A standard for time and labelling problems

• To: NETCDFGROUP%EDU.UCAR.UNIDATA@xxxxxxxxxxxxxx
• Subject: A standard for time and labelling problems
• From: The Maverick <ATAE@xxxxxxxxxxxxxxxxxxxxxxxxxxx>
• Date: Tue, 23 Mar 93 21:46 GMT

G'Day

Is there a standard definition for time that people have been using ? 
Currently I have defined time as the set Year,Day,Hour,Min,Sec,Milsec
which suits my purposes. However if there is an already existing standard
I am certainly willing to change.

I have been struggling with another problem for a week or so, and I would 
sure appreciate your help with it. Suppose I have an variable of rank 2 
with dimension names TIME and FIELD and dimsizes UNLIMITED and 3 eg:

TIME   FIELD
 0     1 2 4
 1     3 4 1
 2     5 6 7

What is the correct (or standard way) of labelling the individual elements of 
FIELD (ie the three columns) ? Currently I have abandoned using a single 
variable since I couldn't figure out a "standard" way of doing it. I guess one
method would be to create new attributes and make up a naming convention, but 
this would not be standard and I am trying to make my data sets as easily 
transportable between different plotting/analysis packages as possible.
I am also not sure if "splitting" FIELD into 3 variables has significantly
affected the time it takes to read a single TIME and FIELD record. Any help
would be greatly appreciated.

        best regards
                Ata <(|)>.

• Follow-Ups:
  • A Brief History of (netCDF) Time
    • From: Dave Fulker