The MIT Voyager Plasma Science Data Documentation

The Voyager data are available in different places, but the same structure and naming conventions are always used. This allows programs to find the wanted data automatically without the user having to know the conventions. It also allows the data to be moved/updated without the changing of using/producing programs. The data are stored in a directory tree with a known base. For MIT internal purposes, this base is pointed to by the environmental variable vgrDATA ; on the www, it is Voyager Data Files ; for the same data on the plasma group machines, it is the voyager data directory .  (Since this directory is exported read-only by faraday, these files can only be modified from faraday.) The   directory structure   from the given base directory is

Space Craft, [ v1, v2]

Type of Data, [edr, sum, vgr, spl, shortsedr, sedr, ha, key, const, misc]

Below this, the structure depends on the type of data at which we are looking. EDRs are subdivided into [ q, f] for quick look and final, which are then divided into separate directories for each year. Not all directories are present in each tree, and occasionally, other nonstandard directories are present. In these directories, usually at the highest level, sit the data files. Data files, when configured in the final way, have a very rigid naming convention so that they may be picked up by the data processing programs automatically. The names of the data files are as follows:
SC_DT_TIME1_TIME2[.SUF]

SC is the Space Craft involved, [ v1, v2]

DT is the Data Type, [sum, edr, ha, key ...] ( Note that this must not disagree with the data type in the data type in the directory tree name, although it may be a modification of it. "keys" for key short, will be in the directory tree under key. A key parameter file of hourly averages will be under ha/key, while a high time resolution key parameter file will be under just "key".)

TIME1 is the time of the beginning of the data (I try to skip over spurious start times in making this determination).

TIME2 is the time of the ending of the data (I try to skip over spurious end times in making this determination). Times are in the format of yyyy.ddd.hh, where yyyy is the four digit year, dddd is the three digit day of year and hh is the two digit hour of day.

.SUF, if present, will indicate postprocessing such as data compression [ .Z, .gz].

The detailed format of the data types not generated at MIT will not be covered here. These are mostly input data types and are too complicated to address in this document. There is a memo (written in troff) describing the Summary Tapes. This memo's chapters are mostly concerned with the Voyager analysis program, vgranl . Most of the file types generated at MIT have man pages written for them, and can be accessed with a call to man (e.g. man key). The directory "const" contains the files needed for mjsanal to run Alan Barnett's analysis, as described in his Ph.D. thesis.

Those binary data types that are produced at JPL (EDR & SEDR), are described in the JPL documentation, of which only a very brief summary is on line. A more complete description is available for Summary Tapes. The binary files produced at MIT ( Hourly Average Tape, vgr Tape, and spl Tape), are available on line.

For the ASCII files, which are generated at MIT, there is documentation with man pages, as well as in the idl file common.types. This file is used by idl and several other programs to recognize different data types. The most common type of data file is the key parameter file, along with its derivatives. The major variation of this file type is how many fields are in the file, although there are some other minor variations possible. In each case the name of the type of file is "key*", where the "*" indicates the type of variation.