[ncl-talk] Visualization of SAF NWC products

Athanasios Karagiannidis thankar at live.com
Thu May 28 02:05:05 MDT 2015


Good
morning to all!

 

Following the
earlier conversation (to which I respond) I think I have the obligation to
clarify some issues, since these conversations are available to all users, and
they are used to solve various kinds of problems. 

After doing
some more tests, I found out that the otherwise elegant and technically
flawless script that was kindly prepared for me, does not illustrates correctly
the CTTH_HEIGHT field. The problem is related to the fact that the line/column
space of the raster data does not refer to a regular lat/lon space. In simple
words the pixels of the raster data are not equal in lat/lon dimensions.

 

Regarding
more technical issues: 

XGEO_LOW_RIGHT,
XGEO_UP_LEFT, YGEO_LOW_RIGHT, YGEO_UP_LEFT are not given in degrees but in meters.

 

dims =
dimsizes(CT) , lat = fspan(minLat, maxLat, dims(0)) and lon = fspan(minLon,
maxLon, dims(1)) are producing  a regular
grid (which is not correct)

 

It would be
really helpful if a script could project correctly Meteosat fields (I think safnwc
is using the Meteosat grid) but I’m not sure that such a thing exists.

 

Kind
regards

 

SakisAthanasios F. Karagiannidis
Physicist, PhD Meteorology and Climatology
E-mail: thankar at live.com

From: thankar at live.com
To: haley at ucar.edu
Date: Sat, 21 Feb 2015 08:29:02 +0000
CC: ncl-talk at ucar.edu
Subject: Re: [ncl-talk] Visualization of SAF NWC products




Mary, thank you for your input!

Sakis

Athanasios F. Karagiannidis
Physicist, PhD Meteorology and Climatology
E-mail: thankar at live.com

Date: Fri, 20 Feb 2015 09:46:02 -0700
Subject: Re: [ncl-talk] Visualization of SAF NWC products
From: haley at ucar.edu
To: thankar at live.com
CC: huangwei at ucar.edu; ncl-talk at ucar.edu

The issue with "namedcolor2rgb" is not really an issue. The code should still work.  The error message is just misleading, but can be ignored.
We have fixed the incorrect error message for the V6.3.0 release coming next month.
--Mary

On Fri, Feb 20, 2015 at 1:46 AM, Athanasios Karagiannidis <thankar at live.com> wrote:



Wei,

Your help is more than appreciated.

I will update my NCL version.

Kind regards

Sakis

Athanasios F. Karagiannidis
Physicist, PhD Meteorology and Climatology
E-mail: thankar at live.com

From: huangwei at ucar.edu
Date: Thu, 19 Feb 2015 09:23:14 -0700
Subject: Re: [ncl-talk] Visualization of SAF NWC products
To: thankar at live.com
CC: ncl-talk at ucar.edu

Sakis,
Attached is script which looks more like what you see with HDFview,but used a whole user-defined color table.
"gsdtol" is a standard color-table name, in NCL 6.2.1. Your NCL of 6.1.2is a little bit old, and I strongly suggest you update your NCL to the newest version.
Compare the newly attached script and the script before,you'll be able to figure out that the difference is the projection applied.So you can turn on/off those options based on your requirement.

A new png file is attached as well.
Regards,
Wei

================================================1850 Table Mesa Dr.Boulder, CO 80307Phone: 303-497-8924

On Thu, Feb 19, 2015 at 4:07 AM, Athanasios Karagiannidis <thankar at live.com> wrote:



Dear Wei

I'm more than thankful for your script! It is excellent in correctly reading my data and more or less what i wanted regarding the visualization.
I have just 3 comments:
1. I would really like to see the drawn field (in example the CT) even over land. How can i do this?
2. When i run the script in my machine i get the following warnings:

(0)    Warning: namedcolor2rgb: 'gsdtol' is not a valid named color.
(0)    Will return missing values for this color.

Is 'gsdtol' a custom made color? In this case maybe i should replace it.

3. In the example HDFView visualization i provided the grid was not "tilted" to the left (a projection issue i guess!). How should i make your visualization, so it looks like the HDFView one?

Thank you in advance!

Sakis

Athanasios F. Karagiannidis
Physicist, PhD Meteorology and Climatology
E-mail: thankar at live.com

From: huangwei at ucar.edu
Date: Wed, 18 Feb 2015 13:26:21 -0700
Subject: Re: [ncl-talk] Visualization of SAF NWC products
To: thankar at live.com
CC: ncl-talk at ucar.edu

Sakis,
You may take a look of the attached script and pdf file to see if these are what you wanted.
Thanks,
Wei

================================================1850 Table Mesa Dr.Boulder, CO 80307Phone: 303-497-8924

On Wed, Feb 18, 2015 at 1:38 AM, Athanasios Karagiannidis <thankar at live.com> wrote:



Dear sirs/madams

I want to visualize SAF NWC output data using ncl. They are included in an h5 file.
These data are not in a simple lat-lon grid.
After an earlier contact with you i managed to visualize them by creating such a grid, but i feel that the image is not accurate.
What i want to help me with, is to tell me weather there is a standard way or an already existing script to visualize correctly that kind of data.
To help you I provide the following:

1. A sample datafile.
2. The printVarSummary output
3. A screenshot of the HDFView visualization of the variable where the Variable info are shown (more or less the same info with printVarSummary).
4. The script i used (although, i do not think that you need that!). The lat/lon limits that i have n my script are not acurate. I used them to make the visualization.


---------- printVarSummary------------------------------

ncl 2> printVarSummary(CT)

Variable: CT
Type: ubyte
Total Size: 370688 bytes
            370688 values
Number of Dimensions: 2
Dimensions and sizes:    [DIM_000 | 512] x [DIM_001 | 724]
Coordinates: 
Number Of Attributes: 11
  CLASS :    IMAGE_INDEXED
  ID :    CTTH_HEIGHTED
  IMAGE_COLORMODEL :    RGBH_HEIGHTED
  IMAGE_SUBCLASS :    IMAGE_INDEXED
  IMAGE_VERSION :    1.0GE_INDEXED
  N_COLS :    724
  N_LINES :    512
  OFFSET :    -2000
  PALETTE :    ��8�3
  PRODUCT :    CTTHE_INDEXED
  SCALING_FACTOR :    200

---------------------------------------------------------------------------------

---------------- My Script ---------------------------------------------------

load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"   
load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"    
load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl"

begin
cdf_file = addfile("SAFNWC_MSG2_CTTH_200812121100_Greece______.h5","r")
CT = cdf_file->CTTH_HEIGHT(:,:)

  res                             = True

   lat    = fspan(50,20,512) ; A.K. -- reverse the matrix
   lon    = fspan(10,30,724)
   CT!0   = "lat"                   ;-- set name of CT dimension 0
   CT!1   = "lon"                   ;-- set name of CT dimension 1
   CT&lat =  lat                    ;-- assign lat array to CT
   CT&lon =  lon                    ;-- assign lon array to CT
   CT&lat at units = "degrees_north"  ;-- set the units of the named coordinate lat
   CT&lon at units = "degrees_east"   ;-- set the units of the named coordinate lon

res at gsnAddCyclic = False
res at tiMainString         = "Default Color"    ; main title
res at cnFillOn             = True               ; turn on color fill
 res at mpMaxLatF                   = 50           ; choose subregion           
 res at mpMinLatF                   = 20
 res at mpMaxLonF                   = 30
 res at mpMinLonF                   = 10

xwks = gsn_open_wks("pdf","CloudTops")
plot = gsn_csm_contour_map(xwks,CT,res)
end

--------------------------------------------------------------------------------

Thank you in advance.


Sakis




Athanasios F. Karagiannidis
Physicist, PhD Meteorology and Climatology
E-mail: thankar at live.com 		 	   		  

_______________________________________________

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 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20150528/b65dacb1/attachment.html 


More information about the ncl-talk mailing list