GNGTS 2016 - Atti del 35° Convegno Nazionale

590 GNGTS 2016 S essione 3.3 in order to estimate the unknown quantities for the scaling function (Eq. 3), that is the coordinates of the vertices of the body , }, where q= 1 ,...,Q . First of all, by looking at points along the ridge, where i =1, …, L , we compute numerically the scaling function there: . So we may form a system of equations along the ridge of the form: (5) The set of nonlinear Eqs. 5 has to be solved for the coordinates of the vertexes of a body. Application on gravity data. We performed a Very Fast Simulated Annealing (VFSA) algorithm (Sen and Stoffa, 1995) in order to solve equation 5. Wide bounds were used as constraints for the source depth. Here we show the results obtained for a synthetic case: a complex shape body, which can approximate a salt dome (dashed line in Fig. 2). The gravity fields and the scaling function of the model were calculated by using the Talwani’s formula. A -2.0 g/cm 3 density contrast is assumed for calculating the gravity field. Scaling function was computed Fig. 1 – Fitting of observed and calculated scaling function. The various scaling function are as: a) of gravity data, b) – d) of vertical field derivative, e) – g) of second order vertical derivative of the field and h) – i) of horizontal derivative of vertical gradient of the field.