NOAA / NWS's Meteorological Development Laboratory
Degrib (NDFD GRIB2 Decoder)
Subject: Degrib 1.40
Date: Wed, 24 Dec 2003 17:47:29 -0500
From: Arthur Taylor
Hi (Happy Holidays)
You all have the unique distinction of being registered on my degrib (aka
NDFD_GRIB2Decoder) mailing list (all 270 of us) (I'd like to welcome: tony,
jon.mills, webyoda, guy.super, mark.jackson, brian.mcinerney, mvevea,
olivier.van.keulen, cmontes, klawsteve, paul_m_ingram, wiscwx, greg.martin,
colle006, yetzer, bboyer, sunholz, dl1050, greg, qwedq, kirby.devore,
gsstrassberg, tarunareddy81, abcdmnop, jbreier, markt, mesaofc, richard,
jkellysr, ctarunaredddy, duncalf, marktorregrossa, smccormick, msullivan,
kari.sheets, develop, svmbpetkovic, Dmaineult, sstraley, fingersbrumley,
ASoto, Mel.Newman, tom.kent, rob, mjg7876, rod.golightly, mickes,
mbousliman, rlepage, ebashi, xingbao, dulal_chakrabarti, tim, Mark.Fenbers,
wmiler, john.schattel, mma, and kelly.whitaker)
Please let me know (email me) if you want to unregister. The latest version
of the program is on: www.nws.noaa.gov/mdl/NDFD_GRIB2Decoder/
----------- General News.
First, some of you may have noticed that the sector grids shifted around
some this week. That was because, thanks to Kari Sheets, we discovered that
the sector points did not line up with the Conus points. This was due to an
error in the packing routine, which has now been fixed. The lower left
corners of the sectors should now be the same as the corresponding Conus
point within 4 or 5 decimals.
Please check if you hard-wired your lat/lons. If you did so, I would
recommend you update them to what is currently being sent in the GRIB2
----------- News about this release:
I have just finished putting together version "1.40" (last one that I
announced was "1.31" on 10/24/2003).
This release contains a DRASTIC shift in the unpacker, so compilation is a
bit different, hence the shift from 1.31 to 1.40. (If you just use the
MS-Windows executable, you probably won't notice anything) I have now put
an "archive" link on the "download.htm" page, so you can get the old release
if need be.
1) I merged the NCEP GRIB2 compression methods with the "official NWS" GRIB2
compression. NCEP introduced to the WMO the idea of using PNG / Zlib
compression (ie gzip) as well as JPEG2000. JPEG2000 comes from the jpg
image format, which can be a lossy compression scheme, but it is possible
(which NCEP did) to choose certain parameters so that no data is lost.
Unfortunately the NCEP library requires 3 other libraries (zlib, libpng, and
jasper), all of which are open source projects. To make it easier for
people to compile degrib, I have included the source for a copy of zlib, and
libpng. In addition I have included the relevant parts of the jasper
library in the "jpeg2000" directory.
NOTE 1: You do not need to use the copy of zlib, libpng, or jasper that I
provide. The URLs for those packages are located in "/degrib/version.txt",
and they may already be on your system.
NOTE 2: You should look at the licenses for those libraries! If you don't
agree with the terms, it is possible to compile and use degrib without using
them, you just won't be able to decode NCEPs JPEG2000 compressed GRIB2
Since there are now 5 libraries and a driver that need to be compiled, I
have created some makefiles in "degrib/src" which should "make" each piece,
and then install it. This should simplify the compilation process,
unfortunately I could only create makefiles for the systems I have access
to. If you create a makefile for a system that I don't have, could you
please send me a copy.
The advantage of all this: NCEP is now encoding their ETA model data in
JPEG2000 GRIB2, as well as their .5 degree world data. "degrib" officially
only supports NDFD, but for those out there who are applying "degrib" to any
GRIB message you can lay your hands on, this will give "degrib" a chance.
More work needs to be done to fully support the ETA model data, but this is
the first step.
2) Non-Projected grids. Degrib can handle most lambert conformal map
projections (as well as polar stereographic and mercator). It was not able
to handle pure lat/lon grids. It now can, but some work is needed to make
sure that the .shp files it creates are in the range of -180..180 longitude
instead of 0...360. Note: the primary use for this feature is to handle
world model data which is on a lat/lon grid. NDFD so far sticks with
3) Dateline issue. With Alaska about to join the NDFD, some work is needed
to split polygons at the dateline. Previously ESRI could not handle "big
polygons" that degrib creates if they crossed the dateline. The effect in
ESRI was weird streaks across the map. Now polygons that cross the dateline
are split at the dateline. Note: Some work needs to be done to clean up
exactly where they get split, and similar work needs to be done for the
"small polygons", prior to Alaska going public.
Please let me know if you have problems.
(Art in MDT (Mountain Daylight Time) [:-)] )
("Big A" for Glasgow's WCM [:-)] )