From vertical gradient to total field in magnetic prospecting: study of the spectral operator application to the archaeological site of Saint Veran (Hautes-Alpes, France).

There are significant advantages in using avertical gradiometer in environmental or archaeological magnetic prospecting (Shlinger, 1990). One of the most interesting lies in the independence of the gradient with respect to the time-depending field variation. In practical terms, the time variations are not recorded and no base station neither second magnetometer is required. What is usually called « gradient » involves two sensors separated from 0.5 to 1 m: in reality, it is a difference. This is therefore an approximation of the true gradient, and we prefer name it « pseudo-gradient » (or PG). Assuming that the total field decays as r (r being the source to probe distance, and n being the « structural index» usually lying between 1 and 3), the PG decays approximately as r providing a better spatial resolution and magnifying the most shallow anomalies. These effects are often sought in subsurface surveys. However, gradiometry leads to the lost of the low spatial frequency (or large wave number) information. The trends and « regional » anomalies are strongly reduced when performing this kind of measurement. Hence, if both the pseudo gradient and total field measurements have their own advantages (depending on the source depth, extend with depth etc.), it can be, in lot of cases, interesting to recover the total field from the pseudo-gradient. On one hand, such a procedure would permit to retrieve the low wavenumber information (due to deep structures). On the other hand, modelling and interpretation of the total field are in common use.