The scale to which modelling is undertaken varies from the global scale to the mine scale. Data are usually provided with a coordinate system ranging from latitude-longitude-elevation coordinates to mine grid coordinates; including the specific AMG system (Australian Map Grid).

Modelling of large scale areas (eg northern margin of the Australian plate) needs, for obvious reasons, to be performed on a spherical projection, but the ability to digitise boundaries of units, or structural elements in a map view or a cross-section view is still needed. New functions have been developed allowing to switch between the following projections (Fig. 4):

1. geodetic,

2. AMG,

3. homolographic (equal area sinusoidal),

4. spherical.

For each of these projections, elevation is still displayed and allows for a better visualisation in any of the projection system.

In the case the seismicity in the Banda Arc region, building an envelope surface bounding the hypocentres required the development of a new function allowing (1) to slice a complex set of points in smaller sets, and (2) represent the selected points in a 2-dimensional space (Fig. 5). Digitisation of a bounding line for each of the slices was performed using common gOcad® tools. The bounding surface was consequently built by linking together each of the lines.

Slicing the set of points is done according to a user-defined value (± tolerance) of a user-defined property assigned to the points (eg longitude). Points falling within the interval [property-tolerance, property+tolerance] are consequently represented in a new space where the slices can be represented as a cross-sections. In figure 5, the new space is defined by X = latitude, Y = longitude and Z = elevation (or depth).