[CF-metadata] WRF staggered grids and vertical coordinates

From: Brian Eaton <eaton>
Date: Wed, 28 Jan 2004 19:06:52 -0700

Hi John,

> so the meaning of two different variables with "projection_x_coordinate"
> attribute. is that they are different samplings of the same coordinate?

That seems reasonable to me.

> so that would solve the problem of actually associating the coordinate
> variables. you still have the problem of requiring that extra lat, lon
> variables must be present in the file. what do you think about allowing
> those to be derived by the client?

Putting the responsibility for deriving them on the client reduces
interoperability. We made a similar decision against doing this in the
case where the projection mapping is supplied and the lat,lon values could
be calculated directly from the projection coordinates.

> >You mentioned in an earlier post (27 Oct 2003) that the pressures are
> >defined on the unstaggered grid. In that case the pressures on the
> >staggered grids are obtained by interpolation and it wouldn't be necessary
> >to include them in the file.
> >
> so if i understand you, its ok to derive the z coordinates from
> staggered x and y coordinates, in the case that the z coordinates are
> dependent on x and y.

I may have misspoken here. The intent of the convention is that each data
value should have associated coordinates that allow it to be spatially
located. If supplying a 3D (ignoring time dependence) pressure field is
the easiest way for WRF to provide vertical locations then the 3D pressure
fields for each staggered grid would be required. But all the hybrid
vertical coordinates listed in Appendix D provide formulas that allow 3D
(spatial) pressure or height fields to be calculated from 1D hybrid
coordinates and 2D surface fields. This avoids having to supply 3D fields
of vertical locations. Does the WRF vertical coordinate have a similar
formula?

Brian
Received on Wed Jan 28 2004 - 19:06:52 GMT