MLOC Native File Formats

MLOC Native File Formats

There are several distinct formats of data that can be read or written by mloc that are especially important. Although they are quite different in detail they are all considered to belong to the family of MLOC Native Formats or MNF; they are distinguished by version number:

  • MNF v1.3: Format for input of individual event files to mloc; also used for output of bulletins of the initial (automatic) and final (command datf) state of the current cluster.
  • MNF v1.4: Format for single-file output of all information related to an entire cluster, designed originally for import to the U.S. Geological Survey’s National Earthquake Information Center (NEIC) ComCat database, but also used for the GCCEL project
  • MNF v1.5 Format for input of differential time data

Detailed specifications for each format are found in other documents linked to the version numbers in the list above.

Version Number Family

The basic functionality and structure of the MNF formats listed above is described by the major.minor version format, but for the standard format v1.3 mloc keeps track of a third field which tracks certain changes in the format. The full version number of the MNF v1.3 “family” is currently v1.3.3 and mloc will issue a warning if it reads a file carrying an earlier version of the format. Depending on the details, a data file in an earlier format may be read and processed without a problem. No format check is done for input files of differential time data in v1.5 format but this would be added in the future if the need for a change in the format arises. Output files in MNF v1.4 format always contain a format line with the format version, currently 1.4.2.

Description and Usage

All three members of the MNF family of formats are fixed formats based on the common concept of a small set of distinct record types with different formats, identified by a single-character flag in the first column. Each record is a single line; each record type has a minimum line length determined by the defined fields (even if they are not all required). There are minimal constraints on the order in which different record types may occur. Each type of information (e.g., event, hypocenter, depth, magnitude, phase arrival) has a specific record type, and there are a few utility record types.

MNF v1.3 is by far the most important format version for mloc and non-specific references to “MNF” always refer to this version. It can be used to carry data for a single event, an earthquake catalog (multiple events, hypocenters only) or a seismic bulletin (multiple events, including phase readings), but for input to mloc only single event files are used. Data files with this format always have the filename suffix .mnf. The bulletin form of MNF v1.3 is used for the ~.dat0 output file created automatically by mloc to archive the input dataset for the run, and for the optional (command datf) ~.datf file that archives the final state of the cluster.

MNF v1.4 and 1.5 are special-purpose formats, used only for output and input, respectively. Output files in MNF v1.4 format are only created by the ccat command and input files in MNF v1.5 format are only read by the diff command. Files in MNF v1.4 format are created with the filename suffix .comcat. Files in MNF v1.5 format may have any naming convention.

Background

In its early development in the late 1980s, mloc used an event data format inherited from Ken Creager’s LOC program, on which mloc was based. In the late 1990s, when mloc began to be heavily developed for application in research on calibrated (“ground truth”) locations in close collaboration with Bob Engdahl, the ISC’s 96-byte fixed format (FFB), on which Engdahl had standardized his processing, was adopted as the default event data format for mloc. The FFB format had to be extended in certain ways to meet the needs of mloc (as well as Engdahl’s codes), and it was always an awkward format for this application, particularly in the fact that very different formats are used for initial and secondary phase readings, on the need to break up station codes longer than 4 characters, and in the strict requirements on ordering of records that is imposed by the concept of carrying in each record the record type for the following record.

As the FFB format was becoming deprecated at the ISC, replaced primarily by IASPEI Seismic Format (ISF), it became necessary and desirable to devise a format that specifically meets the requirements of mloc and which is more compatible with current practice in seismological data formats and exchange mechanisms. This results in data files that are reasonably compact and easy to process and edit in the ways required by the typical processing strategies that have proven to be effective with mloc.