Assessing the accuracy of combined DEM-based lineament mapping and the normalised SL-index as a tool for active fault mapping

Abstract

Satellite image-based lineament mapping in combination with the normalised Stream Length Gradient Index (SL/K-index) is one of the most widely used methods to identify active faults in areas where other verification methods, such as faulted strata, geophysical subsurface data, or seismic activity in the upper crust, are unavailable, and to classify their degree of activity. The accuracy of this approach, however, has not yet been assessed rigorously. We assess its accuracy by comparing the results against 27 thrust fault segments that were identified from six seismic reflection profiles with a total length of 340 km. Over 106,000 SL/K points were calculated from 815 river profiles on basis of the SRTM v.3.0 DEM. Standard deviations (SD) of < -2SD (class 1) to > +2SD (class 5) were calculated from the lognormalised SL/K data and five SL/K classes were made and assigned to the nearby lineaments. A comparison with the fault segments from the seismic profiles showed a good coincidence between the presence of knickpoints, SL/K classes 4 and 5 and to a lesser extent class 3. Lineaments with SL/K classes 1 and 2 never corresponded to actual faults. We identified seven master thrusts of which three are very active to highly active, likely older, structures with important amounts of accumulated slip. The other thrusts are probably younger faults with relatively little accumulated slip, yet they are also active to highly active faults. Minor strike-slip faults were identified as very active to highly active structures. Published slip rates shows that at least some of the SL/K class 4 faults experience slip rates of ~1 to ~4 mm a−1 and the SL/K class 5 strike-slip faults may experience slip rates as high as one metre over the course of several months. Most active faults occur where the basal evaporite layer is abnormally thick.