| commit | 050f959eb060fd4a9e23df1404daae454cf45188 | |
| author | Jim Bresch <bresch@ucar.edu> | |
| Thu, 24 Jan 2013 00:44:41 +0000 (24 00:44 +0000) | ||
| committer | Jim Bresch <bresch@ucar.edu> | |
| Thu, 24 Jan 2013 00:44:41 +0000 (24 00:44 +0000) | ||
| tree | 362a67ed8bd3d0f6bc26c070a12653a03946a676 | treesnapshot (tar.gz zip) |
| parent | c2125dc4926067b2d7e0de4a0d95a9e9176d7ceb | commitdiff |
Bug fix for the decoding of the time-level in rd_grib2.F. rd_grib2 uses
the forecast time of the first field in the grib2 file as the time for
all fields in that file. If an accumulated or time-averaged field was
first, the time was decoded wrongly. The problem was discovered with
CFSR flx files, but could conceivably occur with any grib2 file.
Ungrib assumes only one fcst time per file. So far, all datasets we've
encountered are stored this way. Ungrib would have to be re-written if multiple
forecast hours are ever put in a single file.
Make a minor fix to the netcdf bit determination in configure.
git-svn-id: https://svn-wrf-wps.cgd.ucar.edu/trunk@715 86b71a92-4018-0410-97f8-d555beccfc3a
the forecast time of the first field in the grib2 file as the time for
all fields in that file. If an accumulated or time-averaged field was
first, the time was decoded wrongly. The problem was discovered with
CFSR flx files, but could conceivably occur with any grib2 file.
Ungrib assumes only one fcst time per file. So far, all datasets we've
encountered are stored this way. Ungrib would have to be re-written if multiple
forecast hours are ever put in a single file.
Make a minor fix to the netcdf bit determination in configure.
git-svn-id: https://svn-wrf-wps.cgd.ucar.edu/trunk@715 86b71a92-4018-0410-97f8-d555beccfc3a
| configure | diffblobblamehistory | |
| ungrib/src/g2print.F | diffblobblamehistory | |
| ungrib/src/rd_grib2.F | diffblobblamehistory |