mankoff / sankey

Sankey diagram for ice sheet mass flow

See: http://kenmankoff.com

sankey's Issues

Deposition

sankey/dat/gl_baseline.csv

Line 8 in 8476eb7

DP,Deposition,20,TODO

Need estimate of deposition in Greenland.

AQ mass loss

The current AQ Sankey diagram is all-AQ and therefore designed for mass gain.

We need an AQ mass loss (similar to Greenland) for

• West Antarctica
• Model forecasts when AQ is net loss

Evaporation

sankey/dat/gl_baseline.csv

Line 15 in 8476eb7

EV,Evaporation,100,TODO

Need estimate of evaporation in Greenland

Grounding line retreat

sankey/dat/aq_baseline.csv

Line 10 in 8476eb7

GZRET,Grounding line retreat,100,TODO

Need estimate of GZ retreat in Antarctica

Basal freeze-on for ice shelves

Ice shelf basal freeze-on offsets some of the melting. I think we currently have net melt, and it would be good to increase this by the freeze-on amount, and show a freeze-on input (assuming it is not negligible).

Frontal advance in Greenland

Hi @willkochtitzky

As of 9e2b09b I now have frontal advance AND frontal retreat for Antarctica and show both in all figures (no longer net frontal change). Do you have frontal advance for Greenland?

Your 2022 GRL paper mentions 2.5 % advanced, and gives numbers, but dimensions are length not mass.

Your 2022 nature paper says

For the entire Northern Hemisphere, the total terminus mass loss of 11.57 ± 3.81 Gt a−1 during 2010–2020 consisted of 2.98 ± 0.70 Gt a−1 of terminus advance and 14.54 ± 5.97 Gt a−1 of terminus retreat

I could use that, but I thought net retreat was 50 Gt/year - what I currently use in Greenland for the graphic linked above. 50 Gt/yr comes from your 2023 paper abstract, 510.2 ± 18.6 Gt a−1 for 2010–2020.

Basal melt

sankey/dat/aq_baseline.csv

Line 17 in 8476eb7

BM,Basal melt,210,TODO 10 x Greenland b/c Antarctica is 10 area

Evaporation

sankey/dat/aq_baseline.csv

Line 16 in 8476eb7

EV,Evaporation,10,TODO

Add glacier to schematic?

Should arrows be more connected to a real glacier? See for example

source: https://royalsociety.org/topics-policy/publications/2009/geoengineering-climate/

Or

source: Cogley (2011).

Add references and adjust remaining values

Need to fix place-holder values and add references for
 

• Snowfall
• Rainfall (remains on ice)
• Rainfall runoff (remains liquid)
• Condensation
• Deposition
• Runoff
• Evaporation
   
  I assume most of these come from RCMs @xfettweis, @bpynoel, then we can start writing as per #1

Drifting snow

Discussion

Need discussion on drifting snow. What is the uncertainty? Does each snowflake fall and drift 10x? Do 10 % of snowflakes drift 1x?

Graphic

Perhaps the graphic should be updated with a dashed line for this arrow to show it is approximate?

Condensation

sankey/dat/gl_baseline.csv

Line 7 in 8476eb7

CD,Condensation,30,TODO

Need estimate of condensation in Greenland.

Write a Commentary for J. Glac

I think a J. Glac. Commentary based on this figure might be worth a publication.

No new science here, but could still be useful as an overview figure. Perhaps

• Baseline Greenland
• High mass loss year Greenland (2019?)
• Baseline Antarctica
• East AQ
• West AQ

@lhbeem - you seem interested in small publishable units. This seems to fit that criteria. Would you like to help co-author this?

Graphical uncertainty

Should display uncertainty graphically.

Hatch

• Hatch option shown on p.39 of https://mirrors.mit.edu/CTAN/graphics/pgf/contrib/sankey/sankey.pdf#section.12
  https://www.visualcinnamon.com/portfolio/adyen-report-2019/ or https://www.visualcinnamon.com/portfolio/adyen-report-2019/

This approach seems unlikely to work with flows and curves - if the flow has a different width curves will be different.

Could possibly use TikZ style? The current style

      ice-to-air/.style={fill/.style={fill=none, top color=ice, bottom color=air},},

defines a color gradient, here top to bottom. Is it possible to define a binary step function gradient style where you choose the step location (in % or absolute)? Yes, of course, this is LaTeX.

Shading

It may be easiest to not auto-generate uncertainty. Just do it manually for the publication. Use hatch, dots, or shading. See for example https://www.visualcinnamon.com/portfolio/oecd-global-plastics-outlook/

Frontal retreat

sankey/dat/aq_baseline.csv

Line 9 in 8476eb7

FRET,Frontal retreat,100,TODO

Graphical plan

Options are (hopefully): GL, AQ, E, W, & P. Can also generate GL each year, or GL seasonally.

Possible figures

Option 0

• All 5 of GL and 4xAQ in one full-page graphic with sub-panels (see below, but it includes GL2019).
• Figure 2 is GL 4x seasonal
  For each figure, all widths are proportional within and between.

Other options?

See also #17

Antarctic calving vs. submarine melt

sankey/dat/aq_baseline.csv

Lines 14 to 15 in 8476eb7

	ICE,Icebergs,1150,TODO Half of 2300 DYN
	SUB,Submarine melt,1150,TODO Half of 2300 DYN

Check with TC editors if this graphic (and similar for Antarctica) might be appropriate for Brief Comment

Q for @nbkglaciology

I'm wondering how you as an editor for TC would feel about a Brief Comment submission based on this graphic:

Sankey diagram showing all mass flow (Gt) in Greenland

I'm picturing a small paper (Brief Communication) with

• Baseline Greenland (this figure)
• Baseline Antarctica

And possibly

• Anomalous Greenland year (say 2019)
• East Antarctica
• West Antarctica

Discussion would be short as there is no new science here, hence Brief Communication. But I haven't seen an overview of all mass flow processes captured in one quantitative figure like this, and I think it would be helpful for the community as a reference of which processes are covered in any given research, and possibly as an overview figure for IPCC style reports.

Beyond the figure, a few items could be addressed. For example,

• We don't know how much drifting snow exists (does every snowflake that falls drift 2x? Do 10 % of them drift 10x each?).
• If the RCMs have uncertainty of ~15 % on snowfall, that's significantly larger than many of the other terms combined.
• GMB method doesn't "see" some of these (submarine melt, calving, melt-refreeze loop, BMB) while I/O doesn't consider sublimation, evap, etc.).