Decoder Packages

Robb Kambic
Unidata Program Center
September 2005

The decoder packages are used to manage GRIB and text data by either decoding the data into netCDF files and not storing the GRIB files or by storing the GRIB files and decoding the data on the fly. Recently the size of the GRIB files has been increasing with more parameters and levels; therefore it is more efficient to store the GRIB files for later decoding.

The netCDF Decoders package converts WMO GRIB products and textual bulletins into netCDF formatted files. This package usually works in conjunction with the LDM package to convert incoming products to the netCDF format without the need to store the raw data. The gribtonc decoder converts most NCEP models into GRIB format. There are perl decoders for the following NWS textual report
types: METAR, buoy, upper air, and synoptic products. This package requires the installation of netCDF, UDUNITS, and the Perl-netCDF packages to work.

The GribJava decoder package does not convert GRIB products into netCDF but has three possible utilities: dump the contents; index the products in the file; or extract the data on a parameter basis. The index gives a concise overview of the file contents that can be used in conjunction with the data extraction routines to provide quick data access. GRIBJava is written in Java so it can run on all platforms where Java is installed.

Work : netCDF Decoders package

During the last year modifications were mainly of bug fixes for the netCDF Decoders package and modifying the cdl files (netCDF templates) to accommodate the changes to the NCEP models. The last remaining work is changing the netCDF file format to be CF (netCDF Climate and Forecast Metadata Convention) conventions compliant. There have been three releases in the last year.

Work : GribJava decoder package

The package is still in beta stage, so it is still in bug-fix mode and implementation of options that haven't been coded. The code is used by the netCDF Java 2 package to decode GRIB1 and GRIB2 files. Work that needs to be complete before the release is taken out of beta stage includes: scanning the code for errors and creating Java Docs.