FAULT MORPHOLOGY RECOGNITION BY DIGITAL ELEVATION MODEL PROCESSING
Igor V. Florinsky
Institute of Mathematical Problems of Biology, Pushchino, Moscow Region, 142292, Russia
Commission IY, Working Group 4; Commission YII, Working Group 4
KEY WORDS: Geomatic, Geology, DEM, Analysis, Fault,
ABSTRACT:
Morphology, Recognition, Topography
À quantitative method of morphology recognition of topographically expressed faults is developed. The method is based
on digital elevation models (DEMs) analysis. Lineaments
revealed on horizontal landsurface curvature maps indicate
faults formed mostly by horizontal tectonic motions (i.e., strike-slip faults). Lineaments recognised by vertical landsurface
curvature mapping correspond to faults formed mainly by vertical motions (i.e., dip-slip and reverse faults) and thrusting.
Lineaments recorded on both horizontal and vertical curvatures maps indicate, as a rule, oblique-slip and gaping faults.
The method is tested by processing the DEMs of an abstract area with modelled faults and a DEM of a part of the
Crimean Peninsula and the adjacent sea bottom.
1. INTRODUCTION
Faults can be revealed by several geological, geophysical,
remote sensing and topographical techniques (Slemmons,
Depolo, 1986). As tectonic motions can result in linear
deformations of the landsurface so topographically
expressed lineaments are often used as fault indicators
(Hobbs, 1904; Ollier, 1981). Properties of linear relief
dislocations formed by vertical tectonic motions differ from
properties of topographic lineaments which are horizontal
movement traces (Trifonov, 1983). Qualitative and
quantitative signs of these differences can be used as a
basis for fault morphology recognition.
Qualitative approaches to revealing and morphological
classification of faults by a relief analysis were often
exploited (Trifonov, 1983; Slemmons, Depolo, 1986). Of
frequent use was a visual analysis of topographic map
contours (Hobbs, 1904; Phylosophov, 1960) and remotely
sensed images (Wilson, 1941; Trifonov et al, 1983).
Keller (1986) summed up data on qualitative geomorphic
indices of active faults. Stereophotogrammetric analytical
techniques were applied for fault revealing, morphological
recognition, dip and strike measuring (Vinogradova,
Yeremin, 1971).
Indicated qualitative approaches are not free of
subjectivity. However, due to difficulties in formalization
of fault geomorphic indices there are a lot of quantitative
computer methods of fault revealing by remotely sensed
(Burdick, Speirer, 1980; Masuoka et al, 1988) and
topographic data processing, but there is no quantitative
method for fault morphological classification by outlined
data handling without ancillary geological information.
Digital elevation models (DEMs) and DEM analysis
methods are used for fault recognition as about 90% of
fault geomorphic indices can be defined quantitatively
(Schowengerdt, Glass, 1983). There are techniques of
perspective views (Campagna et al, 1991), thalvegs
revealing (Eliason, Eliason, 1987), landsurface gradient
and aspect mapping (Onorati et al, 1992), reflectance
mapping (Wise, 1969; Schowengerdt, Glass, 1983).
DEMs are applied for measuring dip and strike of known
faults (Chorowicz et al, 1991). However, the use of
indicated methods of DEMs analysis without ancillary
geological data does not permit us to determine a fault
morphology.
Reproducible recognition of lineaments can also be
obtained by calculation and mapping of the horizontal
(Kh).and vertical (Kv) landsurface curvatures with the use
of DEMs (Florinsky, 1992). Statistical properties of
lineaments (i.e., orientation, length, density) recorded on
Kh maps are rather different from statistical properties of
linear structures revealed by Kv mapping (Florinsky,
1992). Taking into account physical and mathematical
senses of Kh and Kv (Evans, 1980; Shary, 1991) we can
reason that lineaments revealed by Kh mapping
correspond mostly to structures like strike-slip faults,
while lineaments revealed by Kv mapping indicate mainly
structures as dip-slip faults and thrusts.
2. THEORETICAL BASIS OF THE METHOD
Let us consider a surface klmp. Kv is the curvature of a
normal section bac of the surface klmp. The section bac
includes a gravity acceleration vector g and a normal
vector n in a given point a. Kh is the curvature of a
normal section dae of the surface klmp. The section dae is
perpendicular to the section bac and includes the normal
vector n in the given point a. If indicated sections are
convex Kh and Kv have positive values, if sections are
concave Kh and Kv have negative ones, if sections are
plane Kh and Kv have zero values (Evans, 1980; Shary,
1991).
Suppose a dip-slip (or reverse) fault is formed within the
surface klmp. Kh and Kv values within the scarp zone will
change and besides Kv will have negative values all along
a fault line. Let us stratificate Kh and Kv values into two
levels with respect to the zero value and paint areas with
Kh and Kv positive values in white colour, while areas
with negative values of curvatures in black colour. An
indicator of the dip-slip (or reverse) fault that is a black
lineament on a white background will be recorded on the
Kv map. A similar lineament will be recognised on the Kv
map if before a vertical motion a surface was plane. If
before vertical motion Kv values were negative the Kv
sign will also change along the fault line therefore a white
lineament on a black background on the Kv map will
indicate a dip-slip fault. A lineament consisting of black
252
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996