4. RESULTS AND DISCUSSION 
4.1 The Abstract Area 
Visual analysis of the maps obtained (Fig. 1) confirms that 
the theoretical basis of the method is mostly correct. After 
dip-slip faults modelling (Fig. 1 d) clear lineament passing 
strictly along fault axes is recorded on Kv map (Fig. 1 f). 
Some vertical motion traces are revealed by Kh mapping 
too (Fig. 1 e) but there are no lineament along fault axes. 
After the strike-slip fault modelling (Fig. 1 g) only Kh 
mapping allows us to reveal a small lineament in the 
lower part of the map (Fig. 1 h). This lineament passes 
strictly along the dislocation axis. The Kv map does not 
contain lineaments (Fig. 1 i) although strike-slip fault 
traces are revealed which are a break of the valley and a 
displacement of its east part northward. 
After the oblique-slip fault modelling (Fig. 1 j) the clear 
expressed lineament, the break of the valley and the 
displacement of its east part northward are revealed on the 
Kv map (Fig. 1 1). Vertical and horizontal motion traces 
are recorded by Kh mapping (Fig. 1 k) but there is no 
theoretically presumed lineament along the fault axis. This 
fact contradicts the theoretical basis of the method and, in 
principle, can be connected with drawbacks of the oblique- 
slip fault modelling. Besides, we may probably suppose a 
lesser "sensitivity" of Kh to horizontal tectonic motions 
(Fig. 1 h, k) as compared with Kv "sensitivity" to vertical 
movements (Fig. 1 f, 1). However, real DEMs processing 
demonstrates that both Kh and Kv mapping allow us to 
reveal lineaments with equal "sensitivity" level (Florinsky, 
1992). Probably, this may be a result of erosion influence 
which can increase topographic expression of faults, for 
instance, in forms of fault-line valleys or scarps. 
After thrusts modelling (Fig. 1 m) clear lineaments 
passing along fault axes are revealed by Kv mapping (Fig. 
1 o). Some overthrust traces are also recorded on Kh 
maps (Fig. 1 n) but there are no lineaments along fault 
axes. 
After the gaping fault modelling (Fig. 1 p) the clear 
lineament is recorded on the Kh map (Fig. 1 q) as well as 
the lineament broken by the valley is revealed by Kv 
mapping (Fig. 1 r). 
4.2 The Part of the Crimean Peninsula and the 
Adjacent Sea Bottom ! 
Kh mapping allow us to reveal a) a complex of near-north 
oriented lineaments in the east and central parts of the 
study site, b) a complex of near-east oriented lineaments 
in the west part of the region, c) some near-north-east 
oriented linear structures in the north part of the study 
site, and d) some near-north-west oriented lineaments in 
the south part of the region (Fig. 2 b). Lineaments 
revealed by Kh mapping correspond to convergence areas 
(Kh negative values) and are connected with a valley 
network. A strong dependence of the Crimean valley 
network on the regional fault network was noted even by 
Muratov (1937). Lineaments recorded on the Kh map are 
mostly interpreted as strike-slip faults (Fig. 2 d). 
Kv mapping allow us to reveal a) a complex of near-west 
oriented lineaments in the east and central part of the 
study site, b) a complex of near-north oriented lineaments 
mainly in the west part of the region, and c) some near- 
north-west oriented linear structures mostly in the south 
and the north part of the study site (Fig. 2 c). Lineaments 
revealed by Kv mapping correspond to relative 
deceleration areas (Kv negative values) and are connected 
with terraces. These lineaments are mainly interpreted as 
dip-slip faults and thrusts (Fig. 2 d). 
Lineaments revealed on both Kh and Kv maps (Fig. 2 b, c) 
are interpreted as oblique-slip faults (Fig. 2 d). 
The map of morphologically classified faults (Fig. 2 d) 
displays a complicated spatial distribution of faults. Dip- 
slip, thrust and strike-slip faults unite, as a rule, into 
complexes. The complex of near-north-striking strike-slip 
faults stretches through areas of different geological 
origins. Dip-slip faults stretch, as a rule, across strike-slip 
faults. There are some complicated faults which include 
dip-slip, strike-slip, thrust and  oblique-slip offsets 
stretching along the same fault line one after another. 
Obviously, the fault map obtained (Fig. 2 d) has a forecast 
and somewhat subjective nature. First, it is a result of 
slightly ambiguous drawing of fault lines. Second, a visual 
analysis of these maps may result in the loss of some 
lineaments. Third, this fault map corresponds only to the 
single matrix step 3000 m. Using a smaller matrix step 
we can obtain a map which will include more faults, while 
using a larger step we will obtain a map with fewer faults. 
A visual comparative analysis of the fault map obtained 
(Fig. 2 d) and factual geological data made it apparent 
that a portion of revealed faults correlates with familiar 
ones. Another portion of revealed faults does not correlate 
with known structures. For the first time the complex of 
near-east-striking strike-slip faults is recognised in the 
west part of the region. Origin of these structures and 
their relationships with the regional tectonics is the 
subject of an individual study. It is very important that the 
most of the faults revealed and morphologically classified 
fit into the main regional fault groups. 
However, familiar faults do not all were revealed. On the 
whole, this is the result of the use of the single matrix 
step. To recognise all the topographically expressed faults 
(and to range faults into trans-regional, regional and local 
groups) we have to use a set of Kh and Kv maps 
correspond to several matrix steps or extents of DEM low- 
pass filtering and smoothing (Florinsky, 1992). 
5. CONCLUSIONS 
For revealing and morphological recognition of 
topographically expressed faults it is necessary a) to 
calculate Kh and Kv by a DEM processing, b) to stratify 
Kh and Kv values into two levels with respect to the zero 
value, and c) to map Kh and Kv. Lineaments revealed on 
Kh maps indicate faults formed mostly by horizontal 
tectonic motions. Lineaments recognised by Kv mapping 
correspond to faults formed mainly by vertical motions 
and thrusting. Lineaments recorded on both Kh and Kv 
maps indicate, as a rule, oblique-slip and gaping faults. 
The method was tested by processing the DEMs of the 
abstract area with modelled faults and the DEM of the 
part of the Crimean Peninsula and the adjacent sea 
bottom. For the abstract area the results obtained mostly 
correlate with the theoretical basis of the method. For the 
real area the comparative analysis of the results obtained 
and factual geological data demonstrates that the method 
256 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996