To realize a pragmatic collaboration and exchange of information between the participants of the Geological Modelling Project (GM) and the Advanced Seismic Modelling Project (ASM) of the EC GEOSCIENCE Project. For this purpose, the TNO Institute of Applied Geoscience will develop a general Data Exchange Format (DEF) for 3D subsurface modelling parameters.
The data in the GEOSCIENCE Project can be subdivided into three groups : well related data, seismic related data, and 1, 2 and 3 dimensional subsurface model data. These data represent geological as well as geophysical information generated in the Geoscience Project. In particular the 1, 2 and 3 dimensional subsurface model data will be of common interest to both the geological and geophysical disciplines in the Geoscience Project, and existing standards (SEGY, UKOOA, LIS, etc...) do not provide common storage and exchange facilities for such data. Therefore the design of the proposed Data Exchange Format (DEF) focuses on the method to properly store and exchange these kind of data. The data analysis includes the inventory of :
1) the data (Universe of Discourse)
2) commonly accepted definitions of the data
3) mutual relations or associations of data and constraints on these relations and/or data values.
The transformation of conceptual models to relational schemes includes a normalization phase resulting in table definitions, field formats and referendal and domain constraints. This scheme is extendible if needed. From the relational conceptual scheme the Data Exchange Format can now be derived. The mandatory information or data coming from different relational tables will be grouped into a sequential structure. Rules how to extend the Data Exchange Format with new data types (attributes) have to be given. In the output of extensive amounts of table information (e.g. surface grids, seismic picks, etc...) the optional information should be omitted, enabling a more condensed way of exchanging and storing data. The data analysis is carried out according to Nijssen's Information Analysis Method (NIAM). This method makes use of a sentence-predicate calculus, whereby the data is defined by sentences having deep structure, in the sense that they can be transformed into a variety of other representations (called s rface structures). The sentences may be presented as natural language statements, tables or graphs.
It is now possible to subsequently derive the relational table/field/ definitions and associations between tables from the information structure diagram in the fifth order normalized form (5NF). These tables can directly be used to fill the data dictionary of a RDBMS and to create the tables to store 1, 2 and 3D subsurfaces models. The information presented in the Data Exchange Format can be compared to information retrieved from a RDBMS into an output report. Exchange information has to be stored in one logical file which can be put on (optical) disc, tape or other media. This implies that information from different tables with one to many relationship to each other has to be assembled and reorganized into on file. The tables with general information and limited number of occurrences will be presented as header records in a header block. The original tables, objects and fields will be recognizable in the output.