A NEW METHOD FOR COMBINING GRAVIMETRIC AND GEOLOGICAL DATA

Application of the gravity method to structural geology makes it possible to highlight local variations of the gravitational field that are caused by density heterogeneities related to the composition and structure of the crust. These density heterogeneities and their boundaries, which correspond to gravity discontinuities, can be identified using the properties of the vertical and horizontal derivatives of the field. Gravity data on the Huelgoate granite in the Armorican Massif (France) were subjected to a new two-step geological analysis. First, the gravity discontinuities derived from Bouguer anomaly data were automatically identified; and second these discontinuities were combined with existing geological data. The first step was made possible by a software program that defines and visualizes the gravity discontinuities through an automatic analysis of the horizontal and vertical gravity gradient functions. The results is a map on which gravity discontinuities separate areas of different densities. The vertical and horizontal derivatives of the field were computed from the gravity data using frequency domain transformations. The second step corresponds to the geological interpretation of the gravity discontinuity map. A petrostructural map is used, which mainly shows the structure and the syntectonic internal fabric of the area, along with fracture data obtained on the granite and its host rocks. The latter illustrates late brittle deformation. The petrostructural map is used to adjust and rearrange the gravity discontinuities in order to obtain a gravity image deduced from the fabric of the rocks, a "fabric/gravity image". After this, a "fracture/gravity image" is derived by using the fracture data as a filter to obtain priority directions in the gravity discontinuities. Superimposing the two images provides a structural sketch map that gives a more integrated insight into the ductile and brittle deformation of the area than can be obtained from field observations alone.