At this point, I would like to pose a more general question:  does 
identification of a standard_name provide any benefit to either model or user 
when the contents are non-standardized strings?  This seems somewhat different 
from the CF variables that are numerical with standardized units, allowing a 
model (or user) to confidently interpret them from CF-standard attributes alone.

To look at this another way, would it be better to have more than one 
cloud_mask_* standard name, each one linked to a specific semantic "scale"?  
For instance, one scale might be based on cloud fraction, another on cloud 
determination confidence.

On Sep 28, 2012, at 9:01 AM, Martin Raspaud wrote:

> Hi all,
> 
> While working with cloud products extracted from weather satellite data,
> we would like to add the following standard names. Tell us what you think!
> 
> Best regards,
> Martin
> 
> "cloud_mask"
> Description: A variable with the standard name of cloud_mask contains
> strings which indicate whether it is cloudy or not, also classes like
> partly_cloudy is allowed. These strings have not been standardised.
> Alternatively, the data variable may contain integers which can be
> translated to strings using flag_values and flag_meanings attributes.
> Units: 1
> 
> 
> "aerosol_mask"
> Description: A variable with the standard name of aerosol_mask contains
> strings which indicate whether there are aerosols present or not. These
> strings have not been standardised. Alternatively, the data variable may
> contain integers which can be translated to strings using flag_values
> and flag_meanings attributes.
> Units: 1
> 
> 
> "cloud_type"
> Description: A variable with the standard name of cloud_type contains
> strings which indicate the type of cloud  e.g. low_level_cloud or
> thin_cirrus. These strings have not been standardised. Alternatively,
> the data variable may contain integers which can be translated to
> strings using flag_values and flag_meanings attributes.
> Units: 1
> 
> "probability_of_precipitation_in_an_intensity_interval"
> Description: "probability_of_X" means the chance that X is true or of at
> least one occurrence of X.
> For probability_of_precipitation_in_an_intensity_interval, this means
> the chance (in percent) that the precipitation intensity is in the
> defined interval. The interval (from x mm/h to y mm/h) should be
> described in attribute description. Usually two or more variables of
> this type are combined, for adjacent intervals.
> Units: 1
> <martin_raspaud.vcf>_______________________________________________
> CF-metadata mailing list
> [email protected]
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata

--
Dr. Christopher Lynnes     NASA/GSFC, Code 610.2    phone: 301-614-5185


_______________________________________________
CF-metadata mailing list
[email protected]
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata

Reply via email to