NOAA / NWS's Meteorological Development Laboratory
Degrib (NDFD GRIB2 Decoder)
Subject: Degrib 1.74
Date: Mon, 31 Oct 2005 07:57:34 -0500
You all have the unique distinction of being registered on my degrib (aka NDFD_GRIB2Decoder) mailing list (all 892 of us). (I'd like to welcome: dwbauer, stephen.cofield, saurin.shah, desrochers.claude, ciau,
wa7tib, lee, kbryson, matthew.powe, rafael, kevin.pace, gfuselier,
mhess, atorres, indaplains, michael.montefusco, mlilly, mouse16jp,
diego, paul.kettle, frigon.anne, ijarman, mbutt, paul, ejneafsey,
hmuller, jeremycamp, nightsky68, cermak)
Please let me know (email me) if you want to unregister. The latest
version of the program is on www.weather.gov/mdl/degrib/ (or
----------- General News:
John Jones announced on Friday 10/28/2005 that on 12/8/2005 the NWS will
transition Wind Direction and Wind Speed forecast elements from
experimental to operational for CONUS, Puerto Rico/Virgin Islands,
Hawaii, and Guam.
On 10/25/2005 the valid selection range of the "NDFD Custom Grid Sector"
changed from a 15 degree by 15 degree latitude / longitude box to a 20
degree by 20 degree latitude / longitude box.
On 10/21/2005 the NDGD increased to include GFS model MOS (Model Output
Statistics) for PoP06, PoP12, and Relative Humidity (primarily for the
----------- News about this release:
I have just finished putting together version "1.74" (last one that I
announced was "1.73" on 10/14/2005).
This release focused on adjusting the custom sector to 20 x 20, introducing 3 new NDGD variables, several NetCDF bugs, and handling time from before 1970.
1) At the request of fire weather emergency managers via Andy Bailey,
the size of the custom grid sector for the NDFD was adjusted from 15x15
degrees to 20x20 degrees. To accommodate this, tkdegrib, and web.tcl
needed some changes in their validation routines.
2) Added PoP06, PoP12, and Relative Humidity to the NDGD. This meant
reconfiguring the control file for tkdegrib and web.tcl.
3) NetCDF bug fixes:
3A) Thanks to Kirby Cook, found an error with my test to see which
flavor of my "-NetCDF 1,2, or 3" was used to create a file. This
resulted in randomly allowing only 1 message per file for "-NetCDF 2,3"
(-NetCDF 1 was designed to only allow 1 message per NetCDF file).
3B) Thanks to Anne Frigon, found out that my method of using the msgNum
to specify which element of the ProjHour coordinate variable to set, was
non-CF compliant. It could result in msgNum - 1 missings, followed by a
valid projection hour. Have now improved "-NetCDF 2" and "-NetCDF 3" to
insert the projection into the correct place in the ProjHour coordinate
variable, hopefully eliminating missing values from the ProjHour
variable, and keeping it monotonic. This should also help keep the
files small, as the variable grids are (X x Y x ProjHour) in size, and
this will keep the ProjHour size to exactly the number of decoded messages.
3C) NetCDF: For -NetCDF 2 and -NetCDF 3 with multiple grids, have
improved the "valid_min, valid_max" attributes to represent the min and
max of all the grids (rather than just the first one).
3D) NetCDF: Thanks to Anne Frigon, found a typo in the NetCDF
attributes: LowerLeftLatiude -> LowerLeftLatitude
4) Thanks to Kevin Pace, was able to get some reanalysis data from
6/1/1952. This caused problems since degrib used "time.h". The problem
is that on some platforms, some of the functions in time.h apparently
assume that dates stored as seconds since 1/1/1970 in a long integer
would be positive. To resolve this, I've switched to using my "clock.c"
code / library which stores dates as doubles since the start of the
epoch, and was designed to handle negative dates. This should allow for
times since the theoretical beginning of time, to close to the end of
time, although it currently doesn't handle calendar switching (ie from
the Gregorian to the Julian calendar). Let me know if you have GRIB
data that should be on the Julian calendar.
5) Thanks to Claude Desrochers, degrib now has sgi makefiles.
6) Added a GetURL.tcl script which when used from the "/degrib/bin"
directory by "tcldegrib web.tcl
downloading data (GRIB or other types) from the web.
7) Added the GRIB1 "parm_table_nohrsc" (NOHRSC = National Operational
Hydrologic Remote Sensing Center") for Chad Shafer who was working with
snow cover maps.
8) Added some capability to handle multiple input GRIB files. Currently
it can handle the "-I", "-C" options. Need to work on the "-P" option.
9) Fixed some bugs with the "-C -Map" option, but is not quite complete.
Please let me know if you have problems.