20001102: McIDAS-X 7.6 SFCPLOT

>From: Mark Tucker <address@hidden>
>Organization: Lyndon State
>Keywords: 200011021724.eA2HOE400457 McIDAS-X 7.6 SFCPLOT

Mark,

>We've found some strange behavior with the SFCPLOT command and hoped that
>you could possibly shed some light on how to fix it.  Here's some examples
>of what we have tried and the results:                
>       
>SFCPLOT GUS OLAY LATEST INT=00:30 DAY=2000307 MODE=X GRIDF=132 GRA=12 
>SF=YES PRO=CONF OUT=PLO GU=GRAPHIC BLANK=NO
>SFCPLOT: Dataset contains more than one schema 

The error indication suggests that at least one of the files in the
RTPTSRC/SFCHOURLY dataset is somehow damaged or contains a schema
different than one or more of the other files in the same dataset.
At this point, I would list out the file information from the dataset
with:

PTLIST RTPTSRC/SFCHOURLY.ALL FORM=FILE

and see if the schema type in the file list varies at all.  I just tried
this while pointing to your server and did not see the problem:

DATALOC ADD RTPTSRC cirrus.lsc.vsc.edu
PTLIST RTPTSRC/SFCHOURLY.ALL FORM=FILE
Pos      Description                        Schema  NRows NCols  Date
------   --------------------------------   ------  ----- ----- -------
     5   SAO/METAR data for   31 OCT 2000   ISFC       72  4500 2000305
     6   SAO/METAR data for   01 NOV 2000   ISFC       72  4500 2000306
     7   SAO/METAR data for   02 NOV 2000   ISFC       72  4500 2000307
     8   SAO/METAR data for   03 NOV 2000   ISFC       72  4500 2000308
PTLIST: Done

While pointing at cirrus, I ran the invocation you list in a 7.613 McIDAS-X
session and got the correct plot of wind GUSts:

SF 12
EG
MAP USA
SFCPLOT GUS OLAY LATEST INT=00:30 DAY=2000307 MODE=X GRIDF=132 GRA=12 SF=YES 
PRO=CONF OUT=PLO GU=GRAPHIC BLANK=NO

Is it possible that there happened to be a bad surface file in the
RTPTSRC/SFCHOURLY dataset when you ran this program and got the error
and that file is now gone?  Is cirrus the ADDE server you are running from?

>SFCPLOT GUS OLAY LATEST INT=00:30 DAY=2000307 MODE=X GRIDF=132 GRA=12 
>SF=YES PRO=CONF OUT=PLO GU=GRAPHIC BLANK=NO DATASET=RTPTSRC/SFCHOURLY.7 
>PTDISP: No data found matching search conditions 

This invocation worked for me also:

EG LEV=3;SFCPLOT GUS OLAY LATEST INT=00:30 DAY=2000307 MODE=X GRIDF=132 GRA=12 
SF=YES PRO=CONF OUT=PLO GU=GRAPHIC BLANK=NO DATASET=RTPTSRC/SFCHOURLY.7

>  This has the same results regardless of the (valid) position of
>  RTPTSRC/SFCHOURLY
>
>SFCPLOT GUS OLAY 12 INT=00:30 DAY=2000307 MODE=X GRIDF=132 GRA=12 SF=YES
>PRO=CONF OUT=PLO GU=GRAPHIC BLANK=NO
>SFCPLOT: Dataset contains more than one schema
>SFCPLOT - Done
>SFCPLOT failed, rc=1

This really does look like there was a bad datafile in the RTPTSRC/SFCHOURLY
dataset that you were pointing at.  Again, what is/was the ADDE server
when you ran this invocation?  To check, run:

DATALOC LIST

on the machine on which you encountered the problem.

>SFCPLOT GUS OLAY 12 INT=00:30 DAY=2000307 MODE=X GRIDF=132 GRA=12 SF=YES
>PRO=CONF OUT=PLO GU=GRAPHIC BLANK=NO DATASET=RTPTSRC/SFCHOURLY
>SFCPLOT - Done
>
>  It works!!

This worked for me also.

>SFCPLOT GUS OLAY 12 INT=00:30 DAY=2000307 MODE=X GRIDF=132 GRA=12 SF=YES
>PRO=CONF OUT=PLO GU=GRAPHIC BLANK=NO DATASET=RTPTSRC/SFCHOURLY.7
>SFCPLOT - Done
>
>  It works!! (for any valid position)

This invocation worked for me as well.

>Without specifying the dataset we get a schema error and the command
>fails.

At this point, I would appreciate to see the output of:

DATALOC LIST

PTLIST RTPTSRC/SFCHOURLY.ALL FORM=FILE

>In addition, the LATEST parameter does not seem to work.

LATEST was introduced for use in the non-ADDE access surface plotting
routine.  It was kept for backward compatibility, but it does not necessarily
give you back a plot of the latest avaiable data for the day.

>The data
>does exist and the last two successful examples I've listed behave as one
>would expect.

The problem for me is that I was unable to reproduce your failure while
pointing my 7.613 session at cirrus.  I need to know if you were getting
the results while pointing to a different machine.

Tom