Add T/S plots

[DaniJonesOcean]

It would be interesting to see T/S plots clustered into profiles. This would have to be done by depth or perhaps over a depth range.

We would probably have to plot them iteratively over different depths, but they could be added to the same plot, of course.

[maikejulie]

I'm not sure I follow here. Do you mean plotting the clusters that are found differently, or clustering in a different space?

[DaniJonesOcean]

I'm thinking about plotting the clusters differently. Like this!

[DaniJonesOcean]

In this example, the green color represents the central Weddell Gyre class. Here I've separated out the data by depth (pressure). So we can see how the T-S properties of each class change with depth/pressure.

You can sort of get the same information from the plots showing what the average profiles look like, but I think this is a nice way to see the relative properties at a glance.

[isazar]

These look great! Have you tried to put all the profiles together, without separating by depth... I think it can be even more informative.

[DaniJonesOcean]

Thanks! Not yet - I will put that on the to-do list. :)

[maikejulie]

Very cool! Following on Isa's comment, maybe one could plot the clusters individually and use depth as the colouring?

[DaniJonesOcean]

Good idea ✨

[DaniJonesOcean]

Hi both! Finally managed to do this. This is just a proof-of-concept for the individual clusters with depth as colouring as suggested by @maikejulie, so don't spend too much time analysing it. The one in the top-right is the "Antarctic class", as you might guess.

I love the patterns here (e.g. the contrast between K=4 and K=5). Potentially a lot of good physical stuff happening, to the extent that the classes really are spatially coherent.

Edit: Ignore the titles. They're wrong and will be fixed on the next iteration.

[DaniJonesOcean]

And here is the single T-S plot (with all depths) as suggested by @isazar. :)

This combined with the above "by depth" colouring make for a good pair. Thanks for the suggestions!

[maikejulie]

Very cool! These are the same as in the spatial plots, no?

[isazar]

They look great!! acouple of things: the labels ("..at 100 dbar") are wrong, aren't it? and, can i be picky (and annoying)? 🙄 can you please swap temperature with salinity axes?

thank you!!

On Tue, Sep 14, 2021, 23:36 Dan(i) Jones @.***> wrote:

Hi both! Finally managed to do this. This is just a proof-of-concept for the individual clusters with depth as colouring as suggested by @maikejulie https://github.com/maikejulie, so don't spend too much time analysing it. The one in the top-right is the "Antarctic class", as you might guess.

[image: compare_TS_class_depths 001] https://user-images.githubusercontent.com/11757453/133336938-bc1f3403-857b-4bed-b146-f81cdca8c859.png

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/so-wise/weddell_gyre_clusters/issues/23#issuecomment-919530522, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEO74ROMMDPEPYORMIKW2JDUB656JANCNFSM5AZXPHRA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[DaniJonesOcean]

Very cool! These are the same as in the spatial plots, no?

Yes! They're the same as for the 10-class model. I'll put together something more coherent soon, with the label maps, T-S, etc. all in one place for easier comparison.

[DaniJonesOcean]

They look great!! acouple of things: the labels ("..at 100 dbar") are wrong, aren't it? and, can i be picky (and annoying)? 🙄 can you please swap temperature with salinity axes? thank you!!

Yep, titles are wrong. And sure. 😉

[DaniJonesOcean]

@isazar, done. ;)

[DaniJonesOcean]

Since I've added T/S plotting capabilities to the software package, I think this is pretty much done. Closing now!