Thwaites Glacier Discussion

[steve s]

For a comparison with the previous post with respect to the retreat of the Thwaites Glacier's ice front, see the ice front on March 4, 2013 in the attached image. The variation in the pattern sea ice loss from year to year is a mystery to me.

[Stephan]

Weeks of cloudiness in Thwaites area. With Sentinel and EOSDIS - no chance of evaluation what is going on there.
Today the western edge of iceberg B-22-A is visible and I calculated its WNW movement between Feb 4 and Apr 7. It has moved around 3-4 km since then which is in my opinion a sign that it has melted a little bit from below and has lost some of its pinning points.

[pietkuip]

https://interactive.pri.org/2019/05/antarctica/submarine-glacier-tour.html

An AUV (autonomous underwater vehicle) was underneath the Thwaites ice shelf.

[kassy]

This Is The Most Important Antarctic Glacier The World Needs to Watch Right Now

in western Antarctica, a glacier the size of Florida is losing ice faster than ever before. Sections of the Thwaites Glacier are retreating by up to 2,625 feet (800 metres) per year, contributing to 4 percent of sea-level rise worldwide.

That ice loss is part of a broader trend: The entire Antarctic ice sheet is melting nearly six times as fast as it did 40 years ago.

In the 1980s, Antarctica lost 40 billion tons of ice annually. In the last decade, that number jumped to an average of 252 billion tons per year.

Now, authors of a new study report that over the last six years, the rate at which five Antarctic glaciers slough off ice has doubled. That makes the Thwaites Glacier a melting time bomb.


https://www.sciencealert.com/antarctic-glacier-on-track-to-irreversibly-melt-which-could-trigger-a-chain-reaction

[vox_mundi]

Vintage Film Shows Thwaites Glacier Ice Shelf Melting Faster Than Previously Observed
https://phys.org/news/2019-09-vintage-thwaites-glacier-ice-shelf.html

Newly digitized vintage film has doubled how far back scientists can peer into the history of underground ice in Antarctica, and revealed that an ice shelf on Thwaites Glacier in West Antarctica is being thawed by a warming ocean more quickly than previously thought. This finding contributes to predictions for sea-level rise that would impact coastal communities around the world.

... The researchers made their findings by comparing ice-penetrating radar records of Thwaites Glacier with modern data. The research appeared in Proceedings of the National Academy of Sciences Sept. 2.

"By having this record, we can now see these areas where the ice shelf is getting thinnest and could break through," ... "This is a pretty hard-to-get-to area and we're really lucky that they happened to fly across this ice shelf."

The researchers identified several features beneath the ice sheet that had previously only been observed in modern data, including ash layers from past volcanic eruptions captured inside the ice and channels where water from beneath the ice sheet is eroding the bottom of ice shelves. They also found that one of these channels had a stable geometry for over 40 years, information that contrasts their findings about the Thwaites Glacier ice shelf, which has thinned from 10 to 33 percent between 1978 and 2009.

Dustin M. Schroeder el al., "Multidecadal observations of the Antarctic ice sheet from restored analog radar records," PNAS (2019)

[Stephan]

Thank you for posting this 

[sidd]

The Schroeder paper on Thwaites is interesting.

doi: 10.1073/pnas.1821646116

"In contrast to the stability of the FRIS basal channel, a similar comparison between a 1978 SPRI profile from the Amundsen Sea Embayment (ASE) of West Antarctica with 2 2009 Operation Ice Bridge (OIB) radar-sounding profiles (27) reveals dramatic changes in the subsurface geometry of the remnant eastern ice shelf of Thwaites Glacier, West Antarctica (Fig. 10), which we refer to hereafter as the Thwaites Eastern Ice Shelf (TEIS). In general, the TEIS begins to float seaward of a potentially stabilizing inland ridge and regrounds on an offshore ridge (28)(Fig. 10). The TEIS currently buttresses a portion of the ASE grounding zone along the Walgreen Coast, which is the boundary between Thwaites Glacier and Pine Island Glacier, 2 of the most rapidly changing and potentially unstable glaciers in Antarctica(2, 29). This ice shelf was previously dynamically coupled to the faster-flowing tongue of Thwaites Glacier. However, after 2006, the section connecting the ice shelves collapsed, leading to divergent flow histories (29, 30). This dynamic event caused TEIS to become the only portion of the ASE with decelerating ice flow, in part due to the portion grounded on an offshore ridge(28). "

"By comparing the thickness of the floating and the regrounded portions of the iceshelf interpreted in the radargrams (Fig. 10), we estimate that TEIS lost∼115±62 m of ice thickness (∼10–33%) from 1978 to 2009 (Fig. 10E)(Methods). This contrasts with the regrounded portion of the shelf, which appears stable within our uncertainty estimate (∼19±43 m) (Fig. 10F) over the same period (Methods), suggesting that basal melting rather than dynamic thinning may be the primary cause of the observed thickness change in the floating portion of TEIS. "

"These results suggest that the reduced ice velocities of TEIS during the past decade will not serve to stabilize the Walgreen Coast (29), but instead the TEIS will undergo further thinning and unpinning (28) that could act to destabilize the rest of the ice shelf in the coming decades."

sidd

[Stephan]

Thank you for that information. 
So we should all have an eye on TEIS in the coming years.

[sidd]

Add fig 10 from Schroeder.

sidd

[Stephan]

Iceberg B-22 also has moved in this Austral winter. I checked at three different places, and its WNW movement is in the range of slightly above 2 km in the time between Feb 03 and Sep 10, 2019. Compared to last year this movement has become faster. I guess it must have lost some more pinning points. I have no detailed knowledge of the bathymetry of the sea, therefore I do not know whether a further movement into that direction will end up in a dead end street ?!?

[gerontocrat]

The data on Thwaites comes from a Jan 2019 Paper - https://www.pnas.org/content/116/4/1095
Four decades of Antarctic Ice Sheet mass balance from 1979–2017

They have extracted from a mass of previous studies, and they are confident enough on discharge data to produce analyses by years for data on nearly 200 sub-regions of Antarctica. But the SMB (mainly snowfall) data is not analysed by year, there are only 39 year totals.

But I am not a scientist worried about peer review. So on the attached graphs I have allocated SMB over years by the yearly average. I did this to show how a relatively small percentage annual increase in discharge leads to a high percentage increase in net mass loss.

In the 39 years of the analysis, discharge is estimated at just over 3,800 GT, but with just over  3,200 GT of SMB gain, net mass loss is only just over 600 GT. Half of that may well have happened in the first 30 years, half in the last 9 years.

The current annual net mass loss guesstimate is 37 GT a year. A pure guess at the future gives that increasing to around 60 GT per year in 10 years.

[baking]

Iceberg B22-A has shifted between September 22 and 26.  It is 44 by 24 nautical miles in size and broke off from Thwaites Ice Tongue in 2002 and drifted about 50 km to where it is grounded today.

What's interesting is that is seems to have caused the end of Thwaites Ice Tongue, about one quarter of its length, to separate slightly from the rest.

B22-A shifted in July 2018 and caused the calving of Iceberg B-45 from the nearby Crosson Ice Shelf. https://www.natice.noaa.gov/doc/PR%20-%2020180730%20-%20B-45%20Discovery.pdf

What's puzzling is the mechanism that could cause an effect on the Tongue 50 km away.  The sea between the Tongue and B22-A is covered with sea ice at this time of year so it is possible that the shifting sea ice eased pressure that was holding back the Tongue.  Although it doesn't seem likely, the appearance is that the Tongue was pulled by the sea ice because the remaining 75% of the Tongue did not move as much as the 25% at the end did.  In other words, it looks as if it was pulled away.

Below are links to full resolution images, six days apart.  A small part of B22-A is seen at the bottom edge of the frame.  Please Note that PolarView only hosts these images for 30 days.

https://www.polarview.aq/images/105_S1jpgfull/S1A_IW_GRDH_1SSH_20190926T043625_648C_S_1.final.jpg

https://www.polarview.aq/images/105_S1jpgfull/S1B_IW_GRDH_1SSH_20190920T043542_1AC1_S_1.final.jpg

[baking]

Here is a high resolution detail of the separation of the end of the Thwaites Ice Tongue between Sept. 20 and 26.  Motion on the right is the normal 5km/year.  Motion on the left represents a separation of about 200 meters over the course of six days across the breadth of the tongue.

[baking]

Here the image is scaled down 50% so the movement of the surrounding sea ice can be seen.

[baking]

Here's a final image scaled down by a factor of 8.  A tiny portion of B22-A can be seen at the bottom edge of the image.  The movement of sea ice over a vast region, along with isolated icebergs, iceberg formations, and a huge iceberg almost the size of Rhode Island cannot be a mere coincidence.

Ocean currents and/or wind had to have been the moving force, but the shifting position of B22-A must have allowed the sea ice behind it to follow along bringing smaller icebergs and formations with it.

It also raises the question of whether the fate of Thwaites Ice Tongue can be tied to Iceberg B22-A.  If B22-A were to ever unground and drift off, would the ice tongue become even more vulnerable?

[blumenkraft]

This is going to be an exciting Antarctic summer...

[oren]

Thanks for these updates, baking, especially for the multiple scales - making it much clearer. Seems like a serious shift in Thwaites, I hope it calms down again.

[baking]

A note on scales.  As far as I can figure, the PolarView high resolution images from Sentinel-1 are 20 meters per pixel.  (The Sentinel-1 specs are 10m/pixel, but I think PolarView reduces file sizes by producing a lower resolution.)  That makes the sizes for the above three pictures 14km, 28km, and 112km on a side.

The distance from Iceberg B-22A to Thwaites Ice Tongue is about 100km at their closest points.

[Susan Anderson]

Goodness gracious. I know it's a visual artifact, but I can't help being reminded of a shaking puddinglike entity. Since this is the beginning of the warming season, I would guess hope for inaction is sadly unlikely.

[nanning]

Nice one Susan. Thanks for all the images that your post invoked 
Action it will be then. Or is there no hope? (whatever 'hope' means)

[steve s]

Another possibility is that we are seeing an artifact of instrument calibration or image alignment associated with stitching together pixels captured from different angles by a moving instrument. Perhaps the next image will offer a useful comparison.

[Stephan]

Thank you baking for this information. I'll keep an eye on B-22A, Thwaites Ice Tongue, Eastern Thwaites Ice Shelf and the fast ice surrounding them in the next months.

[baking]

Another possibility is that we are seeing an artifact of instrument calibration or image alignment associated with stitching together pixels captured from different angles by a moving instrument. Perhaps the next image will offer a useful comparison.

I know my immediate thought was that it was an artifact, but on closer look there is no evidence for it.  As for waiting for another image, here is a comparison of two images, different from the previous two, taken on 9/22 and today, 9/27.  They are from wider view images and are therefore at a lower resolution than the previous images, but they show they same separation.

[blumenkraft]

Another possibility is that we are seeing an artifact of instrument calibration or image alignment associated with stitching together pixels captured from different angles by a moving instrument. Perhaps the next image will offer a useful comparison.

I don't think so either. You clearly see some icebergs not moving (likely grounded). A killer argument for this not being optical issues IMHO.

[Stephan]

And, in addition, if this were an optical artifact or due to processing problems, there would be at least some icebergs that partly move and partly don't. But this animation clearly shows that the moving and the non-moving parts are all divided by cracks between the icebergs (scission line). So this is some sense "digital". Either they move fast (left from the scission line) or very slow (right from the scission line), apart for some "not moving at all" grounded icebergs.

[oren]

This must have been an extremely strong pull, with the thick sea ice of September (SH) acting as the rope.
On the positive side, I keep thinking (naively) if B-22 calved in 2002 and its sub-berg is still stuck around due to being grounded, perhaps Thwaites will have a hard time collapsing quickly with an iceberg armada, as one huge iceberg could delay the whole process by a decade or two

[baking]

On the positive side, I keep thinking (naively) if B-22 calved in 2002 and its sub-berg is still stuck around due to being grounded, perhaps Thwaites will have a hard time collapsing quickly with an iceberg armada, as one huge iceberg could delay the whole process by a decade or two

The way I look at it, B-22 broke loose in 2002, but it was fully formed and floating while still attached to grounded ice in 1992.  However it took 10-20 years to form, which means that ice has been floating for 40-50 years.  The old "Thwaites Tongue Iceberg" floated for at least 75 years before breaking up around 1990.  Meanwhile the newer ice in the Tongue floats for about 6 years before breaking up.  This is evidence to me that the Tongue ice is getting thinner.

Couple that with the recent analysis that the Thwaites Eastern Ice Shelf has thinned 23% since the 1970's and I'm not holding out much hope that the process will be delayed by much.

https://www.pnas.org/content/116/38/18867

[Stephan]

I compared the Sentinel pictures from Sep 14 with Sep 27, 2019.
On the SE edge of the Thwaites Eastern Ice Shelf new cracks (green lines) grew in the last two weeks, generating two new icebergs, each around 1 km² in size.
To the east four icebergs float in the sea ice (circled in orange), turned around by 90°. So you can easily see their internal structure with a darker strip, probably weaker ice, which appears darker because these strips are shaded.

[Andreas T]

......

The water depth is known relatively well and available on the bedmap2 from the British Antarctic Survey.

https://www.bas.ac.uk/project/bedmap-2/#data

The whole ice field is over a roughly 300-400m deep part.This is about half the depth of the surrounding areas. I attached an image of the bedrock overlayed with a coastline mask. The mask is maybe from 2012 when the bedmap2 was created and doesn't have the newest glacier front positions.

Oren, it is worth looking through the thread for information on this.
The reason the broken off ice tongue has been stuck for so many years is the relatively shallow dept there. This does not stop more recently calved bergs from leaving the area and meting out further north. B22 does not melt in situ because the upper 300m of water are cold, warmer more salty water is found beneath that layer. Therefore this water reaches the grounding line of Thwaites via a deeper channel east of the stranded tongue. For calved ice to obstruct this water it would have to pile up much higher than the glacier surface. I can't see how it would do that.

[oren]

Thanks for these responses, Tealight and Andreas.

[Stephan]

The disintegration of the tip of Thwaites Ice Tongue continues. I analysed two pictures from Sentinel on Feb 09, 2019 and Oct 04, 2019 (see attached photographs). I marked the movements on the latter picture, together with scission lines (in red and blue) and zones of fragmentation (yellow circled). Longer/stronger arrows mean faster movement. All movements are "fixed" to the intrinsic movement of the whole ice tongue.
For orientation: N is up, to the right follows the Thwaites Eastern Ice Shelf. The picture has a size of approx. 36x19 km

[baking]

All movements are "fixed" to the intrinsic movement of the whole ice tongue.

The important fact about the Tongue to understand is that it is not entirely free floating.  There is an underwater ridge under the end of the tongue and there are peaks at various points that slow down or ground the movement of icebergs passing over them.  The only way to understand the movements of the Tongue is to view its motion on a fixed reference to see the effects of these peaks.

For example, view this GIF in full-screen and try to place your cursor on the spot that you think is causing the icebergs to drag and cause other icebergs to backup behind them and to split off from the rest of the Tongue: https://twitter.com/StefLhermitte/status/1101870380623511552

I don't want to influence your observation, so I will leave you to find it and then we can talk about the consequences for the Tongue going forward.

[baking]

Spoiler Alert!

Do the exercise in the above post first to see if you come to the same or a different conclusion than I do.  I have identified the approximate locations of four potential peaks in the picture below.  They are in order of importance.  Only the first two can be identified from the 4-year GIF, but the second one may be hard to pick out without knowing where it is.  The other two, lesser peaks, can be detected from a close viewing of the last eight months of Sentinel images.

1.  The iceberg currently at this location has been grounded for over a year, going back to September 2018.  With all the movement and chaos going on around it, that berg has not budged one bit.  The only possible explanation is a fairly high peak under it, which it is currently grounded on.  Before that time, the peak was probably responsible for causing two separate rows of icebergs to break away from the Western side of the Tongue in 2017 resulting in a substantial narrowing of the Tongue.

2.  A peak at this location is beginning to have a larger effect.  When Iceberg B-22A shifted during Sept. 22-26, the sea ice between it and the Tongue moved with it, along with the tip of the Tongue causing the rift Stephen has labeled in blue.  The most likely explanation is that this was built up compression from the Tongue during the Southern Hemisphere winter pushing against the thicker sea ice that was released when B-22A shifted.  The Eastern side of the Tongue expanded more naturally and pushed the rest of the tip to the North while the Western side of the Tongue was held by by Peak 2 causing the rift just to the North of it.

Peak 2 is also probably responsible for Stephen's red rift to the South that may cause a further narrowing of the Tongue in the near future.  There is a second parallel rift forming next to it which may also come into play.  The actual motion of the Tongue is shown roughly with the white arrow.

Peaks 3 and 4 are shown for completeness.  I don't expect either one to have a substantial effect on the Tongue going forward.

3.  The thicker of the two icebergs off the Northwestern corner of the Tongue has apparently been grounded ever since it broke off from the Tongue, but it is most likely about to float free since it has almost passed over it's original grounding point.  Its narrower companion berg does not seem to be grounded, but only sheltered by the other one.

4.  A peak can be detected here from the splitting of icebergs as they pass over it.  There is no sign that they are grounded enough to affect the movement of nearby icebergs.

[Stephan]

Thank you baking for this additional information. And a "Like" earned.

[Stephan]

I don't want to influence your observation, so I will leave you to find it and then we can talk about the consequences for the Tongue going forward.

Thank you for the Stef Lhermite twitter link with the wonderful video.

I have a general question:
With further warming of the planet Thwaites Glacier will probably further enhance its speed. Will that cause icebergs that are thicker as they are now, which will create further grounded icebergs or will the calving further upstream give the calved and glued-together icebergs enough time to melt from below which will make them thinner in the end which will cause less icebergs thick enough to get grounded in the shallow waters?

[blumenkraft]

BTW there was a minor calving also.

[blumenkraft]

I have a general question:
With further warming of the planet Thwaites Glacier will probably further enhance its speed. Will that cause icebergs that are thicker as they are now,

I don't think so. Keep in mind the grounding line is also retreating meaning upwelling warm waters would thin out the tongue further land inwards giving the ice more time to thin before they reach the shallow sea.

Hope that makes sense.

[Stephan]

blumenkraft, your answer does make sense to me.
And thank you for the calving information. I didn't have the time to study Thwaites calving front in detail so far.

[FredBear]

The thing about Thwaites is that it produces a raft of bergs that often don't topple over but move as one mass. The glacier seems to preferentially crack the ice vertically but in wide enough segments that they don't roll over as they run aground in relatively shallow water.
In comparison, Pine Island Glacier produces a more uniform berg that drifts away as a unit but sometimes shatters or breaks up later, not being impeded by shallow seas.
I notice that this year the break-up is starting earlier near the glacier fronts and the ice drifting north is filling the areas that have been more open in the last couple of years. Also that the nearby outer edge of the drifting Antarctic ice is closer to the continent this year, I wonder if the summer melt will hit harder?

[baking]

I have a general question:
With further warming of the planet Thwaites Glacier will probably further enhance its speed. Will that cause icebergs that are thicker as they are now,

I don't think so. Keep in mind the grounding line is also retreating meaning upwelling warm waters would thin out the tongue further land inwards giving the ice more time to thin before they reach the shallow sea.

Hope that makes sense.

The way I think of it is that there is this massive pile of ice and snow that is the West Antarctic Ice Sheet.  The weight of all that ice and snow is causing a slow and steady flow at the edges.  As a first approximation, assume that the flow of ice into the glacier is constant.  The first result of faster flow would be thinner ice.  For example, the maximum speed of Thwaites Tongue used to be about 3km/year and now it is about 5km/year.  As a first approximation, I would expect the ice to be about 40% thinner.

As Eric Rignot has explained numerous times, thinner ice causes the grounding line to recede which causes more seawater contact from underneath and increases melting making the ice even thinner.  Thinner ice moves faster because there is less grounding.  It is a viscous circle.

Eventually the mass loss increases, but I would never expect the ice to get thicker as long as it continues to move quickly.

Of course at some point the calving line will recede deep into the glacier and Marine Ice Cliff Instability could kick in and we will see really massive (and rapid) calving events, but hopefully not in our lifetimes.

[blumenkraft]

A vicious circle indeed. :-/

[baking]

I thought of another way to look at the ice velocity vs. ice thickness vs. ice mass loss equation.

You can think of the mass of ice as it calves and floats off, but then you have to account for the melting that takes place below the ice shelf between the time the ice first floats (at the grounding line) and when it finally calves.

A possibly simpler way to look at it is the mass of ice as it passes the grounding line.  With a fixed grounding line (at an underwater ridge) the mass loss is simply a a function of the thickness of the ice at the grounding line times the velocity of the ice at the grounding line.  We can assume the density of the ice and the width of the glacier do not change substantially.

What I am saying is that if something happens downstream to slow the flow of ice and the velocity at the grounding line drops X%, the ice at the grounding line will thicken, but by something less than X%, so the new mass loss will be lower.  Likewise, if you speed up the velocity by X%, the ice will get thinner but by less than X% and the mass loss will rise.

One thing I really like about this model is that it makes it relatively easy to incorporate a moving grounding line.  If the ice is moving forward at A km/year and the grounding line is receding at B km/year, you can add A plus B and multiply it by the thickness to get the mass loss.

Marine Ice Sheet Instability (MISI) is basically an unstable grounding line caused by a reverse-sloping sea bed, thinner ice, and warmer ocean water.  As ice gets thinner, it becomes easier to float.  As the grounding line recedes over a reverse-sloping sea bed, the water becomes deeper and it can support an even thicker layer of floating ice.  In this case, the ice sheet will be thicker at the grounding line, increasing the amount of mass loss significantly.  This might happen at Pine Island Glacier before it happens at Thwaites.

Marine Ice Cliff Instability (MICI) might happen after a MISI event that exposes ice cliffs over about 90 meters in height which are likely to collapse under their own weight and could lead to high speed runaway ice sheet retreat.  In this case the ice velocity might be an insignificant factor to the ice mass loss since the retreat of the grounding line should be significantly faster.

[baking]

Below is a GIF of the last 8 months of movement of a 10km long "cork" that has been holding back the "melange" of ice between the Thwaites Eastern Ice Shelf and the Thwaites Tongue.  For at least 7 months it has been slowly turning and around the end of August and Early September it started sliding out towards the sea, being pushed by the ice behind it.  There is a strong possibility that it may be "caught" at its right front corner in the next few weeks by a couple smaller icebergs that are pinned against the eastern ice shelf.  Otherwise, it might be moving out to sea.  Something to watch anyway.

The "cork" broke off from the Eastern Ice Shelf during 2013-2014 and has turned almost 180 degrees clockwise as it tumbled in the melange.  It has been relatively stationary for the better part of a year, causing the melange behind it to build up.  Some of the larger pieces of ice behind it have been broken up in that time as the got caught between the rapidly moving Tongue and the stationary cork.

Also see the top of this GIF: https://twitter.com/StefLhermitte/status/1101870380623511552

[Tom_Mazanec]

Is there any order of magnitude how long a Thweites collapse would take?
A month, a year, a decade, a century?

[Stephan]

It will, it must take a while.
AbrubtSLR has posted a lot about a possible Thwaites Glacier collapse including MICI type degradation in the Ice Apocalypse thread today.

[baking]

Respectable scientists try not to be alarmists.  I got interested in Thwaites when I heard Richard Alley say it was something like "50 years give or take 50 years."  Maybe it was somewhere in this talk: but he has moderated his language somewhat since then to say that there is a risk that it could happen in decades but it may never happen.  More of a risk analysis rather than a prediction.

On the other hand, a paper published last December tried to model the collapse of Thwaites and it didn't see anything major for 30 years and the 100 year predictions were for something like a contribution of only 8 inches of sea level. https://www.the-cryosphere.net/12/3861/2018/

But if you read the paper, under section 4.6 "Limitations of the model study" you find "Another limitation is that the ice shelf front migration is not included in our simulations. We assume that the ice shelf front position of TG remains fixed" and later "The eastern ice shelf has been thinning and retreating, which means that the ice shelf could disintegrate in the coming decades."

It is hard to understand exactly what they are saying here.  They are making 100 year predictions assuming that the Eastern Ice Shelf remains fixed, yet they freely admit that the Eastern Ice Shelf could disintegrate in the coming decades.

Anyone who looks closely, knows that the Thwaites Eastern Ice Shelf (TEIS) is doomed to collapse in the next ten years.  The ice is visibly sliding eastward off the forward buttress which is at an angle to the current ice movement and not in a position to stop the forward progress.  (See GIF below.)  Meanwhile, the ice on the Western side of the TEIS has broken away (including the "cork" above) leaving no new ice to form a replacement TEIS.  It's impossible to say if the ice further inland might reform TEIS a few years after the collapse, but in my mind it is hard to find solace in these 100 year models based on the "stability" of the TEIS.

The models say that the Eastern half of Thwaites is the one most likely to collapse first, but the Eastern half is currently buttressed by the TEIS.  If the TEIS collapses and the Eastern half starts moving as fast as the Western half is currently, about 5km/year, I don't see how these 100 year predictions are worth anything.

The academic push has been to say "let's really study Thwaites" and they've done one season of observations so far.  Over the next 5 years expect to see a lot more papers published, but in the mean time, all we can do is watch the ice.

[baking]

Is there any order of magnitude how long a Thweites collapse would take?
A month, a year, a decade, a century?

Let me try to rephrase my answer.  Once the collapse "starts" it could take decades to wipe out most of the West Antarctic Ice Sheet.  But best guesses for WHEN it starts, is on the order of centuries and maybe millenia.

But on the other hand, things are happening at Thwaites.  The Tongue is much shorter and narrower than it was.  It is moving much faster and the ice is thinner.  Frankly, the Tongue could be gone after this Antarctic Summer.  Once the Tongue goes, the melange will probably float off also, exposing the Western side of the TEIS to open water during the summers.  We will see calving on both the western and eastern sides of the TEIS narrowing the ice shelf until it becomes unstable.  Collapse of the TEIS will speed up the ice on the Eastern half of Thwaites, pushing back the grounding line.  Then it will be just a matter of time.

Of course, the events above could be delayed for years or even decades.  Or the Tongue and TEIS may reform after collapsing.  The point is that there is plenty to watch for on the remote sensing front, and plenty of academic papers in the pipeline.  Things may happen slowly, but it won't be boring.

[baking]

Today's Sentinel-1 images show a possible new crack on the Eastern calving front of the Thwaites Eastern Ice Shelf (TEIS) near the pinning point.  It's just a few pixels of shading so it may be an artifact, but it lines up with the crack opposite so is worth following up on in later images.

[baking]

I think it's time for an overview of the whole ice front of Thwaites Glacier so the discussion of various areas can be put in context.  The image below is from October 4 and I discuss the major sections from top to bottom (East to West.)  The image size is 112 km on a side and the width of the front as a whole is about 120 km.

Eastern Calving Front:  This is my designation.  It is usually considered to be part of the Eastern Ice Shelf, but this section does not seem to be directly pinned to the offshore ridge.  However it is slow moving because the ice behind it is probably affected by the pinned ice shelf.

Eastern Ice Shelf:  Ice that is caught directly between the glacier behind it and the undersea ridge in front of it.  This shelf was found to have thinned from 10 to 33 percent between 1978 and 2009 after early films of ice penetrating radar were recently digitized.

Melange:  Irregularly shaped ice that has calved from a transition zone between the slow moving Eastern Shelf and the fast moving Tongue.  It tends to stay trapped between the shelf and the tongue before reaching open water after 5-10 years.

Tongue:  Ice that calves from the fasting moving part of Thwaites Glacier, often called the Main Trunk, and tends to stay in formation until it passes over the submerged offshore ridge.  The trunk and the tongue move at about 5 km/year.  The ice tends to calve in long transverse pieces about 10 km long and 1 km wide, which then breakup into roughly 1km squares and get glued to each other with sea ice over many winters before finally breaking up.

Western Calving Front:  This used to be a slower moving part of the Tongue, but now the calving ice tends to float free although it doesn't always move away quickly.  There is usually a lot of ice just offshore combined with ice from the neighboring Haynes Glacier and the Crosson Ice Shelf fed by the Pope and Smith Glaciers.  The Western Calving Front is very close to the Thwaites grounding line in this sector, about a km at points.  The worst case scenario for Thwaites would be if the entire front were to degrade into a calving front like this sector, just dumping icebergs out near the grounding line and providing no buttressing to the glacier.

[Stephan]

baking,
thank you for that comprehensive description and the differentiation of the various parts of this huge and vulnerable glacier . I also keep an eye on it, but I have to wait for clear pictures from Sentinel to do an evaluation of the things that are going on.