Look at float 6900230 (Portugal)

[AndreaGarciaJuan]

https://fleetmonitoring.euro-argo.eu/float/6900230

• notebook: preproc_gdac_files_before_pyxpcm.ipynb

[AndreaGarciaJuan]

Correction using argo data

Correction using CTD data

[AndreaGarciaJuan]

Classification k=4 with argo data

Spatial distribution

Temporal distribution

Salinity profiles

Temperature profiles

The green class has a maximun of salinity around 800dbar

Correction

[AndreaGarciaJuan]

Classification k=4 with CTD data

Spatial distribution

Temporal distribution

Salinity profiles

Temperature profiles

The grey class has a maximun of salinity around 800dbar

Correction

[AndreaGarciaJuan]

Fit with CTD data + prediction with argo data

[gmaze]

Fit with CTD data + prediction with argo data

are you sure ?
looks like the Argo data fit figure here #23 (comment)

[gmaze]

are all the correction figures:

Classification k=4 with CTD data: #23 (comment)

and

Classification k=4 with argo data: #23 (comment)

computed using OWC with classif ?

[gmaze]

Ok, my understanding is that: "With Argo ref data, there is no need for a correction, while there is one needed with CTD data"

Do we have the same conclusion without using PCM+OWC ?