Model centred approach to scientific visualization

One of the crucial requirements for a Scientific Visualization system is to produce reliable and accurate results. There are many possible sources of errors that could jeopardise these efforts, such as measurement or simulation errors in the pre-analysis stage, or errors generated in the visualization process itself. Our focus here is to control errors introduced during the visualization processes. To this end we propose a conceptual model for visualization known as the Model Centred Approach (MCA). This new paradigm separates the modelling and viewing processes in visualization, and this provides the opportunity to consistently utilise a single modelling function throughout the visualization process. Results show that consistent visualizations are produced by our approach when compared to conventional methods.