Recent technological advances have created volumes of data such that, unless some effective methods are used to analyse them, they will be either wasted or under‐examined for their useful information content. Scientific data visualization is an attempt to use graphical and numerical tools to extract information contained in data and hence to allow its analysis. This paper seeks to present a systematic approach to the development of tools for scientific data visualization.
It is shown that the approach to implement these tools involves four major steps: description of a reference model, validation of the data process, description of the software component and the design and implementation of the visualization tool.
This approach is substantiated by defining conditions suitable for scientific data visualization processes, in a relaxed manner. These conditions are subsequently refined more formally. Definitions and theorems of the proofs are succinctly discussed.
The mathematical description of the visualization process is necessary to understand and maintain some significant reduction in errors in scientific visualization processes.
