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

20010216: IMGPROBE of NEXRAD Level III product data

>From:  Russ Dengel <address@hidden>
>Organization:  SSEC
>Keywords:  200102161418.f1GEIGL16751 McIDAS-X ADDE nexraget

Russ,

>This is a problem I'm seeing of our Nexrad server test machine
>(breeze). IMGPROBE of a nexrad image is giving me invalid values of
>DBZ and VIP (from a N0R product)
>
>
>    Image Name       ID       Day      Nominal Time   Scan Time    Band
>  ----------------    ---    -------    ------------   ---------    ----
>
>  NEXRAD/BREF1        LCH 16 Feb 01047   12:59:00      MISSING       1
>
>                        File     Nominal  Image     RAW     BRIT
>ECHO    VIP
>       Lat/Lon      Line/Element  Line/Element                      DBZ
> 31:36:58 / 94:39:40    63/   92      64/   93         4
>645266304      6

This is weird and also something I have not seen.

>... So, I went into the code to try to fix it. I noticed the
>DEBUG flag in nexrutil.h and set its value to 1. After  I recompiled and
>linked nexraget and nexradir I get resonable values for the DBZ and
>VIP values
>
>
>    Image Name       ID       Day      Nominal Time   Scan Time    Band
>  ----------------    ---    -------    ------------   ---------    ----
>
>  NEXRAD/BREF1        LCH 16 Feb 01047   12:59:00      MISSING       1
>
>                        File     Nominal  Image     RAW     BRIT
>ECHO    VIP
>       Lat/Lon      Line/Element  Line/Element                      DBZ
> 31:07:20 / 94:41:07   118/   89     119/   90         6        96
>30      1

Hmm...  Very strange indeed.

>... I went back to nexrutil.h and turned DEBUG off ( value
>changed to 0 ). and the errors returned.

OK.

>Are you seeing this on your machines also?????

No, but I decided to try turning debug on like you did to see what
would happen.  Apparently something is getting stepped on somewhere and
turning on DEBUG changes the address space so that it doesn't show up
the same way (i.e., appears to work with no error).

I am checking into this to see what I can find.  I will bounce back to
you when I find out what is going on.

Tom

>From address@hidden Fri Feb 16 11:32:17 2001
>Subject: Re: 20010216: IMGPROBE of NEXRAD Level III product data

OK, I'm working on it here also.
Russ

>From address@hidden Fri Feb 16 11:40:10 2001
>Subject: 20010216: IMGPROBE of NEXRAD Level III product data (cont.)

Russ,

Well, almost immediately after sending off my previous email I saw and
fixed a mistake in nexrutil.c.  The problem was introduced when I went
back and "cleaned up" some code.

In CalibrateNexrImgData, I had the following:

#if DEBUG
    for ( i = 0; i < nlevel; i++ ) {
      levels[i] = auxblk[43+i];
      (void) sprintf( dbg, "levels[i]: %d", levels[i] );
      M0sxtrce( dbg );
    }
#endif

It should have been:

    for ( i = 0; i < nlevel; i++ ) {
      levels[i] = auxblk[43+i];
#if DEBUG
      (void) sprintf( dbg, "levels[i]: %d", levels[i] );
      M0sxtrce( dbg );
#endif
    }

After making this change, an IMGPROBE of data from motherlode (adde.ucar.edu;
dataset RTNEXRAD) works with DEBUG turned on or off.

Also, while looking through nexrutil.c, I found a sprintf invocation within
a DEBUG clause that would cause the C compiler on Solaris SPARC to choke.
I corrected this also.

You should grab a new version of nexrutil.c from the pub/mcidas directory
of anonymous FTP on our ftp server, ftp.unidata.ucar.edu.  I put
this new copy in an update compressed tar file that contains a current
snapshot of what is in my distribution: nexrad.tar.Z, nexrad.list.

Please let me know if this fixes the problem you are seeing.

Tom

>From address@hidden Fri Feb 16 11:46:47 2001
>Subject: Re: 20010216: IMGPROBE of NEXRAD Level III product data (cont.)

I'll get it NOW.
Russ