Gravity Gridding in South Australia

Abstract

Creating an image of the acceleration due to gravity over a large area (in this case a large portion of South Australia) is not a trivial process. This paper examines some of the issues involved with creating such an image, and presents some examples. Treating the state gravity as a single dataset highlights outliers and results in ‘dimples’. These are short wavelength features around single points. Treating the data as a compilation of grids and then levelling the grids results in linear artefacts where the survey boundaries meet.

Two new approaches have been implemented, involving removing selected data points (based on proximity to adjacent points) and implementing variable density gridding techniques. The resulting grids still have artefacts (notable when viewing a first vertical derivative of the grid), but are smoother and more geologically plausible.