Adds a CF-compliant time dimension for Omega IO

[philipwjones]

Adds a CF-compliant unlimited time dimension

  - adds support for an unlimited time dimension
  - adds the CF-compliant time field with appropriate units
  - adds additional Field properties to support time-dependent fields
  - extends the support for non-distributed field IO into IOStream
  - updates documentation

This still only supports one time slice per file. A subsequent modification will add support for multiple slices in a file.

No additional unit tests were added, but I have manually confirmed the fields and dimensions appear correctly in the output files. Additional tests will be added as part of the upcoming multi-slice capability.

Checklist

• Documentation:
  • Developer's Guide has been updated
  • Documentation has been built locally and changes look as expected
• Testing
  • A comment in the PR documents testing used to verify the changes including any tests that are added/modified/impacted.
  • Unit tests have passed. Please provide a relevant CDash build entry for verification.

[xylar]

@philipwjones, this looks really excellent! I successfully ran CTests on Chrysalis with Intel and OpenMPI but I'm still working on some meaningful tests in Polaris.

In the meantime, I have some comments that I think will be easy to address.

[xylar]

I'm sorry for the delay in reviewing this. I wanted to test in Polaris, but needed to get E3SM-Project/polaris#231 updated based on recent changed to Omega's develop branch before I could test this out. That is finally done and I'm testing this.

[xylar]

Beautiful!!

I'm seeing:

$ ncdump -v time output.nc 
netcdf output {
dimensions:
	MaxCellsOnEdge = 2 ;
	MaxEdges = 6 ;
	NCells = 2500 ;
	NEdges = 7500 ;
	NTracers = 5 ;
	NVertLevels = 1 ;
	NVertices = 5000 ;
	VertexDegree = 3 ;
	time = UNLIMITED ; // (1 currently)
variables:
...
	double time(time) ;
		time:Description = "time" ;
		time:FillValue = 0. ;
		time:Name = "time" ;
		time:StdName = "time" ;
		time:Units = "seconds since 0001-01-01 00:00:00" ;
		time:ValidMax = 1.e+20 ;
		time:ValidMin = 0. ;
		time:_FillValue = 0. ;
		time:calendar = "noleap" ;
		time:long_name = "time" ;
		time:name = "time" ;
		time:standard_name = "time" ;
		time:units = "seconds since 0001-01-01 00:00:00" ;
		time:valid_max = 1.e+20 ;
		time:valid_min = 0. ;

// global attributes:
		:SimulationTime = "0001-01-01_10:00:00" ;
data:

 time = 36000 ;
}

This is correct: we wanted to write out data 10 hours = 36000 seconds into the simulation.

@philipwjones, my only suggestion would be that 0. is not a safe FillValue and _FillValue. This should be something like -9.99e+30 used for LayerThickness.

I the meantime, I'll be working on making the manufactured solution analysis step in Polaris work with this update.

[xylar]

Suggested change

	create("time", "time", UnitString, "time", 0.0, 1.e20, 0.0, 0, DimNames);
	create("time", "time", UnitString, "time", 0.0, 1.e20, -9.9e30, 0, DimNames);

[philipwjones]

Thanks - done.

[xylar]

Thanks @philipwjones, this looks great!

[brian-oneill]

Seems good to me. Successfully passed the unit tests on Chrysalis and Perlmutter CPU & GPU. If you all are happy, I approve.