A standard data format has been established for experiment, instrument, ocean, acoustic and tomography data as well as for intermediate processing results, upstream compatible with the software package and downstream compatible with the end user (exploitation) requirements. For the tomography data in particular, a set of standard levels of processing has been defined starting with raw data and ending with inversion results: 

Level-0 data: Raw acoustic data as read from the instrument storage medium without any compression or processing. 

Level-1 data: Complex correlated acoustic data (using a replica of the emitted signal) and, additionally, data with optional processing such as normalization, windowing and over-sampling, as well as navigation and clock-drift correction (output from pre-processing toolbox). 

Level-2 data: Estimated arrival times from the acoustic data and, optionally, identified or un-identified tracks of arrival times through the experiment (output from estimation/identification toolbox). 

Level-3 data: Slice-inversion results in the form of sound-speed mode amplitudes and sound-speed / temperature distributions obtained from the previous-level acoustic data (output from inversion toolbox). 

The experiment, instrument, ocean, acoustic and tomography data as well as all intermediate and final results are stored in NetCDF format, which is a self-descriptive, machine-independent format for representing and sharing array-oriented scientific data; NetCDF files can be accessed through the NetCDF library and interface (details for the NetCDF format can be found under http://www.unidata.ucar.edu/packages/netcdf/). 

The flow of information between the various toolboxes of TOMOLAB takes place through NetCDF files referring to :