Details of a fault surface

An investigation of the 3D geometrical details of a fault trace from Mt. Alpi area (Lucania, Southern Italy) is here presented, based on some QuantumGIS plugins and an in-progress Python tool. The southern portion of a well-exposed fault, separating Mt. Alpi Unit carbonatic rocks from Frido Unit metasediments, was mapped in GIS and its possible orientations were defined using qgSurf, in combination with qProf and a still-in-development Python tool that rotates 3D fault traces. Gnuplot and R were used to visualise in 3D the fault trace points.

Analysed fault segment in Mt. Alpi area (Lucania, Southern Italy). The fault segment (red line) separates the Mt. Alpi Unit, made up of carbonatic rocks, from the metasediments of the Frido Unit. Basemap: GoogleEarth; QuantumGIS with OpenLayers plugin used for map visualization.

qgSurf, a QuantumGIS plugin, helps in determining the possible planar orientations of geological surfaces, f.i. fault traces, given a topographic DEM and satellite images and/or geological information. However, experimentations with this tool have evidenced that the calculated orientations can be sensitive to the particular examined segment, particularly when the surface is not planar. More reliable results can be obtained when fault traces are visualised and examined in 3D, in order to preliminary isolate possible planar segments making up a complex fault surfaces.

qProf, another QuantumGIS plugin, has been used to extract the elevation information of a profile from the used DEM, a 30 m Aster. Elevation info were read from qProf output and saved as an xyz text file using the in-development Python tool, so that they could be easely read by gnuplot. With gnuplot it is possible to produce interactive 2D and 3D graphics.


Gnuplot-generated 3D graphic of the xyz data of the fault trace in the Mt. Alpi sector, extracted from an Aster DEM using the qProf plugin.


Trough tentative rotations, 3D visualizations with gnuplot allowed to recognise a probable planar fault section almost 1-km long in the southern section, while the northern section is more complex (see figure below). This complexity could arise from the presence of a set of minor faults, orthogonal to the analysed fault, that are well mapped in the Mt. Alpi Unit, but with unknown prosecutions in the Frido Unit, due to the low relief and reduced exposures of this unit rocks.

.
Analysis of  Mt. Alpi fault trace, using gnuplot to visualize in 3D the topographic trace of the fault. Note that in gnuplot  graphics, the horizontal axes are not in 1:1 scale. Fault traces/points are in red in both GIS map (top) and gnoplot graphics (middle and bottom). In the middle graphics the view is top-to-down, while in the bottom graphics the view is rotated in order to evidence a southern, almost 1-km long planar segment.
Graphic created with gnuplot, gimp and Inkscape.


Using qgSurf on the previously defined fault sections, produced a value of 143.6/18.1 (dip direction/dip angle) for the sourthern planar section, and a mean value of 074.4/28.1 for the northern, more complex section.

Stereonet of the two derived fault sections. Created with Stereonet by Allmendinger.

The still in-development Python tool has the aim of interpolating 3D fault segments from a topographic trace and the local fault attitude a the two extremes of the analysed fault segment.
The tool rotates the fault traces (as xyz point values) in order to plot the trace points along a plane (x'-y') that is perpendicular to the intersection of the fault attitudes at the two extreme points (z' axis), and that contains the start point. In this way, assuming that the fault attitude changes are (almost) cilindrical, the 3D interpolation problem is reduced to a 2D one, that could be solved by user-chosen interpolation parameters.


Rotated basis frame for the fault trace, given its start and end points, coupled with inferred fault attitude at the extremes. The frame is centered on the trace start point. The z' axiz is parallel to the intersection of the two inferred local fault attitudes. The x' axiz is located in the plane containing the z' axiz and the trace end point. The y' axis is, obviously, perpendicular to both x' and z' axes. Fault traces are plotted on the x'-y' plane. 

The result for the fault trace is represented in the figure below. The southern section is well approximated by the 143.6/18.1 (dip direction/dip angle) solution (green line) while the solution for the northern section (red line) is clearly only an approximation that should be improved.


Fault trace points plotted on the x' (horizontal) - y' (vertical) plane. The green line indicates the southern section fault solution, the red one the northern fault solution. Graphic created with gnuplot, gimp and Inkscape.


It planned for the in-development tools to allow the user to interpolate the rotated fault traces in 2D by means of splines, in order to reproduce the local structures of the fault, and then expand the 2D solution to 3D assuming a "cilindrical" fault surface, and meshing it to allow the import in 3D GIS or visualization programs.





Comments