[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20000617: Several GEMPAK questions



>From: Christian Page <address@hidden>
>Organization: UCAR/Unidata
>Keywords: 200006180318.e5I3IET04469

>
>Hi,
>
>  I am a GEMPAK newbie, and I have a few questions about it.
>
>1- For European metar observations, I must use PALT when I want to plot mean
>sea-level pressure, as opposed as PMSL for North America metar obs plots? Are
>these 2 equivalents?

I think you mean "ALTM" for altimeter setting reduced to sea level. "ALTM"
should be Altimeter setting converted to station pressure in millibars.

 GEMPAK-SFLIST>r
 PARM = PMSL;ALTM;ALTI;PALT;SELV                                                
 

    STN    YYMMDD/HHMM      PMSL     ALTM     ALTI     PALT     SELV
    YGK    000619/1200   1022.10  1022.02    30.18  1010.80    93.00
    YHU    000619/1200   1020.80  1020.67    30.14  1017.40    27.00
    YMX    000619/1200   1020.90  1020.67    30.14  1010.78    82.00
    YND    000619/1200   1021.10  1021.00    30.15  1013.28    64.00
    YOW    000619/1200   1021.00  1020.67    30.14  1006.95   114.00
    YOY    000619/1200   1020.80  1020.33    30.13  1000.17   168.00
    YQB    000619/1200   1020.40  1020.33    30.13  1011.53    73.00
    YSC    000619/1200   1021.20  1020.67    30.14   991.84   241.00
    YUL    000619/1200   1020.60  1020.67    30.14  1016.32    36.00

In the examples shown above, you see that ALTM is reasonably close to 
PMSL. Usually, stations that report PMSL use a local formula that takes
the 12 hour average temperature in to account for the aircolumn
that would exist from the station elevation to sea level. So, calculating
ALTM from the reported ALTI will not have this advantage.

PALT is not equivalent to PMSL...it is the station pressure, so is
dependent on the station elevation (SELV).


>
>2- I would like to see an example on how to plot, from metar raw obs, maps of
>mean sea-level contours with station plots overlayed.

To contour the data, you will first have to create an objective analysis
gridded data set from the surface file.

Use either oagrid or gdcfil to create an empty grid file, then oabsfc to
grid the parameters at the time you need. Then, you can display the station plot
obs with sfmap and overlay contours with gdcntr.

http://www.unidata.ucar.edu/packages/gempak/tutorial/barnes2.html
http://www.unidata.ucar.edu/packages/gempak/tutorial/barnes3.html
http://www.unidata.ucar.edu/packages/gempak/tutorial/barnes_exercises.html

Once you have the surface data gridded, then you can use the gd programs
to contour or perform calculations on the grid. Overlaying data is discussed at:

http://www.unidata.ucar.edu/packages/gempak/tutorial/overlay.html

On example using Garp to overlay gridded data on station model plots is shown 
at:
http://www.unidata.ucar.edu/packages/gempak/tutorial/garp.html

>
>3- Is there anything I can do to display rivers out of US? I am in montreal
>Canada, and the big St-Lawrence river is not even plotted, neither Montreal
>Island. This is quite difficult to locate plotted stations as opposed as
>geographical positions.

Try:
$mapfil = states.uni
The St. Lawrence seams to be reasonably well depicted, though the island
still isn't visible. The .uni is a file I created here at Unidata which
covers US, Canada and Mexico.

>
>Thanks a lot for your great support!
>
>Christian Page      finger address@hidden => tel. + adresse
>address@hidden    http://www.sca.uqam.ca/
>
>Assistant de recherche / Research Assistant
>Universite du Quebec a Montreal / Universite McGill / McGill Radar
>
>

Steve Chiswell