Pine Island Glacier (PIG) Calving and Discussion

[paolo]

From uniquorn post
>
"For analysis, best to download it and view frame by frame. (or better, download the original .png files, the ani is half size, cropped)"
I completely agree and moreover it is necessary to juggle between the images because of the gray zones (uncalculated velocity). That's why I hadn't published any animations, but only an example (choosing a date that showed the SIS velocity in their entirety). 

In general you should have an idea to show and choose the images to support it !

>
"x,y coordinate scaling may not be quite right"
I too have just started with panoply (a year ago I even had to ask how to make an animated gif 😊)
For me they are good, in fact for what I understood it is necessary that the intervals of the coordinates X and Y are in the ratio 2 : 1.

>
The second image shows the "standard deviation (STD) of the velocity" and not the velocity.

[uniquorn]

Thanks Paolo. I found 4 scales with the same palette a bit confusing so I tried 3 different ones. The first is 0m-500m to show the outline shape, then 450m-2000m and 1950m-4500m. The overlap is to cut down on white space between the different palettes.
Possible transparency issue with the second palette

Ill add one more image to the collection which might add some reasoning behind the standard deviation image. This one features more contour lines. The gradient along the glacier edge being so steep would probably raise the STD. Contours said more to me than colours in the end.

And a close up. Now I will leave you in peace

[paolo]

The Sentinel2 image of the 16/01 is of poor quality, but allows to focus on the SIS on the SWT side.

Click to enlarge

[paolo]

Calculation of the PIG speeds from the data on the site: http://www.cpom.ucl.ac.uk/csopr/iv/index.php?glacier_number=3 for the months of June and December, from December 2014 to December 2020.

Averages used:
To avoid problems related to doubtful variations in the measurements provided between close points along the transect for the same period and between close intervals for the same position along the transect I calculated the average of the existing measurements for :
> the same interval of points along the transect: 20-30 km, 30-40 km, ..., 120-130 km (I considered that the measurements of the other intervals:  > 130, did not provide sufficiently reliable measurements)
> 5 periods for each month of June and December. For the month of December 2016 and for all months starting in December 2017 we have files that calculate the speeds from intervals of 6 days that follow each other regularly (which makes a global interval of 36 days).This is no longer true for the months prior to December 2016 and for the month of June 2017; in these cases I took the 5 periods closer to the month of calculation, with, as a result, an overall interval between 42 and 108 days and with overlaps between the periods used for the velocity calculations. Moreover for the months of June 2015 and 2016, for one of the chosen intervals I find no valuation of the speed and for these two months I calculate the speed using in reality only 4 intervals.  For the month of December 2015 I find no valuation for the points between 20 and 30 km.

Sources of speed valuations along the transect:
> for the period 2014-2019 I based myself on the text files containing the valuations per point and per period, files contained in the global zip that I downloaded. 
> For the month of June 2020 I am based on the individually downloaded text files.
> for the month of December 2020 no text files are available and I based on the image of the downloaded diagram to value the speeds by interval: 20-30 km up to 60-70 km and I dropped the other intervals for which a reliable valuation was not possible (in the first image I give an example of the images used for my valuations)

Results :
> The measures found are presented in the second image and it can be seen that they are consistent with each other.
> There is indeed a more or less constant acceleration from June 2018. To bring it out more clearly I calculated the averages for the 20-70 km interval that I show in the third image.

Click to enlarge

ADD : In the third image it is the average over the interval 20-70, the values for the interval 20-30 are higher, e.g. for December 2020 there are 4.816!

[interstitial]

Thanks Paolo. I know sorting through that data was more work than it appears.
Did some major obstruction clear around Dec 2017 - Jun 2018?

[paolo]

There was mainly the calving on 23/09/2017, after which the pinning point provided by the SWT lost its strength and it gradually eroded to disappear with the calving of 29/10/2018, after which it remained only an indirect and further upstream pinning point: the Cork, which in turn calved on 09/02/2020.
Since no more action of the SWT on the PIG
To these elements can be added a progressive instability of the ice sheet: MISI (see Rosier et al. 2020).

And it's not going to get any better 

[baking]

There was mainly the calving on 23/09/2017, after which the pinning point provided by the SWT lost its strength and it gradually eroded to disappear with the calving of 29/10/2018, after which it remained only an indirect and further upstream pinning point: the Cork, which in turn calved on 09/02/2020.
Since no more action of the SWT on the PIG

Thank you Paolo for doing this.  I've been putting off doing it myself for a week now.

I agree that the calvings of September 2017 and October 2018 along with the disconnect from the SWT are significant factors.  One important point however is that the constantly increasing velocity in the last three years implies an ongoing and worsening positive feedback effect, such as the worsening situation in the Southern Shear Margin.

Take a look at this 6-year movie of PIG:
https://movie-usa.glencoesoftware.com/video/10.1073/pnas.1912890117/video-1

You can clearly see the relative stability before September 2017 and the rapid collapse after October 2018, so the interesting time to me is the period between those two calvings when the velocity started really accelerating.

What I think is the turning point is in April 2018 when the SWT suffered a major collapse and it no longer supported the Southern Shear Margin.  Also, the calvings in the South have been almost constant since then.

Of course, the collapse of the SWT came about because of the September 2017 PIG calving, but the velocity of PIG did not really pick up until after the SWT collapse.

[oren]

Big thanks paolo for the velocity over time charts. Very useful and extremely disconcerting.

Indeed, from the video it appears the SWT calving was the immediate trigger for acceleration - and it makes perfect sense that such loss of buttressing would result in exactly this outcome. And the SWT calving was triggered by the main PIG calving, which got too near to where it mattered.

[sidd]

Re: Rosier et al. 2020

may i please have a doi for that reference ?

sidd

[paolo]

"The tipping points and early-warning indicators for Pine Island Glacier, West Antarctica"
Sebastian H. R. Rosier, Ronja Reese, Jonathan F. Donges, Jan De RydtG. Hilmar Gudmundsson, and Ricarda Winkelmann
doi.org/10.5194/tc-2020-186
Preprint. Discussion started: 4 August 2020
Review status: this preprint is currently under review for the journal TC

[baking]

"The tipping points and early-warning indicators for Pine Island Glacier, West Antarctica"
Sebastian H. R. Rosier, Ronja Reese, Jonathan F. Donges, Jan De RydtG. Hilmar Gudmundsson, and Ricarda Winkelmann
doi.org/10.5194/tc-2020-186
Preprint. Discussion started: 4 August 2020
Review status: this preprint is currently under review for the journal TC

"Critical slowing is measurable as an increase in the state variable auto-correlation. We measure this here using the lag-1  auto-correlation  function  (Dakos  et  al.,  2008;  Scheffer  et  al.,  2009;  Held  and  Kleinen,  2004)  applied  to  the  grounding line flux over a 300 year moving window preceding each tipping point (hereafter referred to as the ACF indicator)."

This might come in handy after we have 300 years of data.

On a more serious note, another way to look at it is that as long as the velocity is increasing, PIG is in an unstable state.  Only when the velocity levels out can we say that it has at least temporarily stabilized.

[sidd]

Thanks for the Rosier reference.

sidd