Re: Irregular grids in NCL

From: Szymon Jaranowski <szymon.jaranowski_at_nyahnyahspammersnyahnyah>
Date: Thu Aug 23 2012 - 01:22:55 MDT

Dear Mary,
I've verified lat/lon values are read and prepared OK
("grid_var_zoom.png" file in the first mail is a part of the plot of
all points of lat/lon coordinates).

I've added
  print(min(lat) + "/" + max(lat))
  print(min(lon) + "/" + max(lon))
lines to the script - output:
(0) 44.0284/66.8
(0) 0.210999/38.734
looks OK.

I've uploaded 3 data files "LAT_VAR", "LON_VAR", "landmask_pl_full"
(ca 5.5 MB each) (ftp://ftp.cgd.ucar.edu/incoming) and attached
updated script "land.ncl" to this mail so You can run it now. Runtime
is about 10 minutes with commented "res@tfDoNDCOverlay = True" (on my
laptop); I'm using self-compiled NCL version 6.0.0 (but also tried
compiled one with the same effect).

-- 
With best regards,
Szymon Jaranowski
On Wed, Aug 22, 2012 at 3:54 PM, Mary Haley <haley@ucar.edu> wrote:
> It looks like there may be an issue with your lat/lon arrays.
>
> Did you verify that the values look okay?
>
>   print(min(lat) + "/" + max(lat))
>   print(min(lon) + "/" + max(lon))
>
> Can you provide your new script and data files on our ftp site so I can try running it myself?
>
> http://www.ncl.ucar.edu/known_bugs.shtml
>
>  Thanks
>
> --Mary
>
> On Aug 22, 2012, at 3:57 AM, Szymon Jaranowski wrote:
>
>> Dear Mary,
>> I've commented the option; the result is very strange (attachment).
>> Just like the coordinate systems of a map and data were completely
>> different (the data is in geographic coordinates).
>>
>> --
>> With best regards,
>> Szymon Jaranowski
>>
>>
>> On Tue, Aug 21, 2012 at 5:26 PM, Mary Haley <haley@ucar.edu> wrote:
>>> Dear Szymon,
>>>
>>> In land_var.ncl, I noticed you are setting:
>>>
>>>  res@tfDoNDCOverlay        = True
>>>
>>> When you do this, you are effectively telling NCL that you've already set up the
>>> map projection to exactly match the projection of your data, and hence it ignores
>>> any settings of sfXArray and sfYArray.
>>>
>>> Try commenting this line out and rerunning your script.
>>>
>>> --Mary
>>>
>>> On Aug 21, 2012, at 6:06 AM, Szymon Jaranowski wrote:
>>>
>>>> I'm trying to draw a simple landmask using NCL.
>>>>
>>>> First on a regular grid - it works fine.
>>>> Some images (attached):
>>>>
>>>> 1. Plot of grid's nodes (plot of geographic coordinates): grid_reg.png
>>>> 2. Result: land_reg.png
>>>> 3. NCL script: land_reg.ncl
>>>>
>>>> Then I'm doing the same on a second grid. The second grid covers the same area,
>>>> but is more dense and irregular. Zoom of upper left corner:
>>>>
>>>> 4. Zoom of upper left corner: grid_var_zoom.png
>>>>
>>>> As seen on the image, nodes are irregularily spaced, but still this grid is
>>>> a cartesian product of 1D grids. The effect:
>>>>
>>>> 5. Result: land_var.png
>>>> 6. NCL script: land_var.ncl
>>>>
>>>> It seems sfXArray and sfYArray options are ignored - the data is plotted just
>>>> like the spaces between the nodes where the same. Is it possible to draw plots
>>>> on such grids in NCL?
>>>>
>>>> --
>>>> With best regards,
>>>> Szymon Jaranowski
>>>> <grid_reg.png><grid_var_zoom.png><land_reg.ncl><land_reg.png><land_var.ncl><land_var.png>_______________________________________________
>>>> ncl-talk mailing list
>>>> List instructions, subscriber options, unsubscribe:
>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>
>> <land_var_noverlay.png>
>

Received on Thu Aug 23 01:23:07 2012

This archive was generated by hypermail 2.1.8 : Thu Aug 23 2012 - 16:16:15 MDT