f Determination of the optimum gridding interval of a gravity data set

Gravity survey station locations are, in general, irregularly distributed . Although an ideal survey should be spatially homogeneous, this is, in most cases, a talk almost impossible to realize, mainly due to access problems . Moreover, sub-areas are often more densely sampled, either because of their greater geological interest or their ease of acces. This inevitably results in interpolation errors in the computation of a regular grid from this random data distribution . Korvin et al (1990) have shown that the gravity station distribution in South Australia can be characterised by its fractal dimension (1.4), and that, in this specific case, aliasing is unavoidable . The fractal dimension of a gravity network can be used to determine if an interpolated map is aliased at a specific interval and it is often also possible to determine an optimum gridding interval .