Re: Why wind vectors could not be plotted in a sub-area with the same ncl codes?

Ed,

I created a pared version of your script and filed a ticket on this, NCL-1512.

I can't tell if there's a problem with the lat/lon coordinates, or with how NCL plots the vectors, but given that you can plot the vectors when you look at the full map area makes me think it's an NCL issue.

--Mary

On Jul 11, 2012, at 8:14 PM, Yang Yang wrote:

> Hi Mary:
>
> Thank you very much for your help!
>
> Please see my response below for your questions.
>
> "Why does your data have rlat0/rlon0 coordinate arrays, which I would have thought would be lat/lon arrays, yet you are using tlat and tlon on the file?"
>
> The original model outputs are on rotated lad and lon. The model is a regional configuration of the UK MetOffice UM model. I have used um2netcdf (developed by MetOffice) to transfered the rotated data to true lat and lon grids. The description of the data (yy.ctl) is attached, which was created by using "ncdump -h"
>
> "Also, where did you get the map projection parameters from (i.e. the two parallels)? I think you have a rotated grid, and the projection may not be set correctly for your data."
>
> I defined these parameters according to the region I want to plot. This may be an issue, but with small changes in the values of these parameters, It seems that the plotted figures didn't differ.
>
>
> "I noticed that the Corners resources in your two scripts were slightly different. If I use the values from the plot_2.ncl file in the plot_1.ncl file, then I see more vectors."
>
> The corners resources define the area I want to plot. The area in plot_2.ncl is larger, more vectors can be ploted, but cannot be plotted in a small sub-area in plot_1.ncl. This is exactly the question that I want to ask. The odd thing is, it seems that this problem didn't occur for contour plot (the vorticity).
>
> Right now, my question is: for the U and V data I provided on true lat-lon grids, what are the right map projection and resources to plot? The resolution is only 100m for the data. I want to plot wind vectors on a very small area: 1kmx1km. But my ncl codes don't allow me to do that.
>
> Cheers!
>
> Ed
>
>
>
>
>
>
>
> However, there is a strange seam line in your contour plot, like things are not lining up quite right. This is why I think there's something wrong with the map projection used.
>
> Please see the example on our rotated lat-lon grid page:
>
> http://www.ncl.ucar.edu/Applications/rotatedltln.shtml
>
> Note that the example uses values on the file to get the map projection exactly right, without using hard-coded numbers. You may want to try something like this, if you have all the necessary map parameter information on the file. I didn't see it on the file you gave me, but maybe the original file has it?
>
> --Mary
>
> On Jul 9, 2012, at 4:16 PM, Yang Yang wrote:
>
>> Hi Mary:
>>
>> I have sent you my data via the previous email. Attached please find the two pics plotted by the two ncl codes.
>>
>> Please let me know if you have not received the data.
>>
>> Cheers!
>>
>> Ed
>>
>>
>> --- On Tue, 7/10/12, Mary Haley <haley@ucar.edu> wrote:
>>
>>> From: Mary Haley <haley@ucar.edu>
>>> Subject: Re: Why wind vectors could not be plotted in a sub-area with the same ncl codes?
>>> To: "Yang Yang" <yang123yang@yahoo.com>
>>> Cc: "Dennis Shea" <shea@ucar.edu>
>>> Date: Tuesday, July 10, 2012, 1:22 AM
>>> Hi,
>>>
>>> Still no luck:
>>>
>>> -rw-rw-r-- 1 haley CIT\Domain Users
>>> 203812153 Jul 9 11:19 test.nc.gz
>>>
>>> I think the file is getting truncated every time you upload
>>> it. Do you have some place else you can put it,
>>> or can you try making it smaller? Perhaps you can write only
>>> the variables needed for plotting to another file?
>>>
>>> --Mary
>>>
>>>
>>> On Jul 8, 2012, at 3:40 PM, Yang Yang wrote:
>>>
>>>> Hi Mary and Dennis:
>>>>
>>>> Thank you very much for your help. I checked the
>>> compressed file again and had the same problem as yours.
>>> Sorry for the problem.
>>>>
>>>> I recompressed the file, which should be right this
>>> time.
>>>>
>>>> yangy@thotter:TEMP/ $ ls -il test.nc.gz
>>>> 3785158064 -rw-r--r-- 1 yangy niwa-users 308594775
>>> 2012-07-09 09:14 test.nc.gz
>>>>
>>>> I ftped the file again. If you cannot find a file named
>>> test.nc.gz or the file size is different, would you please
>>> let me know? This file is exactly the same file I used to
>>> plot.
>>>>
>>>> Cheers!
>>>>
>>>> Ed
>>>> --- On Fri, 7/6/12, Mary Haley <haley@ucar.edu>
>>> wrote:
>>>>
>>>>> From: Mary Haley <haley@ucar.edu>
>>>>> Subject: Re: Why wind vectors could not
>>> be plotted in a sub-area with the same ncl codes?
>>>>> To: "Yang Yang" <yang123yang@yahoo.com>
>>>>> Cc: "Dennis Shea" <shea@ucar.edu>
>>>>> Date: Friday, July 6, 2012, 10:08 PM
>>>>> Dear Ed,
>>>>>
>>>>> I downloaded your file, but when I tried to unzip
>>> it, I got
>>>>> the error:
>>>>>
>>>>> % gunzip xadala_pi00002011050215_utc.nc.gz
>>>>>
>>>>> gzip: xadala_pi00002011050215_utc.nc.gz: unexpected
>>> end of
>>>>> file
>>>>>
>>>>> The file is 205103998 bytes. Can you
>>> double-check that
>>>>> the file is valid on your end, and put it on
>>>>> anonymous ftp again if possible?
>>>>>
>>>>> Thanks,
>>>>>
>>>>> --Mary
>>>>>
>>>>> On Jul 5, 2012, at 10:57 PM, Yang Yang wrote:
>>>>>
>>>>>> Hi Dennis:
>>>>>>
>>>>>> Thanks for your help.
>>>>>>
>>>>>> I have ftped the file to the specified
>>> directory. It is
>>>>> a compressed file:
>>> xadala_pi00002011050215_utc.nc.gz
>>>>>>
>>>>>> Would you please let me know if you cannot find
>>> the
>>>>> file? Thanks!
>>>>>>
>>>>>> Ed
>>>>>>
>>>>>> --- On Fri, 7/6/12, Dennis Shea <shea@ucar.edu>
>>>>> wrote:
>>>>>>
>>>>>>> From: Dennis Shea <shea@ucar.edu>
>>>>>>> Subject: Re: Why wind vectors
>>> could not
>>>>> be plotted in a sub-area with the same ncl codes?
>>>>>>> To: "Yang Yang" <yang123yang@yahoo.com>
>>>>>>> Cc: ncl-talk@ucar.edu
>>>>>>> Date: Friday, July 6, 2012, 7:22 AM
>>>>>>> Nothing obviously wrong see 'diff'
>>>>>>> below'.
>>>>>>>
>>>>>>>
>>>>>>> It might be best to send the file to:
>>>>>>>
>>>>>>> ftp ftp.cgd.ucar.edu
>>>>>>> anonymous
>>>>>>> <use your email address for the
>>> password>
>>>>>>> cd incoming
>>>>>>> put xadala_pi00002011050215_utc.nc
>>>>>>> quit
>>>>>>>
>>>>>>> Send email to ncl-talk
>>>>>>> after you have successfully completed the
>>>>> transfer.
>>>>>>>
>>>>>>> ================================
>>>>>>> Rather than a double do loop
>>>>>>>
>>>>>>> do i=100, 500
>>>>>>> do j=100, 500
>>>>>>>
>>>>>>>
>>> xx=(v10m(i,j+1)-v10m(i,j-1))/200.
>>>>>>>
>>>>>>>
>>> yy=(u10m(i+1,j)-u10m(i-1,j))/200.
>>>>>>> vor(i,j)=xx-yy
>>>>>>> end do
>>>>>>> end do
>>>>>>>
>>>>>>> Why not use the builtin func
>>>>>>> dd = 200.0
>>>>>>> xx =
>>>>> center_finite_diff_n
>>>>>>> (v10m,200,False,0, 1)
>>>>>>> yy =
>>>>> center_finite_diff_n
>>>>>>> (u10m,200,False,0, 0)
>>>>>>> vor = xx-yy
>>>>>>> ================================
>>>>>>> %> diff plot_1.ncl plot_2.ncl
>>>>>>>
>>>>>>> 91c91
>>>>>>> <
>>> res1@gsnLeftString =
>>>>> "Plot_1 "
>>>>>>> ---
>>>>>>>>
>>> res1@gsnLeftString =
>>>>> "Plot_2 "
>>>>>>> 129c129
>>>>>>> < wks =
>>> gsn_open_wks
>>>>>>> ("eps","plot_1")
>>> ; open
>>>>>>> workstation
>>>>>>> ---
>>>>>>>> wks =
>>> gsn_open_wks
>>>>>>> ("eps","plot_2")
>>> ; open
>>>>>>> workstation
>>>>>>> 139,142c139,142
>>>>>>>
>>>>>
>>> < res1@mpLeftCornerLatF
>>>>>>> = -36.85
>>>>>>>
>>>>>
>>> < res1@mpLeftCornerLonF
>>>>>>> = 174.54
>>>>>>>
>>>>>
>>> < res1@mpRightCornerLatF
>>>>>>> = -36.6
>>>>>>>
>>>>>
>>> < res1@mpRightCornerLonF
>>>>>>> = 174.85
>>>>>>> ---
>>>>>>>>
>>>>> res1@mpLeftCornerLatF
>>>>>>> = -36.75
>>>>>>>>
>>>>> res1@mpLeftCornerLonF
>>>>>>> = 174.64
>>>>>>>>
>>>>> res1@mpRightCornerLatF
>>>>>>> = -36.65
>>>>>>>>
>>>>> res1@mpRightCornerLonF
>>>>>>> = 174.75
>>>>>>> 149c149
>>>>>>> < inc = 10
>>>>>
>>>>>>>
>>>
>>>>> ;
>>>>>>> plot at every "inc" grid point
>>>>>>> ---
>>>>>>>> inc = 2
>>>
>>>>>
>>>>>>>
>>>
>>>>> ;
>>>>>>> plot at every "inc" grid point
>>>>>>> =================================
>>>>>>>
>>>>>>>
>>>>>>> On 7/5/12 3:44 PM, Yang Yang wrote:
>>>>>>>> Hi:
>>>>>>>>
>>>>>>>> I posted the issue two days ago.
>>> Perhaps I
>>>>> didn't
>>>>>>> describe it well. I have not received any
>>> response.
>>>>> So I
>>>>>>> repost it here. Sorry!
>>>>>>>>
>>>>>>>> I used the same ncl codes to plot wind
>>> vectors
>>>>> in an
>>>>>>> area (~30x30km) and a sub-area (~10x10km)
>>> of it.
>>>>> The only
>>>>>>> differences are the parameters controlling
>>> the area
>>>>> to plot.
>>>>>>> However, the wind vectors can be plotted
>>> for the
>>>>> large area
>>>>>>> (this means my data are OK) but not on the
>>>>> sub-area. Can
>>>>>>> anyone help me fix the problem?
>>>>>>>>
>>>>>>>> plot_1.png was plotted by plot_1.ncl
>>> for the
>>>>> large
>>>>>>> area. plot_2.png was plotted for the
>>> sub-area by
>>>>>>> plot_2.ncl.The resolution of the data is
>>> 100m.
>>>>>>>>
>>>>>>>> Thank you very much!
>>>>>>>>
>>>>>>>> Ed
>>>>>>>>
>>>>>>>> --- On Wed, 7/4/12, Yang Yang<yang123yang@yahoo.com>
>>>>>
>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> From: Yang Yang<yang123yang@yahoo.com>
>>>>>>>>> Subject: Odd plot of
>>> wind
>>>>> vectors
>>>>>>>>> To: ncl-talk@ucar.edu
>>>>>>>>> Date: Wednesday, July 4, 2012, 8:14
>>> AM
>>>>>>>>> Hi:
>>>>>>>>>
>>>>>>>>> I have a regional model output
>>> (100m
>>>>> resolution).
>>>>>>> When I
>>>>>>>>> ploted an area about 30x30km
>>> (plot_1.png
>>>>> attached)
>>>>>>> using
>>>>>>>>> plot_1.ncl, the plot seems OK.
>>> However,
>>>>> when I
>>>>>>> tried to plot
>>>>>>>>> an area about 10x10km within the
>>> 30x30km
>>>>> area, the
>>>>>>> plotted
>>>>>>>>> wind vectors are quite odd
>>> (plot_2.png by
>>>>>>> plot_2.ncl), and
>>>>>>>>> wind vectors were not plotted in
>>> most
>>>>> areas. Can
>>>>>>> anyone help
>>>>>>>>> me fix the problem? Thanks!
>>>>>>>>>
>>>>>>>>> Ed
>>>>>>>>> -----Inline Attachment
>>> Follows-----
>>>>>>>>>
>>>>>>>>>
>>>>> _______________________________________________
>>>>>>>>> ncl-talk mailing list
>>>>>>>>> List instructions, subscriber
>>> options,
>>>>>>> unsubscribe:
>>>>>>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>> _______________________________________________
>>>>>>>>> ncl-talk mailing list
>>>>>>>>> List instructions, subscriber
>>> options,
>>>>>>> unsubscribe:
>>>>>>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>>>>>
>>>>>>
>>> _______________________________________________
>>>>>> ncl-talk mailing list
>>>>>> List instructions, subscriber options,
>>> unsubscribe:
>>>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>>>
>>>>>
>>>
>> <plot_1.png><plot_2.png><plot_1.ncl><plot_2.ncl>_______________________________________________
>> ncl-talk mailing list
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
> <yy.ctl>_______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Thu Jul 12 09:22:09 2012