Apologies for my tardy reply Alison.. The historical evolution of salinity measurement from Knudsen S_K (1901-1966) to Cox S_C (1967-1977) to Practical S_P (1978-present) is what drew me to raise the comment below – these variables were intentionally created to maintain an approximate consistency for sea water salinity.
While I don’t disagree that it seems sensible that the “units” for S_P should be 1 (rather than 1e-3) I think at the very least, the change should be noted in a revised description, as S_K and S_C will not be changed by the current request. These variables (S_K, S_C) were in fact equivalent to a mass fraction (g kg-1), denoted by the ppt (parts per thousand, and it is correct to keep the units for these as “1e-3”). And for this reason as S_K == S_C == S_P @ 35 PSS-78 (approximately), changing one salinity variable and not the others raised my awareness. So the proposed unit change will then “break” the approximate consistency between the previously defined variables (S_K, S_C and S_P). I do note that the 1e-3 has been taken literally by some of the groups contributing to the CMIP5 data archive, and consequently clarifying this will ensure that all numeric values of valid salinity for S_P should be included in the range 2 to 42 PSS-78 (the limits of the defined PSS-78 scale). That is of course, if modelling groups use the sea_water_practical_salinity variable, rather than the more vague sea_water_salinity – which is most often the case. I’d be curious what the opinions are of changing the units of sea_water_salinity alongside sea_water_practical_salinity, as these are approximately the same quantity (model vs observation). Cheers, P From: "alison.pamm...@stfc.ac.uk<mailto:alison.pamm...@stfc.ac.uk>" <alison.pamm...@stfc.ac.uk<mailto:alison.pamm...@stfc.ac.uk>> Date: Wednesday, July 1, 2015 at 4:03 AM To: "Durack , Paul J." <dura...@llnl.gov<mailto:dura...@llnl.gov>>, "cf-metadata@cgd.ucar.edu<mailto:cf-metadata@cgd.ucar.edu>" <cf-metadata@cgd.ucar.edu<mailto:cf-metadata@cgd.ucar.edu>> Subject: RE: Salinity units Dear Paul, Thank you for your comments – your input is certainly welcome. There are currently no plans to make any changes to the existing names, units or definitions of sea_water_cox_salinity or sea_water_knudsen_salinity. I am intending to update the standard name table on July 8th, including the agreed changes to existing practical salinity names as summarized below. Just to clarify, are you objecting to the agreed changes or are you suggesting that we need to clarify further the relationship between sea_surface_salinity and the cox, Knudsen and practical salinity names? If the latter, it suggests we need further discussion of the sea_surface_salinity definition but that would not necessarily be a reason to delay changes to the practical salinity names, would it? The discussion on salinity units arose because oceanographers in the OceanSITES project felt that the canonical unit for practical_salinity should be ‘1’ rather than the current value of ‘1e-3’. The main contention is that 1e-3 is in danger of being misinterpreted as parts per thousand or g/kg when in fact neither is correct for practical salinity. The view was also expressed that the change in units would be unlikely to have adverse effects on existing data, as the values would have been recorded in the range 1 – 40, which is the same as one would expect if the canonical units are ‘1’. The change of units was unanimously supported in the earlier discussion. The change would affect two existing names: sea_water_practical_salinity and change_over_time_in_sea_water_salinity. We agreed amended wording in the definitions of both names as follows: ‘Practical Salinity, S_P, is a determination of the salinity of sea water, based on its electrical conductance. The measured conductance, corrected for temperature and pressure, is compared to the conductance of a standard potassium chloride solution, producing a value on the Practical Salinity Scale of 1978 (PSS-78). This name should not be used to describe salinity observations made before 1978, or ones not based on conductance measurements. Conversion of Practical Salinity to other precisely defined salinity measures should use the appropriate formulas specified by TEOS-10. Other standard names for precisely defined salinity quantities are sea_water_absolute_salinity (S_A); sea_water_preformed_salinity (S_*), sea_water_reference_salinity (S_R); sea_water_cox_salinity (S_C), used for salinity observations between 1967 and 1977; and sea_water_knudsen_salinity (S_K), used for salinity observations between 1901 and 1966. Salinity quantities that do not match any of the precise definitions shoul d be given the more general standard name of sea_water_salinity. Reference: www.teos-10.org; Lewis, 1980 doi:10.1109/JOE.1980.1145448.’ The intention of the wording change is definitely not to alter the interpretation of the practical salinity names but rather to clarify even further the correct interpretation of practical salinity units. It was also felt that the modified wording is less verbose and therefore easier to interpret for CF users. We have been careful to retain text referring to TEOS-10 and the Cox and Knudsen names. Do you perhaps feel that we should not have removed the sentence referring to remotely sensed salinities? If so, then I think we could reinstate it before the changes are committed to the standard name table. There was a brief discussion of whether the units or definition of sea_water_surface_salinity should be changed, but no strong view was expressed except that the units should definitely not be changed to psu. Therefore no changes are currently planned for this name. I hope this is a useful summary of the current situation. It would be helpful if you could explain your concerns about the planned changes further as I am not clear at present why we should not proceed with what has been agreed. Best wishes, Alison ------ Alison Pamment Tel: +44 1235 778065 Centre for Environmental Data Analysis Email: alison.pamm...@stfc.ac.uk<mailto:j.a.pamm...@rl.ac.uk> STFC Rutherford Appleton Laboratory R25, 2.22 Harwell Oxford, Didcot, OX11 0QX, U.K. From: CF-metadata [mailto:cf-metadata-boun...@cgd.ucar.edu] On Behalf Of Durack, Paul J. Sent: 30 June 2015 18:23 To: cf-metadata@cgd.ucar.edu<mailto:cf-metadata@cgd.ucar.edu> Subject: [CF-metadata] Salinity units Apologies for coming to the party so late on this one folks.. I do believe that before progressing on changing the units for any of these variables, there needs to be a consideration/revisitation of what is currently in place. Back in 2011 when the TEOS-10 standard names were proposed we spent a fair amount of time attempting to clean up the mess that was (and is) salinity and “units”. For historians, there’s a full trail of the discussions available at http://mailman.cgd.ucar.edu/pipermail/cf-metadata/2011/022581.html and linked messages previous and subsequent. For this reason 3 “historical” salinities were proposed and accepted (along with a bunch of other quantities) which included “sea_water_knudsen_salinity”, “sea_water_cox_salinity” and “sea_water_practical_salinity” which were intentionally added to attempt to give some context in how “salinity” had been measured through time. For each of these quantities, there is considerable descriptive text included in order to put these quantities into perspective: sea_water_knudsen_salinity: Knudsen Salinity, S_K, is defined unitless as a mass fraction per mil (0/00) or "parts per thousand", and was calculated from the titration of inorganic salts from a sample of sea water after a commission to study the problem of determining salinity and density was initiated by the International Council for the Exploration of the Sea (ICES) in 1899. S_K was the standard salinity measure until S_C (Cox Salinity) was established in 1967. Since the work of Knudsen (1901), chlorinity is converted into Knudsen Salinity using S_K = 0.030 + 1.805 Cl. This type of salinity was called simply "salinity" from 1901 to 1966. From the 1960s on, electrical conductivity began to be used to estimate the Knudsen Salinity, rather than chemical titration. Knudsen Salinity was replaced by Cox Salinity in 1967 which was in turn replaced by Practical Salinity, S_P, in 1978. Conversion of Knudsen Salinity to Practical Salinity follows S_P = (S_K - 0.03) * (1.80655 / 1.805). Reference: Knudsen, 1901; Thomas et al., 1934 doi: 10.1093/icesjms/9.1.28; Lyman, 1969 doi: 10.4319/lo.1969.14.6.0928; Wooster et al., 1969 doi: 10.4319/lo.1969.14.3.0437; Lewis, 1980 doi: 10.1109/JOE.1980.1145448; Millero et al., 2008 doi: 10.1016/j.dsr.2007.10.001. There are standard names for the precisely defined salinity quantities sea_water_cox_salinity, S_C (used for salinity observations between 1967 and 1977), sea_water_practical_salinity, S_P, sea_water_absolute_salinity, S_A, sea_water_preformed_salinity, S_* and sea_water_reference_salinity. Salinity quantities that do not match any of the precise definitions should be given the more general standard name of sea_water_salinity. sea_water_cox_salinity: Cox Salinity, S_C, is defined unitless as a mass fraction per mil (0/00) or "parts per thousand". S_C was the standard salinity measure from 1967 until Practical Salinity, S_P, was established with PSS-78 (1978). Chlorinity, Cl, is calculated from the conductivity of a sea water sample and since the work of the Joint Panel for Oceanographic Tables and Standards (JPOTS; 1966) is converted into Cox Salinity using S_C = 1.80655Cl. This type of salinity was called simply "salinity" from 1967 to 1978. Cox Salinity was replaced by Practical Salinity in 1978. Cox Salinity is converted to Practical Salinity following S_P = S_C, however the accuracy of this is dependent on whether chlorinity or conductivity was used to determine the S_C value, with this inconsistency driving the development of the Practical Salinity Scale 1978 (PSS-78). Reference: Cox et al., 1967 doi: 10.1016/0011-7471(67)90006-X; Lyman, 1969 doi: 10.4319/lo.1969.14.6.0928; Wooster et al., 1969 doi: 10.4319/lo.1969.14.3.0437. There are standard names for the precisely defined salinity quantities sea_water_knudsen_salinity, S_K (used for salinity observations between 1901 and 1966), sea_water_practical_salinity, S_P, sea_water_absolute_salinity, S_A, sea_water_preformed_salinity, S_* and sea_water_reference_salinity. Salinity quantities that do not match any of the precise definitions should be given the more general standard name of sea_water_salinity. sea_water_practical_salinity: Practical Salinity, S_P, is defined on the Practical Salinity Scale of 1978 (PSS-78) and is calculated from the electrical conductivity of sea water (as well as temperature and pressure). Officially S_P is dimensionless so that, while convenient, and while it is common practice, it is not officially sanctioned to say S_P = 35 psu. Often authors use PSS-78, as in S_P = 35 PSS-78. If salinity was measured using remote sensing techniques and not conductivity, then it is recommended that additional metadata (calibration/validation information) be described in the variable comment attribute. This name should be used to describe salinity observations made from 1978 onwards (Practical Salinity is the salinity quantity stored by national data centres for post-1978 observations). The only exception to this is where the observed salinities are definitely known not to be recorded on the Practical Salinity Scale. There are also standard names for the precisely defined salinity quantities sea_water_knudsen_salinity, S_K (used for salinity observations between 1901 and 1966), sea_water_cox_salinity, S_C (used for salinity observations between 1967 and 1977), sea_water_absolute_salinity, S_A, sea_water_preformed_salinity, S_*, and sea_water_reference_salinity. Salinity quantities that do not match any of the precise definitions should be given the more general standard name of sea_water_salinity. Reference: www.teos-10.org<http://www.teos-10.org>; Lewis, 1980 doi:10.1109/JOE.1980.1145448 We need to be a little careful with any changes, as sea_surface_salinity which is now measured by satellite has been calibrated back to in-situ measurements which themselves align to PSS-78 (as do all ongoing measurements that are being and will be stored in oceanographic databases now and in the future). Cheers, P
_______________________________________________ CF-metadata mailing list CF-metadata@cgd.ucar.edu http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
