Correct documention regarding NEMO floats and pressure auto-correction

[delphinedobler]

Following a discussion between Olaf Boebel, Birgit and I, we learned that NEMO floats never applied any auto-correction for pressure, whatever the serial number or firmware/software version, but only the NEMO documention had changed to detail this aspect, which brought the confusion.

• Argo documentation should be updated accordingly, in particular within Surface pressure correction table, January 13th 2015.

During this discussion, we also clarified which internal pressure_offset field should be recorded within the Argo Tech file PRES_SurfaceOffsetNotTruncated_dbar, depending on the format version of the intermediate ascii files (containing NEMO decoded data). Indeed, Nemo floats transmit this pressure surface offset at two moments: when the float reaches the surface after ascent and just before if dives for the descent. For the full match-up used to re-generate Tech files, the python code can be found here: https://github.com/delphinedobler/NEMO_legacy_handling/blob/main/TECH/scrpy_modification_of_Argo_techdata.ipynb

Please note that the correction has been performed for almost all NEMO floats (now in Argo TECH format_version 3.1), and that 12 have additionnal issues to solve before completing (they remained in Argo TECH format_version 3.0).

• this discussion raised a question: Shall we continue to apply the pressure offset correction ? considering that 1) offsets are often very small for NEMOs (<0.5 dbar) 2) there are large under-ice profile time series fir which a correction is not possible, which could lead, if pressure is adjusted, to un-consistent whole-life time series 3) given the amplitude of the offset, it has no impact on the salinity. This is for discussion, any comment or consideration are welcome.