Odum-esque schematic, framework

forc-db / ERL-review

Analyses supporting Environmental Research Letters review paper

Creative Commons Attribution 4.0 International

0 stars 0 forks source link

Odum-esque schematic, framework #61

Closed teixeirak closed 3 years ago

teixeirak commented 3 years ago

For reference, here's Odum Fig. 1:

teixeirak commented 3 years ago

Odum's figure is too simplistic. I've made a couple sketches of potential schematics:

option 1

option 2

After starting out with the first one, I came up with the idea of the 2nd, which I think I like quite a bit more. I'd add the second column (biome differences) to option 2.

@bpbond , I'd love to hear your opinion.

teixeirak commented 3 years ago

@CookPatton , I'm flagging you here too to see if you like my proposed schematic (see above).

CookPatton commented 3 years ago

@teixeirak - based on reviewer comments you want to make sure to show at the very least (1) how NEP changes through time and (2) how this might differ for tropical versus temperate forest types. I prefer option2, but you would probably still need the bar charts to the right for the different biomes. Maybe you intended that already...so option2!

bpbond commented 3 years ago

Hi @teixeirak I like those concepts. I wonder if it would be possible to plot partially transparent ForC data on top of the Odum figure? Hmm.

A related possibility might be to do a GCP-style graph, e.g. Figure 3 in Friedlingstein et al. 2019: Screen Shot 2021-01-14 at 3 26 11 PM

...except here it'd be GPP on top, I guess, and then either the respiration fluxes below the x-axis (if emphasizing C balance), or the stocks (for partitioning). Not sure if this would work or not, just a thought.

teixeirak commented 3 years ago

I wonder if it would be possible to plot partially transparent ForC data on top of the Odum figure?

Actually, I find the Odum figure too simplistic, and the terms ambiguous, but its general structure did inspire my version 2.

I love the Friedlingstein example, but don't think it works as nicely here. I like showing that the turnover components of NPP go to respiration.

teixeirak commented 3 years ago

I've started making a schematic in Matlab:

I've just been working on the structure so far, but this gives the general idea. I next plan to basically replicate option 2 above in the left column (and then the right column is just a scaled version of that).

It would be really cool to use our actual regression equations to generate this, BUT mathematically it won't work if we describe each flux with a power function, as we did in the analysis. Ideally, we'd apply a different function to some of them, but not sure it's worth the work of changing it, or how well the budgets would close. That would be a worth exercise, but probably tough to do that much analysis ahead of the re-submission deadline.

teixeirak commented 3 years ago

To do:

[x] ensure schematic is representative of results
[x] add error bars on biomes plot to indicate variability (very high in temperate)
[x] format
[x] make sure it aligns with this:
[x] add dashed lines representing decomp of legacy C, or mention in caption that these are left out.
[x] Note repeated variables in bottom right legend – not sure what is meant here.
[x] The two different uses of dark blue in the top right panel are confusing.

teixeirak commented 3 years ago

I'm happy with this now. May tweak it a bit later, but can close this issue.