Analysis (Integration and modelling)

Analysis, integration and modelling of gamma-ray spectrometric data are usually best supported by Geographical Information Systems (GIS) with some image processing capability. These systems offer a variety of analysis methods that can assist with the interpretation of gamma-ray spectrometric data. Images can be overlaid with polygons (e.g., soil, regolith or geological units), lines (e.g., faults) and points to explore the spatial relationships between radioelement abundances with other complimentary and disparate datasets. Gamma-ray grids can also be integrated into 3D GIS packages and interpreted with other sub surface datasets (see Figure 15 below).

Zonal statistical techniques can be used to explore the relationships between radioelement distributions with particular geological or regolith units. Mean differencing using geological units as regions of interest have been used to map regolith materials and associated salt stores (see Bethrungra 3D interactive model). Gamma-ray imagery are also used in GIS-based, quantitative environmental correlation approaches to map specific soil attributes. For example, Gessler et al. (1995) and Ryan et al. (2000) used statistical models to develop predictive relationships between terrain attributes and other explanatory variables such as gamma-ray datasets to map soil properties.

Standard edge-enhancement methods can be used to assist with the delineation of boundaries that may be related to changes in bedrock or soil type. Multi-variate linear regression techniques were used to separate bedrock and weathering gamma-ray responses to highlight potential areas of alteration associated with mineralisation (Dickson et al. 1996). Principal component analysis can be used for noise removal (Minty and Hovgaad 2002) as well as an exploratory tool for identifying the main trends in multivariate data. Clustering and classification (see section on Geological Mapping) are mapping techniques for automatically identifying areas of similar radiometric character. These provide unbiased generalizations that can be used as a basis for more detailed interpretation of the map data. Anderson-Mayes (2002) demonstrates how a combination of simple exploration analysis, principal components analysis and unsupervised classification can be use to model gamma-ray datasets with other geophysical and terrain datasets for land management application.

Help Rotate Zoom Pan Reset

Figure 15 (above). This interactive 3D VRML model shows the integration of gamma-ray grids with other datasets including DEM,
magnetics and conductivity depth imagery using 3D modelling software. 3D visualisation helps to facilitate a more holistic
approach to data integration and interpretation. Windows XP users may need to click in the white space to activate the
model if it does not appear on first load. Use the navigation buttons above to interact with the model. Click and drag
the top surface upwards to better view the solid geological units inside. Press the Help button for more information.