both years the ice was particularly crushed, so there must have been a lot of fresh and newly frozen leads, and not many melt ponds.

[epiphyte]

i wonder if melt pond drainage through leads contributed to the 2013 and 2014 melt seasons.  iirc, both years the ice was particularly crushed, so there must have been a lot of fresh and newly frozen leads, and not many melt ponds.

I think that this is quite true. You can see the fault lines of old leads under the snow in the Sentinel 1 SAR imagery.
 - Where the ice is thick, the surface is so fractal as to be almost a blur.
 - Where it is new but unbroken, there are far fewer visible features.
    ... but in some areas, the ice is both thin (per PIOMAS/HYCOM), and very granular in appearance - almost like the older ice. (If I'm looking at these images right, the shade of the surface doesn't  tell you much on it's own, it's more about the contrasts in the pattern of reflections, which change, presumably due to the sensor viewing angle, from one side of each image to the other.

....Looking at the same area on MODIS just looks like a plain white sheet until one day, suddenly you see something like this... (in the CAB near the pole about a week ago). No melt ponds, and hundreds of visible holes in the ice (not small ones, either.)

[Wipneus]

One of the more important reasons to calculate extent and area myself, was to get a proper derived calculation that of compactness of the ice pack. Compactness, the ratio of area and extent, is an indication of open water in those grid cells that have a minimum concentration of ice, typically 15% of the grid cell area. Using concentration as a source I can guarantee that extent and area are calculated in a compatible manner which is not the case with using different sources for area and extent (like the CAPIE index).
From the beginning I published a graph (see below) with compactness calculated from data from three main sources ( NSIDC, Jaxa and Uni Hamburg). The graph shows that in the current season compactness has been on the high side, compared with years 2012-2014.

The other reason for the "home brew" effort was to provide regional information to the hemispheric totals.

What i did not do until now was the obvious combination, regional compactness. That has now been rectified, see the image below. I left out the Jaxa compactness, the graphs are very crowded with just NSIDC and Uni Hamburg as it is. Since some regions melt-out during (some) summers making it impossible to define compactness, I set it to zero when the regions extent drops below 30k.

Looking at the graph, I do not think the high compactness (especially UHA) will have such a positive (good for ice) influence on the minimum area and extent. Recognizing that ice in the arctic basin (CAB, Beaufort, Chukchi, ESS and Laptev) will be most important, only one region (Laptev) now has a clearly unusual high compactness. Other regions are within the pack or can even considered to be quite low (Chukchi). Regions that have unusual high compactness are to be found outside the limited arctic basin: Barents, Baffin, Greenland Sea. The latter is no doubt related to the continuing Fram Strait export, not beneficial to the minimum at all.



LINK



LINK

[Wipneus]

Update 20150617.
Changes from 20150610.

Extent: -441.3 (-215k vs 2014, -516k vs 2013, +389k vs 2012)
Area: -493.2 (-101k vs 2014, +3k vs 2013, +513k vs 2012)

Total extent and area declined this week, average rates about -63k/day for extent and -70k/day for area. Those are high numbers but nor unexpected as we are near mid summer. Arctic summer to be unambiguous. Compared with previous years, 2015 extent is somewhat below 2013 and 2014 but well above 2012. For area, 2015 is at the same level as 2013 and 2014 and well above 2012.
Regionally ice in the Kara Sea declined most, nearly 25% from the total. Hudson and Baffin regions are starting to look "slow" while Hudson was well ahead of schedule earlier this year. Normally these regions have melted out (virtually) by the end of June. That should give strong support for strong declines in the coming weeks.
Further the area drop in the Canadian Archipelago  is notable. This is no doubt caused by melt (ponds) as can be seen on the ADS-NIPR Jaxa AMSR2 thickness/melting maps. Melting increased greatly this week, due to above zero temperatures in almost all of the Arctic.
Taking all evidence together, including the regional compactness data presented in my previous post and the continuing Fram Strait export, I am slightly more inclined to think that in the remaining part of the melt season 2015 may end "lowest since 2012". It all depends of course.
 
You will find the updated graphs in the top post

The details (in 1000 km²):


Extent:
   Central Arctic Basin       East Siberian Sea              Laptev Sea
                  -42.5                   -10.9                    10.1
               Kara Sea             Barents Sea           Greenland Sea
                 -120.0                   -57.2                   -11.6
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                  -89.1                     0.0                   -53.9
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                  -17.4                    13.5                   -47.4
             Bering Sea          Sea of Okhotsk            Total Extent
                   -4.5                   -10.4                  -441.3

Area:
   Central Arctic Basin       East Siberian Sea              Laptev Sea
                   -4.6                   -34.9                    29.3
               Kara Sea             Barents Sea           Greenland Sea
                 -147.0                   -47.2                    25.1
Baffin/Newfoundland Bay            St. Lawrence              Hudson Bay
                  -70.2                     0.0                   -69.4
   Canadian Archipelago            Beaufort Sea             Chukchi Sea
                  -88.3                   -10.0                   -62.4
             Bering Sea          Sea of Okhotsk              Total Area
                   -3.5                   -10.0                  -493.2

[Neven]

Wow, that regional compactness graph is wow, wipneus. 

[Wipneus]

Winds over the Hudson did favor ice compaction earlier this season, making ice extent dropping unusually fast. Now the winds have turned, the decline has stalled but it is to be expected that most of his ice will be gone in 2 weeks or so.

[Richard Rathbone]

Nice addition to the graphs.

[Neven]

I've added the regional compactness graph to the ASIG's regional graphs page, but maybe I should start calling it the Wipneus page.