ESPC-D-V02 incorrect temp/salinity values

[Taiki Sakai - NOAA Affiliate]

I noticed a strange issue with both the t3z and s3z temperature/salinity data, I get some values that are exactly 20.0 that are clearly incorrect (for example temperature at depth jumps from 7.04 to 20 then back down to 5.56). It looks like the add_offset value for both of these variables is 20, so that's likely where the number is coming from. But should these be treated as NA values instead?

I'm accessing the data using OPeNDAP in R using the ncdf4 package.

An example can be found in the ESPC-D-V0/t3z dataset at:

 time=2024-10-31 12:00:00 UTC

Latitude=34.44

Longitude=237.68

The values of water_temp at depth coordinates 350 & 400 are 20.00, a sudden jump from the depths before and after.