IOS UBC Vocabulary

[JessyBarrette]

related to HakaiInstitute/cioos-pacific-erddap-public#19 ERDDAP dataset https://pac-dev2.cioos.org/erddap/tabledap/IOS_UBC_profiles.html

Unknown Units

A number of files present variables with no units.

• [ ] sigma-t (n/a) 2236 files -> list. Typically the units associated with sigma-t are kg/m^3

Missing vocabularies

The following variables are missing a BODC vocabulary

• [ ] CONDUCTIVITY_RATIO (n/a): BIO has a similar variable which
• [ ] SIGMA_T (n/a)
• [x] Vitamin_B12 [m-microg/l]
• [ ] Phosphate(inorg) [umol/L]

[JessyBarrette]

Vitamin_B12 [m-microg/l] Drop variable since no real value is associated in the data

[JessyBarrette]

Sigma_T will confirm that n/a should be kg/m^3

[guanlu129]

Vitamin_B12 [m-microg/l] Drop variable since no real value is associated in the data Vitamin_B12 [m-microg/l] is existing in several files. E.g. https://www.waterproperties.ca/osd_data_archive/UBC/6802/68020002.UBC https://www.waterproperties.ca/osd_data_archive/UBC/6805/68050001.UBC csv_file_download_list.csv

[guanlu129]

Sigma_T will confirm that n/a should be kg/m^3

In UBC CTD files, unit used for Sigma-t is kg/m**3

[guanlu129]

Sigma_T will confirm that n/a should be kg/m^3

In the "Comments" of UBC files: Sigma-t values computed using UNESCO Equation of State of Seawater. The unit for in-situ density should be kg/m^3.

[guanlu129]

One question - what does BIO use for CONDUCTIVITY_RATIO (n/a)?

[JessyBarrette]

@guanlu129

Vitamin_B12 [m-microg/l] is existing in several files. E.g. https://www.waterproperties.ca/osd_data_archive/UBC/6802/68020002.UBC https://www.waterproperties.ca/osd_data_archive/UBC/6805/68050001.UBC csv_file_download_list.csv

hum I guess I'm having issues with this specific variable then.

One question - what does BIO use for CONDUCTIVITY_RATIO (n/a)?

BIO calls that variable CRAT which is 42.814*CRAT/10 = Conductivity [S/m] We haven't presented that variable within ERDDAP itself but it's been used to retrieve the conductivity data (BODC = CNDCST01 see here)

[guanlu129]

Need to make a correction here: sigma-T is a unitless concept. It is not really density, first because it does not account for pressure (only depth is available from the .UBC files) and second because it had 1000 subtracted from it. Using kg/m^3 would be misleading. Can we leave the unit blank in nc file?

[JessyBarrette]

I think sigma-t should still have units of kg/m^3 since it is referring the density a specific water mass for a set temperature and salinity at sea surface. Subtracting 1000 kg/m^3 from the unit do no change the units themselves.

This sigma-t convention aligns with the cf term sea_water_sigma_t (see here)

Standard Name	Canonical Units	AMIP	GRIB
sea_water_sigma_t: Sigma-t of sea water is the density of water at atmospheric pressure (i.e. the surface) having the same temperature and salinity, minus 1000 kg m-3. Note that sea water sigma is not the same quantity as the dimensionless ocean sigma coordinate (see Appendix D of the CF convention), for which there is another standard name.	kg m-3

[JessyBarrette]

This sigma-t variable with no unit is also present within the TOB files HakaiInstitute/cioos-pacific-erddap-public#20

[guanlu129]

The following variables are missing a BODC vocabulary

CONDUCTIVITY_RATIO (n/a): Conductivity ratio reported is the ratio of conductivity of the sample to that of standard sea water at 20 degrees Celcius. Should be no unit if it's a ratio calculated. Didn't find matched BODC code.

SIGMA_T (kg/m^3): SIGEQST

Vitamin_B12 [m-microg/l]: removed

Phosphate(inorg) [umol/L]: PHOSAAZX or PHOSZZXX (don't have enough information for analysis method).

[JessyBarrette]

Sounds good @guanlu129, based on all this:

Vocabularies

1. The variables CONDUCTIVITY_RATIO and Vitamin_B12 will be omitted from the erddap dataset.
2. Phosphate(inorg) is present within (~25 files). We could potentially omit it. Otherwise I think using the broader term PHOSZZXX is probably better. That been said it would also match the phosphate variable also available within those files (see UBC/6007/60070062.UBC as an example)

Default missing unit

3. Regarding SIGMA_T, the main issue is related to the missing unit. We can certainly default to kg/m^3. This is affecting 2238 files

[IOS-OSD-DATA]

Sounds good for using PHOSZZXX would be good for Phosphate (inorg).

[JessyBarrette]

Oh actually regrding the Phosphate (inorg).

Those files generally regroups both variables: Phosphate and Phosphate (inorg)

Phosphate is already assigned to the term:

SDN:P01::PHOSAAZX: Concentration of phosphate {PO43- CAS 14265-44-2} per unit volume of the water body [unknown phase] by colorimetric autoanalysis

I feel like this potentially wrong (at least for the UBC dataset) since we have no idea by which method that Phosphate was generated. Perhaps, Phosphate should be associated to the BODC term PHOSZZXX instead.

This doesn't resolve the Phophate (inorg) variables which do not have a specific BODC term. I would personally vote to omit it for now since it is related to such a small number of files (~25/3000) and has no clear BODC match.