Are all lineaments the surface expression of faults and fractures? A novel analysis using tunnel face mapping data from Norwegian road tunnels

Lineaments are elongated elements in spatial data such as valleys and ridges on topographic maps, or linear lows and highs in aeromagnetic data. Topographic linear depressions (topolineaments) are generally considered as the morphological expressions of easily erodable, elongated rock bodies situated within a mechanically stronger rock mass. In most circumstances topolineaments are even directly interpreted as faults and fractures, which forms the basis for lineament analysis study to understand brittle deformation patterns. However, topolineaments may also be formed by other tectonic and non-tectonic causes, such as alternating layers, foliation traces, dikes etc., or river- and glacial erosion not controlled by any bedrock features. This mix of potential causes begs the question: How robust is actually lineament analysis for characterizing and quantifying faults and fractures?. Testing the topolineament vs. fracture-relationship is not straight forward since topolineaments are usually occupied by rivers or creeks and covered with colluvium, which prevents direct observation of rock types and bedrock structures. Underground excavations allow for continuous logging of bedrock types, rock mass quality and fracture density and orientations, which is done routinely at tunnel face during tunnel construction. Here we make use of such underground data from a large dataset of Norwegian road tunnels to compare the position of topolineaments spatially and statistically with rock fracture density and orientations in the subsurface. The tunnel dataset comprises data from across Norway in areas with widely varying bedrock geology, tectonic evolution, and geomorphology (e.g. etched surface, alpine, lowlands), which allow for an evaluation of the robustness of lineament analysis in various settings. Topolineaments are acquired using a newly developed algorithm (OttoDetect) run on both 10x10m and 50x50m resolution digital elevation models. The algorithm ensures that tunnel data is compared to a homogeneous and reproducible lineament dataset without operator or hillshade illumination biases. Preliminary results from tunnels in areas with etched geomorphology show that c. 75% of all topolineaments correspond to weakness zones in the bedrock (i.e. very high fracture densities/very low rock mass quality compared to the surroundings). Hit rate increases for longer lineaments, which generally correspond to thicker fault zones. At the same time, only up to c. 60% of all weakness zones mapped at tunnel face can be spatially associated to a topolineaments, which demonstrate that significant brittle deformation is not expressed as topolineaments. Further analysis will be carried out to build a statistically robust dataset on the validity of lineament analysis.