Calculation of sea water density is not as straightforward as you suggest (i.e. it is not accurate to use just the salinity value to infer the water density). This is because as you move downwards in the water column, the compression due to increasing pressure, changes water properties (e.g. in-situ temperature). There is a well defined standard for these types of calculations used by the oceanographic community, which is based on the thermodynamic properties of sea water. It is called TEOS-10 and you can find out more here:

https://www.teos-10.org/

However, if you just need the average salinity of the water column, a simple solution would be to calculate the weighted-average salinity for each grid-point, using the layer's thickness as the weight.

To do this, divide the sum of the product of salinity times layer thickness, with the total column depth (see figure).

Arguably, using the above method (salinity weighted-average) to calculate water density is more accurate than using a fixed value of 1035. Nevertheless, a more accurate calculation of the total density of the water column would be to calculate the density for each individual layer using TEOS-10, and then calculate the weighted average of those densities for the whole water column.

Best,

Stamatis

Thank you very much for your answer Stamatios Petalas . Now I understand much better this concept.

Best regards,

Marga