+6 converters tropical temp/saln, -3 old

[givelberg]

3 tropical moorings converters are replaced with 6 converters, where each converter for
temp and saln is split into 2 converters, one for temp and one for saln.
Since saln is available only up to 100m depth and temp is available up to 500m depth,
this will allow to extract more data from the mbuoyb bufr file.

The issue
#200 (comment)

[sanAkel]

I took the liberty of requesting a few changes. 🙏 for making those changes/adding more flexibility.

[guillaumevernieres]

I took the liberty of requesting a few changes. 🙏 for making those changes/adding more flexibility.

@sanAkel , your request seems reasonable, but FYI: The obs error are set elsewhere in the system and at run time. Where did you get these numbers from @givelberg ?

[sanAkel]

but FYI: The obs error are set elsewhere in the system and at run time.

@guillaumevernieres / @givelberg in the case of ⬆️ would you please ➕ a description to the variable saying that those values are for "reference" or whatever. It seems the code that reads these output (netcdf) files expects certain fields and to make it 😄 you are filling some values!

[givelberg]

I took the liberty of requesting a few changes. 🙏 for making those changes/adding more flexibility.

@sanAkel , your request seems reasonable, but FYI: The obs error are set elsewhere in the system and at run time. Where did you get these numbers from @givelberg ?

The numbers were most likely taken from some casual conversation with you.
For other converters they were sometimes inherited from the previous versions of the converters.

[guillaumevernieres]

@givelberg :

• Can you address @sanAkel 's comments, use 0.5K for temp error and 0.2 for salinity. Slightly more realistic as a base error than what you are using in this PR. Would these work @sanAkel ? We don't uses these so feel free to suggest something that makes more sense to you.
• Has this been tested yet?

[sanAkel]

@givelberg :

• Can you address @sanAkel 's comments, use 0.5K for temp error and 0.2 for salinity. Slightly more realistic as a base error than what you are using in this PR. Would these work @sanAkel ?

You can go ahead with whatever that works for you.

We don't uses these so feel free to suggest something that makes more sense to you.

• Has this been tested yet?