[Met_help] METv0.9 testing

Matt Clark mclark at rush.aero.org
Thu Jul 19 10:02:12 MDT 2007


I've d/l'd the MET tarball, and have compiled/installed on a Linux system,
using GCC compilers g++ and gfortran.

There were a couple of compile glitches, and I've attached the changes I
made, as a patch file.  The changes are trivial, but required to meet
standard FORTRAN syntax (other compilers may have different default
settings, but these changes will be transparent.....)

I've encountered problems running the pb2nc tests;  the input files are 
expected to be "blocked FORTRAN", which is very system-dependent, and 
generally an outmoded form. 

I assume that the files were created on some system which 
supports this I/O style, and this is likely to be non-portable, making it
difficult to read arbitrary binary files. (Not to mention problems with
byte-order, word length, etc).

What tools produce the input files for pb2nc (the PrepBufr point obs
file)?

Thanks,

Matt Clark
Aerospace Corporation
(310) 336-1205
-------------- next part --------------
--- readpb.prm.orig	2007-07-18 13:07:23.000000000 -0700
+++ readpb.prm	2007-07-18 13:07:50.000000000 -0700
@@ -17,6 +17,7 @@
 	PARAMETER	( MXSTRL = 80 )
 C*					Maximum size of a string
 C*
+        INTEGER		FILEMXSTRL
 	PARAMETER	( FILEMXSTRL = 256 )
 C*					Maximum length of a file name
 C*
--- readpb.f.orig	2007-07-18 13:09:40.000000000 -0700
+++ readpb.f	2007-07-18 13:10:00.000000000 -0700
@@ -9,7 +9,8 @@
 
 C-----------------------------------------------------------------------
 C-----------------------------------------------------------------------
-	SUBROUTINE READPB  ( lunit, subset, idate, iret, cnlev, chdr, cevns )
+	SUBROUTINE READPB  
+     *          ( lunit, subset, idate, iret, cnlev, chdr, cevns )
 C
 C*	This subroutine will read and combine the mass and wind subsets
 C*	of the next station report in the prepbufr file.  It is styled


More information about the Met_help mailing list