| Re: North American Regional Reanalysis (NARR) projection [message #76852] |
Tue, 05 July 2011 16:57 |
MarioIncandenza
Messages: 231
Registered: February 2005
|
Senior Member |
|
|
On Jul 5, 3:07 pm, Ed Hyer <ejh...@gmail.com> wrote:
> Has anyone succeeded in transforming the NARR data to a real
> rectangular coordinate system for matching to other data?
OK, after some intense Googling and trial and error, we found the
information on the spheroid used by NARR, which is not the IDL
default. We arrived at a solution, shown here for posterity:
narr_proj=map_proj_init('Lambert Conformal
Conic',standard_par1=50,standard_par2=50,center_lat=50,cente r_lon=-107,semimajor_axis=6371200,
semiminor_axis=6371200)
We calculated a rectangular grid and checked it like so:
narr_gridxy=map_proj_forward(narr_lons,narr_lats,map=narr_pr oj)
gridbinx=percentiles(narr_gridxy[0,*]-shift((narr_gridxy[0,* ]),
1),val=0.5); median grid spacing
gridbiny=percentiles(narr_gridxy[1,*]-shift((narr_gridxy[1,* ]),
1),val=0.5); median grid spacing
narr_grid_newx=(findgen(349)*gridbinx)+min(narr_gridxy[0,*]) +(gridbinx/
2.0)
narr_grid_newy=(findgen(277)*gridbiny)+min(narr_gridxy[1,*]) +(gridbiny/
2.0)
narr_grid_new_latlon=map_proj_inverse(rebin(narr_grid_newx,
349,277),rebin(transpose(narr_grid_newy,349,277),map=narr_pr oj)
Those values match very closely (within about 3m in projected space).
--Edward H.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
