20000412: Problem reading MD file FTPed from UW archive (cont.)

>From: Mark Anderson <address@hidden>
>Organization: University of Nebraska-Lincoln
>Keywords: 200004111444.IAA02473 McIDAS-X 7.60 SFCPLOT ADDE

Mark,

>I put the files in the anonymous ftp account on zephyr under
>/pub/anderson/UNIDATA

OK, thanks.  I grabbed them a little while ago and have found out why
spatial plots can't be made from them:  LAT and LON values are missing
from the files!  Not only that, but the station elevation, ZS, state,
ST, and country, CO, are also missing from the file:

PTLIST AM/SFCOBS.1
Row :      1  Col :      1
TYPE        =         0             | DAY         =     99339 SYD         |
TIME        =         0 HMS         | NREC        =      1323             |
ID          =      NZCM             | LAT         = _missing_             |
LON         = _missing_             | ZS          = _missing_             |
ST          = _missing_             | CO          = _missing_             |
MOD         =         0             | HMS         =    235500 HMS         |
CIGC        = _missing_             | CC1         =         1             |
CC2         =         1             | CIGH        = _missing_             |
ZCL1        =     1500. FT          | ZCL2        =    15000. FT          |
VIS         =       7.0 MI          | WX1         = _missing_             |
WX2         = _missing_             | T           =    268.16 K           |
TD          =    265.16 K           | DIR         =       350 DEG         |
SPD         =       4.1 MPS         | GUS         = _missing_             |
PSL         =    964.78 MB          | PCP         = _missing_             |
SNO         = _missing_             | PRE         = _missing_             |
---------------------------------------------------------------------------
Number of matches found = 1
PTLIST: Done

I slowly realized that this was the problem after having correctly
produced meteorograms in the same way that you reported:

SFCMG KDEN DAT=AM/SFCOBS DAY=1999339

SFCMG doesn't do SELects on LAT or LON, so it never sees the problem
with the data files.  PTDISP, on the other hand, must do SELections on
the spatial locations of the data in order to do horizontal plots.

I don't know how these files got produced withoug LAT,LON information,
but they are obviously unusable as they stand.

I then FTPed to the UW archive machine to see what the extent of the
problem was.  I FTPed data from 1999082024 (August 20, 1999) and found
the same problem.  A data file from 1998030924 (March 9, 1998),
however, is OK.

The bottom line is that sometime after March 9, 1998, at least some of
the surface MD files in the Unidata-Wisconsin archive do not contain
needed LAT, LON, ZS, ST, or CO information.  The question is whether
all of the files are bad after that point?

There is a way that these files could be made usable; the procedure
would require me to write a script or better yet a program that would
replace the missing values in the file with information from the IDMSL
file (or MASTERID.DAT).  This wouldn't be too difficult, and I am
willing to do it if you really want the data.  Please let me know.

Tom