Latest PIOMAS Volume update

[seaicesailor]

Animation of the daily ice thickness for February. Ice coming from the region north of Greenland can be seen exiting through the Fram Starit.

Click to start the 3 Meg animation.

February thicknesses along the coast of Siberia reach 3 m and 4 m at some locations, whereas in the video these don"t even reach 2m.

[jdallen]

Instead of focusing on models, you should be focusing on the actual sea ice.  It is not solid -- it is full of large holes like swiss cheese.  The ice now is nothing like it was in 2012. 

We've been getting plenty of information that it is soft like butter.  It used to be very solid and hard.

It is simply not the same animal.

what do you expect to be inside those "swiss cheese holes"?

Brine perhaps?

[Andreas T]

But how would the brine inclusions be different from previous years? Younger ice could be a reason. Compaction would change ice structure, maybe affecting draining of brine. Ice temperature is of course an important factor but how different that is, is hard to say. Unfortunately the mass balance buoys are few and most are not reporting data.

[Andreas T]

some information on brine inclusion and brine drainage in sea ice http://pure.au.dk/portal/files/84136596/Galley_et_al_2015_MRI.pdf

[Frivolousz21]

The sea surface temperature shows anomolies over wide areas, presumably through the cracks

this shows an anomoly of .5C so -1.3C and above freezing[-1.8] wayne of http://eh2r.blogspot.ca/
a frequent commenter on the ASI blog, and a keen observer of the ice has stated a few times that you need an air temp. of -11C to get sea water to freeze, so where the graph says winter one assumes temperarures below that. Despite frequent air temp. anomolies of +15C or more I'm not sure it's gone above -11C in Beaufort at all. If you check out waynes blog he has a post up about his doubts on Piomas. If the sst anomolies persist then I doubt we'll get a 'normal' refreeze in Piomas's terms.
http://earth.nullschool.net/#current/ocean/surface/currents/overlay=sea_surface_temp_anomaly/orthographic=-14.55,108.73,1106/loc=-132.129,70.653

If that is modeled it's likely wrong.

[Frivolousz21]

Noob question but how accurate is the PIOMAS model? If you look at the picture the DMI (center) and HYCOM (right) models look totally inaccurate over the CAB when compared to the Cryosat measured data. It's not reassuring that if those models are inaccurate at high latitudes that PIOMAS is.

The middle image is incredibly wrong.  Looks like mid 1990s thickness's. 

The right image is a lot closer to reality than the middle image.

 And the left image is certainly the most accurate.

Cryosat might be off  in some areas due to snow cover, snow drifts, and snow banks up against ridges.

But it's been calibrated extensively and ice bridge will be put up against it the next month and we can compare them.

[jdallen]

some information on brine inclusion and brine drainage in sea ice http://pure.au.dk/portal/files/84136596/Galley_et_al_2015_MRI.pdf

Purely speculative on my part.

Anecdotally, something appears to be rendering the ice less structurally sound.  What are the physics that could do that?  Voids/brine may not be the cause.  What might?

[Peter Ellis]

Instead of focusing on models, you should be focusing on the actual sea ice.  It is not solid -- it is full of large holes like swiss cheese.  The ice now is nothing like it was in 2012. 

What do you mean by "large"?  What size are these holes, where are they, and what methods were used to detect them?  Without giving more detail on what physical processes you're talking about, it's very difficult to have any kind of meaningful discussion.

some information on brine inclusion and brine drainage in sea ice http://pure.au.dk/portal/files/84136596/Galley_et_al_2015_MRI.pdf

Purely speculative on my part.

Anecdotally, something appears to be rendering the ice less structurally sound.  What are the physics that could do that?  Voids/brine may not be the cause.  What might?

Since you're not even sharing the detail of the anecdotes, it's really hard to comment.  If this relates to Professor Barber's description of "rotten ice", then he was quite clear about what was going on.  A mixture of first year ice  "glue" holding together the rubble of older ice presents a surface signature quite similar to old ice, meaning that remote microwave sensors can be misled about what type of ice is present in a given region. This has absolutely nothing to do with brine pockets per se (except that first year ice has more brine).  Brine channels are very small structures - of the order of millimetres across - and cannot be detected directly by satellites!

If you're referring to the breakout effects currently visible in the Beaufort - i.e. a mobile, fractured pack with lots of leads - then there is again a question of scale, this time in the other direction. These are giant structures in the Arctic ice that can only really be seen from orbit.  Remember that a "vast floe" in sea ice term is a floe that's 2-10 km across, i.e. up to two pixels on the AMSR-E maps.  Professor Barber's observations of vast floes of rotten ice are far too small to be meaningfully compared to the microwave data.


So I pose the same question I put to Tenney.  What holes are you talking about, and when you say "structurally sound", what scale are you referring to?  What is the underlying evidence, and is there good enough data from a multi-year time series to say that there is an ongoing trend in structural soundness?

[jdallen]

<snippage>

some information on brine inclusion and brine drainage in sea ice http://pure.au.dk/portal/files/84136596/Galley_et_al_2015_MRI.pdf

Purely speculative on my part.

Anecdotally, something appears to be rendering the ice less structurally sound.  What are the physics that could do that?  Voids/brine may not be the cause.  What might?

Since you're not even sharing the detail of the anecdotes, it's really hard to comment.

<snippage>

So I pose the same question I put to Tenney.  What holes are you talking about, and when you say "structurally sound", what scale are you referring to?  What is the underlying evidence, and is there good enough data from a multi-year time series to say that there is an ongoing trend in structural soundness?

Sorry to be frustratingly unclear Peter.  I was musing vaguely rather than trying to provoke a definitive replay, for which I apologize.

As the question posed is tied to an anecdotal observation, it is quite reasonable in fact to expect no such thing actually exists, and I'm quite willing to accept that conclusion.

And in fact, at the micro scale, I'm reasonably confident that the ice forming mechanisms haven't changed, and FYI is still FYI as it has been all along; there's just a lot more of it.  When I presented my vague and rhetorical questions, that's really what I was aiming at.

AS to the macro scale, I believe you are alluding to the two images of the Beaufort I posted elsewhere (March 13 this year vs April 19th last).  There I'm not suggesting holes, but instead suggesting there is similar over-all structure and composition in the pack across the region captured.  That I think may have implications for the coming melt season, as if the Beaufort ice is a month or more ahead of schedule, we skip conditioning previously required to prepare the ice for losses later in the season.

It's far from a perfect correlation though, as the attendant conditions - temperature and insolation specifically - are much different as we haven't yet reached the equinox.  It's unclear to me as yet what it actually means.  But speculating (that dangerous word again...), I'd say current conditions of the ice (in the Beaufort at least) are such that given equivalent weather conditions the melt season will ramp up a week or two earlier than last year.

[Wipneus]

Yes, PSC has updated their PIOMAS  daily volume data. I updated my graphics, see the top post

Monthly and daily gridded thickness data also updated , I will post the graphics here later.

[jdallen]

So, volume of 22.33^10x3 KM3 on April 1st divided by CT area of 12.65^10x6 KM2 for the same date gives us an average thickness of about 1.77 meters.

[Wipneus]

Animation of the daily ice thickness for March.

Click to start the 3 Meg animation.

[Wipneus]

And the (average) thickness map for March, and the difference with 2007-2015. 2016 seems to be different to each of them, the ice in the eastern Beaufort is about the thinnest of the lot. Ess and Chukchi have remarkably thick ice cover.

[Wipneus]

BTW. I fixed the animation for February as well. I was wrong by one year. The excuse is that the piomas files are sometimes confusing at the moment caused by the leap year. There are four months now with the fourth month apparently only one day in it.  I (have to) do the same, sometimes ignoring the leap day and sometimes not.

[Neven]

Ess and Chukchi have remarkably thick ice cover.

I'm still wondering about this. How can it be so much thicker there than in previous years after such a warm winter? There was no ice there at the end of the melting season, and first-year ice doesn't get much thicker than 2m, right?

The radar and CryoSat comparisons I made for the 2015/2016 Winter analysis also don't show thicker ice there:

[AmbiValent]

Once again, different models predict very different volume and thickness values... but given the development of area and extent, this promises to be an interesting melt season - interesting as in terrifying to those who can imagine what a mostly ice-free Arctic Ocean will do to NH weather...

[seaicesailor]

Ess and Chukchi have remarkably thick ice cover.

I'm still wondering about this. How can it be so much thicker there than in previous years after such a warm winter? There was no ice there at the end of the melting season, and first-year ice doesn't get much thicker than 2m, right?

The radar and CryoSat comparisons I made for the 2015/2016 Winter analysis also don't show thicker ice there...

Just to add as another (very local but reliable) observation that buoy 2015F reports a thickness of about 189 cm at a  location 81 N, in sector north of Beaufort, where Piomas predicts between 2 and 3m at the end of simulation. Remind also that the buoy in anchored in MYI, not FYI.

So yeah maybe PIOMAS went a bit wild in some areas?

PS. I am forgetting the pressure ridges, that constitute a big proportion of volume and average thickness. Maybe just as Beaufort got stretched and opened new cracks, ESS got compressed and piled up ice, and PIOMAS captured that drift effect.

[DrTskoul]

Ess and Chukchi have remarkably thick ice cover.

I'm still wondering about this. How can it be so much thicker there than in previous years after such a warm winter? There was no ice there at the end of the melting season, and first-year ice doesn't get much thicker than 2m, right?

I am with u Neven. Those radar images have proven a great proxy for thickness variation between different years. Reality would not be far off.  The question is what data does PIOMAS lack resulting to the thickening trajectory of the model? Water temperatures? Sth else?

[jdallen]

Ess and Chukchi have remarkably thick ice cover.

I'm still wondering about this. How can it be so much thicker there than in previous years after such a warm winter? There was no ice there at the end of the melting season, and first-year ice doesn't get much thicker than 2m, right?

I am with u Neven. Those radar images have proven a great proxy for thickness variation between different years. Reality would not be far off.  The question is what data does PIOMAS lack resulting to the thickening trajectory of the model? Water temperatures? Sth else?

Good questions, DrTskoul.

Judging from some discussion elsewhere about buoy deployment et. al., it seems there have been some cutbacks in funding for Arctic research - at least in so far as the placement of sensors is concerned.

The models are only as good as the initial data they are provided.  Between those points, they have to generalize, using what global information they have available - such as various satellite data streams.  It's far from perfect, but imperfect as it is, it is pretty amazing and even with those imperfections paints a pretty stark picture.

This year, with the astonishing weather, weird snow cover, highly variable coverage of ice (e.g. - the Beaufort and Chukchi thicknesses may be getting thrown off because of many, smaller chunks of MYI ice embedded in the mix), and then Things We Don't Know all contributing that it may be off is absolutely no surprise.

Now that we have daylight, we can watch Worldview, wait for Icebridge data, and things will become quite a bit more definitive.

[Michael Hauber]

Ess and Chukchi have remarkably thick ice cover.

I'm still wondering about this. How can it be so much thicker there than in previous years after such a warm winter? There was no ice there at the end of the melting season, and first-year ice doesn't get much thicker than 2m, right?

The radar and CryoSat comparisons I made for the 2015/2016 Winter analysis also don't show thicker ice there:

[/quote]

The warmth was largely towards the Atlantic, and also towards Beaufort.  While ESS was still definitely above average, it was probably something much more typical of what we have seen in the last decade.  Combine that with stronger than normal wind towards Siberia (consider how cold Sea of Okhostk has been and how high ice compared to recent years) causeing higher ice transport, then it would be no surprise that we see thicker ice in this region.  However is not showing thicker ice here, so maybe measurement difficulties on Cryosat's part.  Or perhaps the Beaufort Gyre is deflecting ice around the corner towards Laptev instead of piling it up towards Siberia as Piomas is presumably modelling.

[seaicesailor]

Ess and Chukchi have remarkably thick ice cover.

I'm still wondering about this. How can it be so much thicker there than in previous years after such a warm winter? There was no ice there at the end of the melting season, and first-year ice doesn't get much thicker than 2m, right?

I am with u Neven. Those radar images have proven a great proxy for thickness variation between different years. Reality would not be far off.  The question is what data does PIOMAS lack resulting to the thickening trajectory of the model? Water temperatures? Sth else?

Still can CRYOSAT resolve thickening in ridges by compression? Serious doubt we should address.
PIOMAS can estimate it (not resolve the ridges directly but estimate the increased effective thicknesss by ice accumulation).
The other are not radar strctly speaking and do not show thickness but different quality (age) of ice.

[seaicesailor]

This is an example of recent ice drift: observe the divergence of drift vectors at Beaufort (which causes coast water opening, stretching of ice and opening of leads). This does not contribute to ice volume increase given the overall expansion in that given closed area (rather decrease and export out of the Beaufort, even when the leads and openings refreeze). However note the converging of ice drift vectors at CAB side of ESS. That causes ice compression and subsequent pressure ridges. For that given closed area, volume increases just by drift ice compression.

The question that is not clear to me is that if CryoSat can resolve that ice accumulation that goes to pressure ridges, since, given the small area that pressure ridges occupy with respect to the rest of the ice cap, the radar signal of these ridges (reaching back the satellite slightly before the main signal) will be extremely weak.

However, it is known that pressure ridges can accumulate a vast amount of ice volume (reference missing)*.

I should look for a map of overall drift of the past months, but just this as an illustrative example.

* Wikipedia Entry gives this reference:
Leppäranta, M. (2005). The Drift of Sea Ice. Springer-Verlag, New York, 266 p.
that claims that half the volume of ice could be stored in ridges, but I have not read it, and is a bit old given the evolution of the Arctic

[Neven]

I should look for a map of overall drift of the past months, but just this as an illustrative example.

Here's average sea ice drift for February, sis:

[seaicesailor]

I should look for a map of overall drift of the past months, but just this as an illustrative example.

Here's average sea ice drift for February, sis:

Thank you!
Wherever the streamlines diverge, or go parallel and the vectors accellerate along the streamlines, the divergence is positive,  the ice expands. That would be in Canada coast and the usual Laptev.
Wherever the streamlines converge and the vectors accelerate, the ice is displaced but the net divergence may be zero or small. That is Fram and (it seems) Chukchi.
And where streamlines converge with no vector change, or vectors along the streamlines decelerate abruptly, there is compression. That seems to be ESS in this map, and very strong, and the usual Greenland/CAA wall.

The electronic version of that book is 200+ €.

[Andreas T]

...

Still can CRYOSAT resolve thickening in ridges by compression? Serious doubt we should address.
PIOMAS can estimate it (not resolve the ridges directly but estimate the increased effective thicknesss by ice accumulation).
The other are not radar strctly speaking and do not show thickness but different quality (age) of ice.

Chris Reynolds has written last monthhttp://dosbat.blogspot.co.uk/2016/03/piomas-29-february.html on his blog about the thickening in ESS.
I don't know how drift is assimilated into PIOMAS but the increase in very thick ice which can only originate from pressure ridges shows that PIOMAS "sees" this happening on that side. Cryosat by the way also shows thick ice right against the coast but not as large an extent of it (and therefore less volume)
Temperatures along the Siberian coast have not been much warmer than usual as far as I can tell. I have posted some info in the regional thread but don't have time now to dig for data.

[seaicesailor]

...

Still can CRYOSAT resolve thickening in ridges by compression? Serious doubt we should address.
PIOMAS can estimate it (not resolve the ridges directly but estimate the increased effective thicknesss by ice accumulation).
The other are not radar strctly speaking and do not show thickness but different quality (age) of ice.

Chris Reynolds has written last monthhttp://dosbat.blogspot.co.uk/2016/03/piomas-29-february.html on his blog about the thickening in ESS.
I don't know how drift is assimilated into PIOMAS but the increase in very thick ice which can only originate from pressure ridges shows that PIOMAS "sees" this happening on that side. Cryosat by the way also shows thick ice right against the coast but not as large an extent of it (and therefore less volume)
Temperatures along the Siberian coast have not been much warmer than usual as far as I can tell. I have posted some info in the regional thread but don't have time now to dig for data.

Well, Chris Reynolds amazing analysis as always. ESS has a lot more ice and a lot more ridged thick ice imported compared to 2015, by his analysis using Gice (so PIOMAS model indeed can output thickness distributions). Beaufort however is weakest, more than 2012.

Still cannot find a site on if ridged ice will "escape" from Cryosat radar (anybody? : - ) )

[Timothy Astin]

The radar footprint given here (https://directory.eoportal.org/web/eoportal/satellite-missions/c-missions/cryosat) is "about 15km" radius, so about 700km2.

Radar reflections are an "averaged" response from this footprint. In areas of sparse pressure ridges, the dominant response will be from large relatively uniform areas of undeformed ice.

The exact shape of the reflected radar wavelet will include contributions of features smaller than this window, like pressure ridges. How much modification will depend on their individual widths, spatial densities, and the distribution of pressure ridge thicknesses.

So there will be areas where pressure ridge contributions to the dominant radar reflections are effectively hidden.  There will be some thresholds of pressure ridge density and depth distribution which means they start to increase the thickness estimate to be larger than that of the areas of undeformed ice.  (There is a large literature on this topic in radar and seismic reflection physics.)

[seaicesailor]

Thank you Timothy

[Frivolousz21]

It's far more likely Cryosat is right and piomas isn't in terms of ice thickness distribution

[Neven]

I'm going to try and guess the PIOMAS monthly change for April.

Here's the April volume change for other years:



The four years with lowest change were 2010, 2013, 2007 and 2015. If I then compare this year's weather patterns (SAT and SLP) for the first and second half of the month (see attachments) with those of previous years, I conclude that volume change will probably be negative. Difficult to say by how much, perhaps -100 km3. Around -50 km3 will be enough to overtake 2011 and go lowest on record.

[Wipneus]

I'm going to try and guess the PIOMAS monthly change for April.

Here's the April volume change for other years:



The four years with lowest change were 2010, 2013, 2007 and 2015. If I then compare this year's weather patterns (SAT and SLP) for the first and second half of the month (see attachments) with those of previous years, I conclude that volume change will probably be negative. Difficult to say by how much, perhaps -100 km3. Around -50 km3 will be enough to overtake 2011 and go lowest on record.

Impressive. Calculating from gridded daily thickness data I get for day 121 22,268 km³. That is a drop of about -70 km³
Strictly it is not enough to beat 2011, on day 121 2011 volume was 22,247 km³. Remember that PIOMAS does not know anything about leap years/days.

[Wipneus]

Having said that, gridded data was released (despite the F17 troubles) so I can show you the thickness maps. We are still waiting for the daily volume data.

Here is the thickness animation for April. Click to start.

[Wipneus]

From the monthly (average)  thickness data, here is the April 2016 thickness map and a summary of the differences with previous years. Click for bigger pictures.

[Juan C. García]

Having said that, gridded data was released (despite the F17 troubles) so I can show you the thickness maps. We are still waiting for the daily volume data.

Here is the thickness animation for April. Click to start.

On your thickness animation for April, it seems that the F17 made some days with distortion on Baffin and Greenland Sea. Maybe they will come back and correct them.

Your animations are great, Wipneus. Thanks for all this work!

[Frivolousz21]

There is absoutely no way there is 4m ice that  big of an area In the CAb

[crandles]

There is absoutely no way there is 4m ice that  big of an area In the CAb

Cryosat scale goes to 3.5m, less but fairly similar. Why "no way"?

[magnamentis]

as long as the measurement method remains the same we can easily follow the trends but i as well concur that something seems not right with their numbers. i have this feeling for quite some time, for example does it seem a bit off that as soon as a warm front enters a specific area, that the concentration can drop significantly within 2-3 days where is 4m thick ice, as well, even if that would only appear so due to melt ponds etc. that ice appears too fractured to be solid 4m thick in large areas. i suspect that ridges play a big role in this as well as relying on experience from earlier years 15-30 years ago which are applied in their algorithms that calculate the how deep a specific ice mass is reaching as an average. time will show, it's just a subjective impression based on various observations and not scientifically founded :-)

[slow wing]

Without disagreeing with your points, it should be pointed out that the white regions are modeled thickness>3.75m, not 4m.

[magnamentis]

sure, the numbers were not meant absolute, one user posted a 4M the other 3.5 and you 3.75, it was more a general expression of a feeling that those numbers (which ever LOL) should perhaps be taken with a prise of salt, last but not least, which is what i forgot to mention, when i look at 5 graphs i get 3-4 different impressions as to volume and/or thickness. thanks for feedback however and as i said. the trend is fully visible as long as we compare apples with apples (same charts and images) that are using the same algorithm.

[seaicesailor]

Really thick ice in ESS, but in the comparison it is very similar to what it was in 2012
Beaufort on the other hand... like waiting for "le coup de grace"

[Tor Bejnar]

Re Wipneus' "From the monthly (average)  thickness data, here is the April 2016 thickness map and a summary of the differences with previous years."

I find the thin ice 'shadow' next to Wrangel Island (off of eastern Siberia) an interesting feature.  My guess is that late winter/spring ice from Beaufort, Chukchi and East Siberian Seas (Beaufort Gyre) was split by Wrangel Island and ice piled up both east and west of the 'shadow', leaving thicker than normal ice along the (eastern) Siberian coast except for behind Wrangel Island.  This suggests the Beaufort Gyre moved more ice than normal this last month or two.

[Andreas T]

unfortunately the "shadow" is not covered by this sentinel radar image but it may show the effect of ice pushing against Wrangel island. http://www.polarview.aq/images/106_S1jpgsmall/201605/S1A_EW_GRDM_1SDH_20160503T182915_34F5_N_1.jpg

[werther]

Well, on the Polar Science website I saw April coming in. And it's in an on/off battle with 2011 for lowest volume. Somehow, I still think PIOMASS is at the conservative side of history...
This is going to be a nail-biting summer.

[crandles]

Also looks like anomaly below straight line trend.

[Wipneus]

Yes, PSC has updated their PIOMAS  daily volume data. I updated my graphics, see the top post

Day-to-day changes in volume are a bit erratic, no doubt the F17 troubles are due to this. Perhaps this will be corrected some day, now (preliminary, but much better) F18 based data are available.

[Andreas T]

from the graph it looks like April average is below 2011 with maximum just above 2011. From what you say, Wipneus,  the numbers are in a bit of doubt but not surprising in view of the general picture from other parameters.
Looking at the trend for year to year April volume which is lower than the trend for the yearly average, 2016 actually lies clearly below the trend (that Polar Science Center graph does not show April 2016 yet)
http://psc.apl.uw.edu/wordpress/wp-content/uploads/schweiger/ice_volume/BPIOMASIceVolumeAprSepCurrent.png

[wanderer]

Latest value: 2013-3-31 21.612
Latest value: 2013-4-30 21.273

Latest value: 2016-4-1 22.337
Latest value: 2016-5-1 22.268

Am I misinterpreting something? Why is 2016 lower in the graphs than 2013?

[Nick_Naylor]

Latest value: 2013-3-31 21.612
Latest value: 2013-4-30 21.273

Latest value: 2016-4-1 22.337
Latest value: 2016-5-1 22.268

Am I misinterpreting something? Why is 2016 lower in the graphs than 2013?

If you are misinterpreting, so am I

It sure looks like 2016 is neck-and-neck with 2012:

[crandles]

Latest value: 2013-3-31 21.612
Latest value: 2013-4-30 21.273

Latest value: 2016-4-1 22.337
Latest value: 2016-5-1 22.268

Am I misinterpreting something? Why is 2016 lower in the graphs than 2013?

I think you are missing the switch to version 2.1 in 2014.

V2.1 2013 day 90  22.850
v2.1 2013 day 120  22.828

Edit:
It is very close to 2011 figures not 2012.

2016 day 121 22.268, 2011 day 121 22.247
pretty sure that is statistically indistinguishable.

[crandles]

Day-to-day changes in volume are a bit erratic, no doubt the F17 troubles are due to this. Perhaps this will be corrected some day, now (preliminary, but much better) F18 based data are available.

Yes I see.