Re: Plotting a non-standard WRF projection

Hi Don,

For some reason, the WRF_po_1.ncl example has access to the 2D lat/lon
values (lat2d and lon2d), but is not using them to do the
overlay. Instead, it is attempting to set up the map projection
correctly, and then just draw the contours straight on top without any
transformation taking place (this is what setting tfDoNDCOverlay to
True does). This is actually a valid way of doing things, but if
you don't set up the projection exactly right, then, as you saw,
the data won't align quite right.

Try this instead, and see if you get different results:

   Right after you retrieve lat2d and lon2d, add the following lines:

   x_at_lat2d = lat2d
   x_at_lon2d = lon2d

then:

   res_at_tfDoNDCOverlay = False (or else just remove this line since it
                                 is False by default.)

Meanwhile, I'm cc-ing Dennis Shea in case he has an explanation about
this example. If you continue to have problems, please email me
directly, and we may ask for your script and data.

Thanks,

--Mary

On Wed, 16 May 2007, Don Morton wrote:

> Howdy,
>
> Through my own stupidity several years ago I blindly set
> up a WRF domain for Alaska using WRFSI, and did so
> producing a non-standard polar stereographic projection.
> This seems to give me problems when using the NCL
> WRF examples - the actual data from the netCDF is
> mis-placed relative to the geographic boundaries.
>
> Specifically, some of the relevant parameters for my
> projection are as follows. Note that CEN_LON is significantly
> different than STAND_LON
>
> :WEST-EAST_GRID_DIMENSION = 320 ;
> :SOUTH-NORTH_GRID_DIMENSION = 225 ;
> :BOTTOM-TOP_GRID_DIMENSION = 75 ;
> :DX = 7500.f ;
> :DY = 7500.f ;
> :GRIDTYPE = "C" ;
>
> :CEN_LAT = 65.4f ;
> :CEN_LON = -160.2f ;
> :TRUELAT1 = 67.5f ;
> :TRUELAT2 = 1.e+20f ;
> :MOAD_CEN_LAT = 65.40001f ;
> :STAND_LON = -150.f ;
>
> When I use the examples from
>
> http://www.ncl.ucar.edu/Applications/wrfpo.shtml
>
> as templates, I get nice looking plots, but unfortunately
> the data isn't in the right place. In particular, if I plot HGT
> (terrain height) I end up with some of Alaska's most significant
> mountains out in the ocean.
>
> I've played around a bit and found that if I set my center longitude
> to the value of STAND_LON, it looks a LOT better, but there's still
> a sizable offset.
>
> Does anybody know of a way to resolve this issue with NCL? It
> seems like, at least in the examples, there is no mechanism for
> specifying center and standard longitudes which are different.
>
> Thanks,
>
> Don
>
> --
> Arctic Region Supercomputing Center / U. Montana
> http://weather.arsc.edu/
> _______________________________________________
> ncl-talk mailing list
> ncl-talk_at_ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
_______________________________________________
ncl-talk mailing list
ncl-talk_at_ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Thu May 17 2007 - 10:49:55 MDT