20010906: McIDAS; SFCMG and off-hours reports

>From: Christian Page <address@hidden>
>Organization: Universite du Quebec a Montreal
>Keywords: 200109060800.f8680j120055 McIDAS-X SFCMG

Christian,

>I try to plot with SFCMG some meteograms. It works fine for stations that
>report metars around :00. Example:
>http://www.sca.uqam.ca/cgi-bin-loc/get_meteogram?LFBO  (France Toulouse)
>http://www.sca.uqam.ca/cgi-bin-loc/get_meteogram?CYUL  (Canada Dorval)

OK.

>But if I try for Geneva, Switzerland:
>http://www.sca.uqam.ca/cgi-bin-loc/get_meteogram?LSGG
>I get almost no data. Listing corresponding METARs show that reports are done 
>at :20 and :50. Is this the problem?

The reports at :20 are specials.  This can be seen from a PTLIST listing
(this is an abbreviated listing created from the Data Lists action under
the Information dropdown list in the MCGUI):

SFCLIST LSGG DATASET=RTPTSRC/SFCHOURLY SPAN=DAY DAY=2001249 TIME=7 19 OPTION=ALL
 Day Time StCo  Stn   T  Td  Dir Spd Gus AltSet  Vis  Weather  Ceil  Cld1  ...
     hhmm        id  [F] [F]     [ kts ]  [mb]   [mi]          [8ths/100s o...
---- ---- ---- ----- --- --- --- --- --- ------ ----- -------- ----- ----- ...
   6 0650   SW LSGG   48  43 000   0     1020.0  7.00                3/035 ...
S  6 0720   SW LSGG   52  45 100   1     1020.0  7.00                3/035 ...
   6 0750   SW LSGG   54  45 150   1     1020.0  7.00                3/035 ...
S  6 0820   SW LSGG   54  45 070   3     1020.0  7.00                3/030 ...
   6 1050   SW LSGG   59  45 070   5     1019.0  7.00                3/035 ...
S  6 1120   SW LSGG   59  45 020   6     1019.0  7.00                3/035 ...
S  6 1220   SW LSGG   61  43 090   5     1019.0  7.00                3/035 ...
S  6 1320   SW LSGG   63  43 010   2     1018.0  7.00                3/035 ...
   6 1350   SW LSGG   63  41 080   3     1018.0  7.00          5/100 3/035 ...
S  6 1420   SW LSGG   63  41 060   4     1018.0  7.00          5/100 3/080 ...
S  6 1520   SW LSGG   61  41 060   3     1017.0  7.00          5/100 3/080 ...
S  6 1620   SW LSGG   61  43 070   3     1017.0  7.00          5/110 3/070 ...
   6 1650   SW LSGG   61  43 050   3     1017.0  7.00          5/110 3/070 ...
S  6 1720   SW LSGG   61  45 060   1     1017.0  7.00          5/110 3/040 ...
   6 1750   SW LSGG   59  45 050   1     1017.0  7.00 W        5/100 3/038 ...
S  6 1820   SW LSGG   59  46 000   0     1017.0  7.00 W        5/055 3/038 ...
S  6 1920   SW LSGG   57  50 080   2     1018.0  7.00 RW-      5/040 3/010 ...
Number of reports = 17
SFCLIST: done

The entries above that are specials are taged with an S in column 1.  You
can see that these are all from :20 past the hour.

SFCMG is just plotting the routine (hourly for METARs; three-hourly for
SYNOPTIC obs) reports only.

>Also, for LFBO, reports are also available at :30. Are they included in
>the plot?

Since these are also specials, they are not included.

>Can the INT keyword be set so that it plots all the data?

No.

>I tried with INT=0 or INT=0.5 or INT=ALL without success...

I looked into the code and see that all special reports are thrown
out after the data has been received from the server.  It seems to me
that one might like to be able to specify that all observations are
to be plotted, not just the standard hourly observations.  Given this,
I hacked up sfcmg.c to see if this could be done.

The results of the test can be seen in the followin:

standard SFCMG invocation:

http://www.unidata.ucar.edu/staff/tom/gifs/sfcmgstd.gif

enhanced SFCMG
http://www.unidata.ucar.edu/staff/tom/gifs/sfcmgall.gif

Since I believe that this is what you are looking for, I will roll the
new sfcmg.c code into my next addendum for 7.80, 7.802.  You can
upgrade your system to use the new version of SFCMG as follows:

<login to your workstation as 'mcidas'>
cd mcidas7.8/update
ftp ftp.unidata.ucar.edu
  <user> umcidas
  <pass> XXXXXX
  cd unix/780/bugfix
  binary
  get mcupdate.tar.Z
  quit
./mcunpack
cd ../src
make mcx
make install.mcx

After installing the new code (last step above), you should give the
new SFCMG a quick test:

SF 1
SFCMG LSGG
SF 2
SFCMG LSGG TYPE=ALL
LB 1 2
TERM L ON

The new addition is the TYPE= keyword.  This allows one to produce either
the standard plot (TYPE=STD, TYPE=X; this is the default) or one with
all observations on it (TYPE=ALL).

Let me know if this is not what you had in mind.


Tom