Adaptation of the relational model to large volume geoscience databases

Over the last half decade, the Relational model has become the preferred model for Data Base Management System (DBMS) design and the Hierarchical model has come to be regarded as obsolete. The very large volume data sets from geophysical surveys are, however, intrinsically hierarchical - in both their data structures and in the data manipulation requirements of the digital compilation and cartography processes. Converting an interpolated map grid from its conventional form to a normalised relation, would (at least) double the data storage volume and (at least) quadruple the data access time. With only moderately sized surveys this is an unacceptable penalty. The relational model does, however, have distinct advantages over its predecessors in its clarity and simplicity of data definition and in the rigor with which data manipulation operations can be specified. These advantages could be of substantial benefit to geoscience data processing if the attendant disadvantages could be overcome. One means to accomplish this is to employ an Algebraic data structure model in conjunction with the Relational model. The Algebraic model treats relational domains as vectors. It too employs a rigorous formal definition of data structure and manipulation processes. The relationships between domains in the same relation and between different relations are defined by arithmetical operators used in a logical sense. The data base can thus be described by an algebraic expression and data manipulation processes can exactly be modelled by algebraic manipulation of this expression. The addition of the Algebraic model overcomes the above noted disadvantages of the Relational model as it permits Geoscience data sets to be viewed externally as fully normalised relations yet treated internally in their conventional, condensed form. (author). 8 refs