Developpement de techniques en gravimetrie vectorielle aeroportee pour l'exploration de ressources

A new geophysical method, vector gravimetry, is proposed for resource exploration. Raw data from an airborne gravity system based on inertial navigation system technology are processed and analysed showing the potential and viability of vector gravimetry in resource exploration. The main objective of this thesis is to extract the three components of anomalous gravity from raw airborne gravity system data and to validate airborne vector gravity data for resource exploration in the absence of reference (ground) data. The secondary objectives are to develop simple vector gravity interpretation tools and to investigate the possibility of improving the spatial resolution of the three components of the airborne vector gravity anomaly.; Multi-level wavelet analyses of raw airborne gravity system data are carried out. The goal of the analyses is to better understand the content of the raw data and the propagation of errors in the GPS and accelerometer data toward the three components of the gravity anomaly. Wavelet analyses of the three components of gravity anomaly are also carried out. The analyses show that the main source of error limiting the amplitude resolution of the horizontal components is platform misalignment. Nonetheless, the horizontal components can be use, together with the vertical component (scalar gravimetry), in geophysical interpretations of the geology.; The amplitude resolution of vector gravity anomaly data is estimated by comparison of their Laplaciens with the Laplacians of synthetic data containing known noise. The estimated errors in the horizontal components are greater than the estimated error in the vertical component. For example, the standard deviations of the errors in deltagx, deltagy and deltagz, for the Timmins survey (Ontario), are 0.07 mGal, 0.1 mGal, and 0.05 mGal, respectively. To improve the amplitude resolution of the horizontal components, the estimation of errors in gyroscopic measurements, used in feed-back loops to maintain platform alignment, is required.; Simple vector data interpretation tools are proposed: vector plots, the gradient tensor and Euler deconvolution. In the Euler deconvolutions, joint inversion of the three components is performed to obtain solutions of the 3D coordinates of anomalous bodies. In theory, joint inversion of the three components provides more accurate solutions than independent inversions of the three components. Vector plots also allow locating anomalous bodies in 3D but appropriate anomaly separation in the three components is required. The gradient tensor calculated from vector gravity data does not offer the same spatial resolution as measured gradients, but has the advantages that all nine elements of the tensor are automatically available and the low frequency content of the gradients is preserved.