search

so-wise / weddell_gyre_clusters

Unsupervised classification of Weddell Gyre profiles
MIT License
2 stars 1 forks source link

Estimate temporal variability of class properties #73

Closed DaniJonesOcean closed 2 years ago

DaniJonesOcean commented 2 years ago

This was a question at OSM22 from Erin Thomas:

Did you look into any temporal patterns? Is there any seasonal dependency to the location of the classes? Or do they have any longer term trends of the classification location?

We could look at location statistics (moments of the distribution, with error informed by the number of profiles perhaps). We could also look at the depth of Tmax and other quantities. How do they change with time?

Also should plot how the profile labels change distribution over time.

DaniJonesOcean commented 2 years ago

I've got the seasonal cycle pretty well mapped out.

DaniJonesOcean commented 2 years ago

I've got some code that does this now! We can estimate the mean and the 95% confidence bounds. Can expand on it, but the basic code is in place. 🎉