The effectivness of edge detection techniques depond strongly on derivatives calculation for rxemples methodes using dx dy dz calculated with matlab don t gives the same results when using oasis soft in my opinion the best filter is LTHG and THDRB
Dear Hichem Zerrouki, if you try to understand the structural interpretation of gravity and magnetic anomaly maps, I recommend a set of combination of attributes, geometrical attributes such as:
> First vertical derivative (DZ) (Nabighian, 1984);
> Tilt derivative (TDR) (Miller and Singh, 1994);
> Horizontal tilt angle (TDX) (Cooper and Cowan, 2006),
> Tilt angle of the horizontal gradient (TAHG) (Ferreira et al., 2013) and
> Curvature analysis (Roberts, 2001).
The last one (Roberts, 2001) is a classical research in 3D seismic interpretation, but you can apply in gravity and magnetic interpretation.
The length, width and depth extension in the maps is a clue in the analysis and interpretation of the individual faults, system-faults, and / or lineaments, and this geometric attribute have determined the geometry of your zone of interest.
(I have a lot of original idea in connection to it, but like an exiled researcher, I do not have the opportunity to test them… When somebody who has no opportunity for modern tool testing such a case he developing such methods which can be used same well… (something the same as in case of developing of martial art In East Asia and Japan) So if you have good geomorphological and geo-structural knowledge you can easily determine the buried faults… example if you understand well the attached article… In case of covered bodies same you will observe anomaly, but only in one case you can interpret it successfully if the subbasement ‘homogeneous’.
There is no best filter for all scenarios. Consider a hypothetical scenario where the regional dip varies with depth, each depth would have a different optimal filter. To adequately interpret gravity data one must understand the geology of the area being studied, at least to the point of the density of the rocks involved. One must visualize the density contrasts at various depths and select the filter best suited for that depth. For a complete interpretation several filters might be required.
If you have the gravity isostatic residual anomaly (free air corrected for topography/bathymetry and differences in crustal density between continental and oceanic crust), its total horizontal derivative offers much insight. The GI-THD maxima are located directly above the most rapid lateral changes in density distribution, typically at faults or carbonate bank edges. If you are working on a craton with relatively uniform crustal thickness and density, the Bouguer Anomaly THD works the same way. A general knowledge of basin geology will inform your interpretation by indicating where and how the most noticeable density variations will be located - changes in stratigraphy vs fault offsets vs intrusions of higher or lower density.
You can try to determine vertical gradients of gravity anomalies and use wavelet filters to smooth gradients for detecting tectonic structures. Some a priori knowledge is needed to assist your search.