The REA directory contains phase readings and derived source information like hypocenters, fault plane solutions etc. The REA directory has one or several subdirectories corresponding to separate databases (see Figure 2.1 for an example with two databases). The database names can have between 3 and 5 characters. If less than 5 characters are used, the character `_' is added in the file system to make it 5. The user does not have to put the `_' when running a program, they will be added by the software. If a directory is made manually, the `_' must be put in. It is assumed that a database is always present in the system. The name of the default database is given by an environmental variable (see section 3.1), however if not set, it will default to AGA for agency. Here, BER will be used as an example throughout the manual. A database has a duplicate storage of the events. For quick reference and interactive work the events are stored in single files (S-files) in yearly directories and monthly subdirectories. When new data is entered into the database, it comes in as individual event files. However, once the interactive work has finished, the single event files are overwritten with the final location and additionally stored in monthly files, which are only changed when updating (UPDATE command, see section 14). The monthly files, called CAT-files for catalog, are stored separately in the CAT directory and primarily used for quick searching and backup for the single files. In addition to the event data, there is also a LOG directory in each database to keep a log of the data processing, see section 14.
S-file database structure
The structure for the single file storage is as follows (Windows example):
|\REA\BER__\||Main readings directory, all data|
|\REA\BER__\1999\||Data for 1999|
|\REA\BER__\1999\01\||Data for January 1999, each event in one file|
On Unix, the last line would have been /REA/BER__/1999/01
The structure works back year 0000. Each event contains original phase readings in the
Nordic format (
) which includes file
names of all corresponding waveform files. One event is one file.
Each event has an ID line. The ID line contains a unique ID, which
will follow the event through all COLLECT and SPLIT operations
(see section 12 and 13). The ID line
also contains status information about the event like last action,
when it was updated etc. The ID-number can be fixed, which is useful
if data is taken out from the database, processed on another computer and later put back into the database, since otherwise the ID of an event might be changed and the existing file would not be overwritten. An example of an S-file name is :
The S-files are used as input for the location program and, when making a permanent update, also for output, see 7. The letter in front of the "." indicates the event type and can be L, R or D for local, regional or distant event respectively. It is the same indicator as given in the header line of the S-file, see the Nordic format page . The remaining numbers give (in order) day, hr, min, sec, year and month.
As mentioned above, the system can contain many other databases, which may function exactly like the BER directory. A data base can be used to store a subset of data or data from different networks. Data can be moved between databases or in and out of the databases, for details, see description on EEV (5.3 and 5.4).
Monthly location files, the CAT directory
Events located in monthly files are in a
in addition to the individual S-files. Additional databases like
e.g. NAO will have epicenters stored under
/SEISMO/REA/NAO__/CAT. The monthly epicenter files are called 199901.CAT for e.g. January 1999. Although the files generated by SEISAN normally are monthly files, the CAT directory can also contain yearly files or any other time interval. The only rule is that the name of the file must give the year and month of the first event in the file. This is because the search program SELECT uses the file names to search requested time intervals. If a user has a historical catalog, this can be added as an individual file. If the historical catalog starts in 1820, the file name would be 182001.CAT. The files in CAT do not need to be continuous in time, but they must not have overlaps in time and each file must have data in chronological order. The format of the CAT files is the same as for the S-files. Additionally, CAT files can also be compact files, meaning just the header lines of the S-files (see also section 2.3).