Search

The scant difference in the electromagnetic (EM) reflectivity of barefaced terrain often imposes challenges in differentiating between such terrain types and deployment of synthetic aperture radar to oil sand exploration. Microwave remote sensing has a proven ability to provide valuable information about targets. However to derive geoscientific information, a profound understanding of the EM interaction with terrain is vital. The challenge is to identify scattering characteristics relevant to oil sand fields. While various terrain identification methods and signature databases have been developed in the optical domain, only few examples of barefaced terrain discrimination in the microwave domain have been reported. In this thesis a three step multi-sensor approach has been used to identify EM signature of barefaced terrain encompassing homogeneous and heterogeneous materials, in the optical and microwave range. The combined method also led to the development of a large database of hyperspectral reflectivity, dielectric and backscattering data relevant to geointelligence analysis. The geochemical signature identification and prediction (GSIP) process required spectral data acquisition, chemometric model implementation and postprocessing to determine the spectral fingerprints and components of two strains of Nigerian oil sands. The results were compared with available hydrocarbon databases and four new features of Nigerian oil sands were observed. The dielectric discrimination statistical model (DDSM) involved three studies of the dielectric properties of oil sands and other barefaced terrain with different weight percentage of moisture and statistical processing of data to identify the 1 - 2 GHz and 5 - 7 GHz as most suitable frequency bands for microwave imaging. The GSIP and DDSM provided new empirical data on the geochemical and electrical behaviour of oil sand particularly the contrasting effects of bitumen, sand and moisture. Finally computer EM (CEM) models of barefaced terrain and sensors were used to identify the backscattering behaviour of the terrain for analysis in 2D/3D format. The results provided good agreement with classical surface roughness models particularly the Surface Perturbation and Kirchoffs Scattering model. They also enabled the investigation of the effect of wide variations in the sensor and terrain parameters on backscattering in order to evolve a radar signature necessary for identification of oil sand terrain for petroleum exploration. A laboratory scatterometer system (LSS) was developed and deployed in three imaging scenarios to verify aspects of the derived microwave EM signature of the terrain. The LSS measurements and the results from the CEMs were complimentary. more...