Correct erroneous standard_name and comment in Omon tos variable

[durack1]

Copied from PCMDI/cmip6-cmor-tables#54

The current dataRequest 01.00.06 includes an erroneous standard_name and comment in the tos entry suggesting that CMIP5 tos variables included the incorrect standard_name:

1.2 MIP Variable: [tos] Sea Surface Temperature
    procnote[i]: ('',)

    description[i]: temperature of liquid ocean. Note that the correct standard_name for this variable is "sea_surface_temperature", not "surface_temperature", but this was discovered too late to correct. To maintain consistency across CMIP5 models, the wrong standard_name will continue to be used.

    title[i]: Sea Surface Temperature
    prov[i]: OMIP.Omon
    unid[i]: [units] kelvin [fd70554e-3468-11e6-ba71-5404a60d96b5]
    label[i]: tos
    provmip[i]: [mip] OMIP [OMIP]
    1.8 CF Standard Names [sn[i]]: SurfaceTemperature
    units[i]: K
    procComment[i]:
    uid[i]: 0e5d376315a376cd2b1e37f440fe43d3

It does not appear to be sourced from the google sheet https://goo.gl/Rhrcwp

This has propagated across to the CMOR cmip6-cmor-tables

"standard_name": "surface_temperature",
"comment": "temperature of liquid ocean.  Note that the correct standard_name for this variable is 'sea_surface_temperature', not 'surface_temperature', but this was discovered too late to correct.  To maintain consistency across CMIP5 models, the wrong standard_name will continue to be used.", 
...

This wasn't a problem in CMIP5, as the following ncdump output from a file written with CMOR 2.0.1 (circa July 2010) shows:

netcdf cmip5.HadGEM2-AO.historical.r1i1p1.an.ocn.Omon.tos.ver-1.latestX.1860-2005 {
dimensions:                                                                        
        time = UNLIMITED ; // (146 currently)                                      
        bound = 2 ;                                                                
        lat = 216 ;                                                                
        lon = 360 ;                                                                
variables:                                                                         
...
        float tos(time, lat, lon) ;
                tos:comment = "Converted to annual from monthly mean data" ;
                tos:original_name = "TOS" ;
                tos:associated_files = "baseURL: http://cmip-pcmdi.llnl.gov/CMIP5/dataLocation gridspecFile: gridspec_fx_HadGEM2-AO_historical_r0i0p0.nc areacello: areacello_fx_HadGEM2-AO_historical_r0i0p0.nc" ;
                tos:long_name = "Sea Surface Temperature" ;

                tos:standard_name = "sea_surface_temperature" ;

                tos:cell_methods = "time: mean (interval: 1 month)" ;
                tos:cell_measures = "area: areacello" ;
                tos:units = "K" ;
                tos:missing_value = 1.e+20f ;
                tos:history = "2012-10-18T09:37:26Z altered by CMOR: replaced missing value flag (2e+20) with standard missing value (1e+20)." ;

// global attributes:
                :Conventions = "CF-1.4" ;
...
                :project_id = "CMIP5" ;
                :institute_id = "NIMR-KMA" ;
...
                :tracking_id = "2f1e2649-d426-4fae-9c5a-338bdb5f8483" ;
                :cmor_version = "2.0.1" ;
...
                :modeling_realm = "ocean" ;
                :table_id = "Table Omon (08 July 2010) 141295f2779d3084924ae119f2fe4bbc" ;

@dnadeau4 @taylor13 @martinjuckes

[martinjuckes]

Corrected now (the error was inheritted from the CMIP5 "day" table ... and spread to "Omon" in the CMIP6 request ... both now corrected).

There is an ambiguity in the definition of tos in regions of floating ice shelves ... the standard name definition of "sea_surface_temperature" says that it includes the temperature of the ocean under sea-ice, but does not mention ice shelves. Can you add a clarification in the definitions?

[durack1]

@martinjuckes for the tos variable (physics sheet https://goo.gl/Rhrcwp, Omon, D17) we have the description:

This may differ from "surface temperature" in regions of sea ice. For models using conservative
temperature as prognostic field, they should report the SST as surface potential temperature,
which is same as the surface in situ temperature.

What further identifier is required?

[martinjuckes]

What should be reported in areas of floating ice shelves? This is not sea ice, but the surface of the ocean is covered by ice -- does tos include the temperature of the ocean underneath floating ice shelves?

[durack1]

@martinjuckes we could amend this to:

This may differ from "surface temperature" in regions of sea ice or floating ice shelves. For
models using conservative temperature as the prognostic field, they should report the top
ocean layer as surface potential temperature, which is the same as surface in situ temperature.

@StephenGriffies @taylor13 do you agree with this amendment? I am curious if any models include a depth component of icebergs for e.g., so in the higher vertical resolution models would these upper layers be reported as missing_values rather than valid ocean temperatures?

[StephenGriffies]

Hi,

I am fine with the amendment.

However, I am unsure of the answer to your query about icebergs that penetrate into the ocean interior. My presumption is there will be no CMIP6 models with that sophistication. But I may be wrong, in which case we will need to revisit this issue.

@StephenGriffies https://github.com/stephengriffies

On Thu, May 11, 2017 at 3:26 PM, Paul J. Durack notifications@github.com wrote:

@martinjuckes https://github.com/martinjuckes we could amend this to:

This may differ from "surface temperature" in regions of sea ice or floating ice shelves. For models using conservative temperature as the prognostic field, they should report the top ocean layer as surface potential temperature, which is the same as surface in situ temperature.

@StephenGriffies https://github.com/stephengriffies @taylor13 https://github.com/taylor13 do you agree with this amendment? I am curious if any models include a depth component of icebergs for e.g., so in the higher vertical resolution models would these upper layers be reported as missing_values rather than valid ocean temperatures?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/cmip6dr/CMIP6_DataRequest_VariableDefinitions/issues/162#issuecomment-300892326, or mute the thread https://github.com/notifications/unsubscribe-auth/ACBam3KLDF3P9NrQ7jXNtcxGNUZt-Sxpks5r42DwgaJpZM4M_h50 .

-- Dr. Stephen M. Griffies NOAA Geophysical Fluid Dynamics Lab 201 Forrestal Road Princeton, NJ 08542 USA

[durack1]

@martinjuckes this change has been made in the physics sheet (https://goo.gl/Rhrcwp) so please close this once you've reviewed the changes