Minor corrections to Ex5.10, S9.5 & App F

[RosalynHatcher]

A few inconsistencies:

Example 5.10:

• missing ; on lines 8 & 12

Section 9.5:

• cf_role attribute is only allowed on coordinate variables by table in Appendix A. I think paragraph 2 in 9.5 should say "Where feasible a coordinate or auxiliary coordinate variable with the attribute cf_role should be included."

Appendix F:

• Table 1: Attribute grid_mapping_name should be type 'S' not Numeric.

Happy to create a PR if people agree with this.
Cheers,
Ros.

[dblodgett-usgs]

Example 5.10: http://cfconventions.org/cf-conventions/cf-conventions.html#british-national-grid
Looks good.

Section 9.5: http://cfconventions.org/cf-conventions/cf-conventions.html#coordinates-metadata
I'm not sure cf_role would ever be on what we might call an auxiliary coordinate variable?

Table F.1 doesn't have an appendix entry...
It is just above: http://cfconventions.org/cf-conventions/cf-conventions.html#_revision_history
Looks good.

[JimBiardCics]

@dblodgett-usgs If the variable is not named with its dimension it is, by definition, an auxiliary coordinate variable, even if it is 1D. This is also true if it is not monotonic. Example H.2 is a case in point. The station_name variable is not a coordinate variable. I think @RosalynHatcher's proposal is good.

[dblodgett-usgs]

👍 Got it. That makes sense. @RosalynHatcher -- want to open a PR?

[RosalynHatcher]

@dblodgett-usgs - Yes I'll open a PR

[evertrol]

Example 5.10 also has

float pres(z, y, x) ;
      temp:standard_name = "air_pressure" ;
      temp:units = "Pa" ;
      temp:coordinates = "lat lon" ;
      temp:grid_mapping = "crsOSGB: x y crsWGS84: lat lon" ;

That probably should be pres:... (looks like a copy-paste error).
It doesn't result in an error when using e.g. ncgen, but attributes for pres are lost, and those for temp will be overriden with the incorrect values.

Similarly, example 5.6 has

float rlat(rlat) ;
    rlat:long_name = "latitude in rotated pole grid" ;
    rlat:units = "degrees" ;
    rlon:standard_name = "grid_latitude";

I assume the last line should be rlat:standard_name = "grid_latitude"; instead?

Since the PR is still open, I hope this can be corrected in addition to the other fixes?
Otherwise I'll put in a separate PR (or someone else can).

[dblodgett-usgs]

@RosalynHatcher -- do you want to add these fixes to your open PR?

[RosalynHatcher]

@dblodgett-usgs: Yes I will do.

[dblodgett-usgs]

Closed with #189