Gravity Inversion Method Example
In general, the gravity profiles are modelled using the following procedure:
- Where seismic sections are available, the key horizons in the cover sequence and basement are picked from the seismic interpretation. If there are no seismic sections available, the models are based either on neighbouring profiles with existing seismic data, or on existing structural cross sections on each profile.
- The horizons derived from the seismic interpretation are depth converted and imported into ModelVision. These are then used unmodified to create an initial geological model between the surface and an arbitrary datum.
- Standard densities are assigned to these bodies. The values are either provided by the client (through the inspection of density logs from well completion reports) or taken from published data. For each profile, a first iteration using these constraints generates an initial model.
- The geometry of the horizons, and if required, densities are progressively modified and adjusted to find a fit between the observed and calculated gravity anomalies. In the profiles where seismic information is available, the shape of the uppermost horizons are not modified as they are believed to be quite accurate. The final model is the one with the minimum misfit between the observed and calculated gravity anomalies.
All depth estimation methods are limited by the inherent non-uniqueness of inverting from an observed field variation to the causative sources. The more complex the field variations, with effects from different sources that explain parts of the field variation, the greater the degree of non-uniqueness.
The quality and reliability of the depth estimates vary from survey to survey according to the quality of the data, the geometry and densities of the gravity bodies. Where available, modelling is done on line data. Otherwise, profiles are extracted by interpolation from the Bouguer gravity grid.
The initial model based on seismic information where available. Note that the fit between the observed (Black) and modelled (Red) profiles is poor.
Cross-section pre modelling
Cross-section after modelling
In order to produce a better fit between the observed and modelled results, the shape and average densities of the bodies are progressively modified and iterated in ModelVision. A close curve match indicates a high quality model. The regional gravity signature (pink) is calculated on each profile by a second order polynomial spline function of the gravity data.
The bottom section presents a set of polygons representing source bodies. These bodies are derived from depth converted seismic horizons where available or from sketch cross sections. In the example above, the solid lines in the bottom section show the original depth converted seismic horizons. The shape of these horizons are modified when building the model to get a better fit between the observed and modelled gravity curves.
In most of our project areas, the topography is taken from a compilation of client-provided and public domain data. Public domain topography data may exist at different resolutions ranging from 1 Arc second (approximately 30m) to up to 1 minute (approximately 2km) depending on the project area. When data is supplied by the client, it's quality is checked (QC) and may be stitched with existing public domain data in specific areas. FrOG Tech uses the most up-to-date and best possible topography data for project work.