Re: wmstnm improvements

Thank you Mr. Clare for your response.

In the WMO Guide on the Global Data-processing System:

http://www.wmo.ch/pages/prog/www/DPS/Manual/WMO485.pdf

In Volume I, Part II, Appendix II-4: GRAPHICAL REPRESENTATION OF DATA,
ANALYSES AND FORECASTS (pages 49-63) is a detailed description of the
surface plotting model (the SYNOP symbol) and there it reads on
page 50:

VV - Horizontal visibility at surface, the code figures are plotted.

I can also point you to this German guide:

http://www.4robert.de/synop/synop.html

Norwegian weather page pointing to this guide:

http://www.mir-co.net/wetter/wetterkartensymbole.htm

We are also using other plotting packages and all of them plot the
visibility as it is in the SYNOP code. I would think that since you are
including the option of using the metric system (UNT 1) for your
customers outside of the US, it would be better to follow the standard.

Regarding the rain/snow dot size: it is not a showstopper, I'm plotting
postscript and then converting to png for the web. I had to increase the
size of the images to 1600x1200 points for the symbols to be easily
distinguishable. You will maybe keep this in mind if later on symbols
will be redigitized.

Best regards,
Bolli

On Tue, Dec 15, 2009 at 01:09:01AM -0700, Fred Clare wrote:
>
> On Dec 13, 2009, at 6:26 AM, Bolli Palmason wrote:
>
> >Dear NCL users,
> >
> >thanks to NCAR for a very nice graphics package!
>
> Thanks for the comment.
> >
> >I'm using the wmstnm function to plot SYNOP data and there are two
> >things that could improve (I'm using UNT 1 - metric):
> >
> >1. The visibility should be the numbers from the SYNOP code, not
> >translated to km. 70 in SYNOP code should be plotted as 70, not 20
> >(km).
>
> In implementing wmstmn I encoded the data as per the SYNOP Data Format
> (FM-12) as described at:
>
> http://weather.unisys.com/wxp/Appendices/Formats/SYNOP.html
>
> which indicates visibility is encoded in km. What source for SYNOP
> codes are you using? Since wmstnm has been in our package for over
> fifteen years with the visibilities coded in km and yours is the first
> suggestion to code as per another scale, I am hesitant to change the
> code at this time. I could define a user-settable parameter that
> would control how visibility is encoded. In the meantime, you could
> define a function that converts km to the SYNOP codes you want to use--
> this would not be precise, but it would be close.
> >
> >2. The dots/stars for rain/snow should be bigger or different size, it
> >is hard to see the difference.
>
> I can see that at small sizes there could be difficulty in
> distinguishing these symbols. These symbols were digitized from scaled
> up symbols that appear in NOAA's publication "Explanation of the
> Weather Map." Are the dots/stars for rain/snow the only ones you find
> that are causing confusion? Again, I am hesitant to redigitize those
> as to make them inconsistent with the sizes of dots and stars used
> elsewhere, and redigitizing all of the dots and stars would be a bit
> of bother. Also, the spacings would have to be readjusted. What
> sizes do you normally plot these symbols at and what is your output
> device? Plotting the symbols to a Postscript file and printing seems
> to resolve the symbols fairly well for me. All of these characters
> are drawn as lines and filled areas and not as raster images scaled
> for device-specific resolutions; that could definitely cause confusion
> on lower-resolution output devices.
> >
> >Hope NCAR considers these changes, must be bothering others also :)
>
> You are the first to bring these issues up, but I can see your point
> and will file this for future reference.
>
> Thanks for your input.
>
> Fred Clare
>
>
>
> >
> >Best regards,
> >Bolli
> >_______________________________________________
> >ncl-talk mailing list
> >List instructions, subscriber options, unsubscribe:
> >http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>

-- 
Bolli Palmason
IT Department
Icelandic Meteorological Office
Tel. +354 5226103
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk