Question: What are the small disparities in Absolute Salinity results between releases 3.4 and 3.6?

[soutobias]

Hello,

I'm currently utilizing this library to convert Practical Salinity to Absolute Salinity within a package I'm working on. Upon upgrading from version 3.4.2 to 3.6.16, I encountered some test failures.

Upon closer inspection, I observed minute discrepancies in the Absolute Salinity results. Please refer to the variations below:

import gsw
gsw.SA_from_SP(SP=24.835115, p=0.5038634337828468, lon=-19.888672, lat=40.066406)

In version 3.4, the output is 24.952401390208948, while in version 3.6, it is 24.952399876617687. The difference is 0.00000151359 g/kg. Although I acknowledge the marginal nature of this difference, I am keen to understand the reasons behind it.

Thanks in advance for your help

[DocOtak]

This is almost certainly related to upstream fixes here: TEOS-10/GSW-C#53. as part of that work, the underlying SAAR (Absolute Salinity Anomaly Ratio) data was updated. This data is a lookup table for the global ocean based on observations, see the Notes on gsw_SAAR pdf for more on that. I myself wasn't able to find any notes on what might have changed in the SAAR data, hopefully someone here has some knowledge on that. The notes on SAAR do state that as more observation data becomes available, this calculation is expected to change/improve with time.

Many check values in that PR did change slightly including the ones for SA_from_SP.