The GFS/NAM/RUC have the following grids which I believe is an approximation
of the surface elevation (correct me if I'm wrong).
grid: 201:19947514:D=2009060800:HGT:sfc:kpds=7,1,0:6hr fcst:winds are
N/S:"Geopotential height [gpm]
000_000.inv (daily 0Z GFS run)
grid: 4:1314242:D=2009060800:HGT:sfc:kpds=7,1,0:anl:winds in grid
direction:"Geopotential height [gpm] (daily 0Z NAM run)
00.inv (daily 0Z NAM run)
290:26084294:D=2009060800:HGT:sfc:kpds=7,1,0:anl:winds in grid
direction:"Geopotential height [gpm]
000.inv (daily 0Z RUC run)
The "inv" files are generated using wgrib2.
(http://www.cpc.noaa.gov/products/wesley/wgrib2/). Hope this helps you out.
Let me know if you have any more questions.
Seems like I should know this, but I don't. I wrote a gdvint script that
interpolates GFS/NAM/RUC data to height above ground for 300 meters above
ground. It then occurred to me that I don't know what the particular model
thinks is the elevation at my chosen grid point. I did a gdinfo for the
nam212, gfs212, and ruc255 and did not see anything that might give me model
terrain height at a given grid point (say using gdpoint). Does anyone know
to do this?