Browsing by Author "Mpongo, Elton"

Filter results by typing the first few letters
Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Integrated seismic interpretation of the Larder Lake area, Southern Abitibi greenstone belt, Ontario, Canada
    (2020-10-21) Mpongo, Elton
    The southern Abitibi greenstone belt is characterized by a series of complex metavolcanic and metasedimentary rocks intruded by granitic plutons and batholiths, which makes the area a hardrock environment. In the Larder Lake area, these stratigraphic units are truncated by two major breaks, the Lincoln Nipissing shear zone (LNSZ) and Cadillac-Larder Lake deformation zone (CLLDZ) which trend NE-SW and E-W, respectively. This thesis is focused on the quantitative interpretation of a ∼44 km long seismic transect acquired as part of the Metal Earth project in the Larder Lake area. Seismic imaging and interpretation in a hardrock environment are challenging due to the lack of continuity of reflections and smaller acoustic impedance contrasts between different stratigraphic units. Hence, structural interpretation of the seismic data is favoured rather than stratigraphic interpretation. The application of curvelet transforms and seismic attribute analysis significantly increased the signal to noise ratio (SNR) of the seismic data. Seismic data have pitfalls in imaging the subsurface geology in a hardrock environment due to a strong degree of structural heterogeneity and complex geometry of targets. These pitfalls are overcome by the integration of seismic data with complementary geophysical methods. This study aims to determine the structural architecture of the area using seismic data and other depth resolving geophysical methods using an integrated modelling approach. The integrated modelling is achieved by extracting anomalies from the geophysical inversion models. The study uses the empirical relationship between physical properties derived from all the datasets and integrates these in a spatial domain. Physical properties from the individual models are differentiated by applying unsupervised learning algorithms for characterization. The recovered physical properties show correlations with the seismic data along deformation zones.