timeaveraging tool should use standard time variables only

[ceblanton]

Is your feature request related to a problem? Please describe. FMS history output has traditionally included time bounds metadata in a non-standard convention. i.e.

• 3 time bounds variables: average_T1, average_T2, and average_DT
  • average_T1 is the start time for the averaging period (in the same time units as time)
  • average_T2 is the end time for the averaging period
  • average_DT is the length of the averaging period, in days

    double average_T1(time) ;
    average_T1:_FillValue = 1.e+20 ;
    average_T1:missing_value = 1.e+20 ;
    average_T1:units = "days since 1979-01-01 00:00:00" ;
    average_T1:long_name = "Start time for average period" ;
    double average_T2(time) ;
    average_T2:_FillValue = 1.e+20 ;
    average_T2:missing_value = 1.e+20 ;
    average_T2:units = "days since 1979-01-01 00:00:00" ;
    average_T2:long_name = "End time for average period" ;
    double average_DT(time) ;
    average_DT:_FillValue = 1.e+20 ;
    average_DT:missing_value = 1.e+20 ;
    average_DT:units = "days" ;
    average_DT:long_name = "Length of average period" ;

• These 3 variables are referenced as a variable attribute in each diagnostic. e.g.

      dis_liq:time_avg_info = "average_T1,average_T2,average_DT" ;

The problem is that there is also a standard way to specify the same time bounds information, specified in the CF standards.

• One time_bounds variable, with an extra nv dimension, that species both average_T1 and average_T2.

  double nv(nv) ;
      nv:long_name = "vertex number" ;
  double time_bnds(time, nv) ;
      time_bnds:units = "days since 1979-01-01 00:00:00" ;
      time_bnds:long_name = "time axis boundaries" ;

• This time_bounds variable is refernced as a variable attribute of time:

      time:bounds = "time_bnds" ;

Describe the solution you'd like

1. The timeaveraging tool should read and use the CF standard time:bounds attribute to find the time bounds variable (and not use the nonstandard average_(T1|T2|DT) variables.
2. The timeaveraging tool should write output time bounds in the standard convention (and not write the nonstandard variables)

Describe alternatives you've considered Continuing to output duplicate time bounds information is a bad idea, and the diag manager rewrite is a good opportunity to remove unneeded legacy conventions.

Additional context https://cfconventions.org/Data/cf-conventions/cf-conventions-1.11/cf-conventions.html#cell-boundaries