next up previous contents
Next: Input namelist and files Up: GRIB2V5D USER'S GUIDE Converting Previous: Installing and compiling   Contents


Running grib2v5d

The program grib2v5d reads an unique grib file, looking inside it for fields of variables specified in a namelist file, then outputs these fields in a Vis5d file. The namelist file is named by default v5d.naml. The philosophy of grib2v5d is to guess as much information as possible from the grib file, so grid parameters, vertical coordinate type and initial date are all deduced from the content of the grib file. All the other information must be provided in a namelist file described in chapter 4, however, for an initial use, that chapter can be skipped and one can directly refer to chapter 7 and use the standard namelist files provided, keeping most of the configuration to its default status.

Command-line arguments

The syntax for executing grib2v5d is:

grib2v5d [-n <namelist file>] [-i <grib file>] [-o <Vis5d file>]
[-mi <grib file1> [<grib file2> ...]]

All arguments are optional, if -n is not specified v5d.naml is used as namelist file; if -i or -mi are not specified, then the input grib file name should be set in the namelist names (see section 4.2), while if -o is not specified, the input file name with suffix .vis5d is used; if -mi is specified then all the following arguments, possibly specified by shell wildcards, are taken as file names, and each of them must coincide, in the right order, with the time levels specified in the namelist file.

Format of the grib input file

Since, as explained above, some information is deduced from the content of grib input file, grib2v5d makes some assumptions about the correctness of that file. If the input level type (typel, see chapter 4) is not specified in the namelist, the grid parameters, the vertical coordinate type and the initial date are determined from the first grib field of a requested variable at an upper air requested level (i.e. whose grib level type is one of 100, 103, 107, 109, 110, 119) which lyes on Arakawa H points. This means that, in order to avoid unpredictable results, all the fields in the grib file should have the same grid (apart from fields on U/V points of an Arakawa C-type grid for which the program makes a proper interpolation), initial date and vertical coordinate type (apart from surface fields). If typel is specified, then the field providing the information is looked for according to the previous rule while also matching the vertical level type provided. Once gathered this information from the file, the search of the requested fields is made sequentially with a possible rewind of the file, so the order of the fields inside the grib file is unimportant (provided that there is a unique field matching each set of parameter-level-date-forecast time in the file). For accumulated fields the forecast time matched is the one relative to the end of the accumulation period. When a requested field is not found in the file, a tentative explanation of what didn't match is printed by the program.

next up previous contents
Next: Input namelist and files Up: GRIB2V5D USER'S GUIDE Converting Previous: Installing and compiling   Contents