"height" in stdatmus_p2tdz, stdatmus_z2tdp

From: Tom Roche <Tom_Roche_at_nyahnyahspammersnyahnyah>
Date: Tue Jan 08 2013 - 12:12:37 MST

summary: is the "height"

* returned by stdatmus_p2tdz
* input to stdatmus_z2tdp

geometric? and is it above ground level (AGL), above mean sea level
(AMSL) or other? (Apologies for any lack of meteorological literacy.)


I'm working toward a 3D-regrid of source data that is horizontally
global and spherical (lon-lat) and vertically hybrid sigma-pressure
(hσp) toward a target that must be regional (over CONUS) and Cartesian
(LCC). (I must merge a netCDF data variable--N2O concentration--into a
target file containing concentrations of many other species, but not
N2O.) Unfortunately ESMF_regrid(...) cannot do this directly; I must
first convert my source/input to a Cartesian grid.

I can convert hσp -> pressure using NCL::pres_hybrid_ccm


and pressure -> "height" using NCL::stdatmus_p2tdz

> If p is a scalar, then a one-dimensional array of length 3 will be
> returned. Element (0) will contain the temperature [C]; (1) will
> contain the density [kg/m^3] and (2) will contain height [m].

But I'm wondering,

* is this height geometric? (which presumably I need for a Cartesian
  transform). I'm guessing it is, from the units, but would like to be

* if geometric, what is the reference datum for that height? I'm hoping
  it's MSL, but need to be sure.

Presuming the height is AMSL, then presumably by

1. cropping the source grid to the horizontal bounds of my target grid

2. converting the source hσp to geometric height AMSL (gh_AMSL) via
   (pcode) (stdatmus_p2tdz(pres_hybrid_ccm(hσp)))[2]

I could convert my global source to a Cartesian form suitable for
ESMF_regrid'ing into horizontal over CONUS LCC with vertical gh_AMSL.
Call that "the intermediate grid."

I can also determine the target layers' gh_AMSL more directly: in
addition to specifying the hσp of layer middles and tops, the target
also specifies the mean (surface) terrain elevation (AMSL) for each
gridcell, and the height AGL for the middle and top of each (gridcell,
layer) tuple. Presuming the intermediate and target gh_AMSL match (to
some reasonable precision) I could then in good conscience merge the
concentrations from the intermediate grid into the target grid.

Am I missing something?

TIA, Tom Roche <Tom_Roche@pobox.com>
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
Received on Tue Jan 8 12:12:57 2013

This archive was generated by hypermail 2.1.8 : Mon Jan 14 2013 - 14:31:38 MST