int. j. remote sensing, 2002, vol. 23, no.

13, 2527 –2536

Technical note

Remote sensing and GIS technologies for improvements in geological

structures interpretation and mapping

A. K. SARAF*, P. MISHRA, S. MITRA, B. SARMA and

D. K. MUKHOPADHYAY
Department of Earth Sciences, University of Roorkee, Roorkee 247 667, India

(Received 4 December 2000)

Abstract. In a highly rugged terrain like the Himalayas, due to accessibility

problems, conventional Želd-based geological mapping is a Herculean task. Rock
exposures and geological structures are often studied only at a few places and
later interpolations between observations are made to acquire continuity of geolo-
gical formations and structures. With the advent of satellite images having synoptic
view, higher spatial and better spectral resolution, it has become possible to
recognize deformation structures having tectonic signiŽcance in a better and more
reliable way. Hitherto, the minor structural changes could not be realized on
ground mapping, which has always a limited coverage. This limitation acted as
a hindrance to recognize tectonic signiŽcance of the structural features. An integ-
rated approach of remote sensing and Geographic Information System (GIS)
technologies provides better opportunities to prepare geological maps with higher
accuracy and in a cost-eVective manner. Using relatively high resolution multispec-
tral digital data, interpretation of Želd observations can be done very eYciently
without any vague interpolation between observations. Further, an integrated
approach of remote sensing and GIS can be employed to make improvements in
existing geological maps. Such an integrated approach becomes complementary
to conventional Želd-based geological mapping techniques. In the present study,
an integrated remote sensing and GIS-based approach has been developed and
successfully demonstrated to make improvements in existing geological maps and
result in better identiŽcation and interpretation of geological structures. The study
has led to the modiŽcation of existing maps of Nahan Salient, Himachal Pradesh,
India, after proper Želd veriŽcations.

1. Introduction
The Himalayas represents a classic example of collision-type orogenic belt. It
was formed as a result of the northward drift of the Indian plate after its split from
Gondwanaland, the consumption of intervening oceanic crust of the Neotethys and
the collision with the Tibetan plate during the Eocene period, around 50 Ma ago.
The collision was followed by a continued northward convergence of India against
Tibet resulting in crustal shortening both on the northern margin of India as well

*e-mail: [email protected]

Internationa l Journal of Remote Sensing

ISSN 0143-116 1 print/ISSN 1366-590 1 online © 2002 Taylor & Francis Ltd
http://www.tandf.co.uk/journals
DOI: 10.1080/01431160110070311
2528 A. K. Saraf et al.

as on the southern end of Tibet (Kumar et al. 1991). The compression that was
responsible for the India–Asia collision and consequent formation of the Himalayan
orogen, though subdued, has not yet ceased. A large number of earthquakes occurring
between the Main Central Thrust (MCT) and Himalayan Frontal Fault (HFF) bear
testimony to the continuing compressional stress regime (Routela and Sati 1996).
IdentiŽcation of active and existing faults in the Himalayas is extremely signiŽcant
as these reect the Himalayan continental collision. Active faults have moved repeat-
edly during the Quaternary, resulting in the dislocation of many landforms, such as
streams, alluvial and piedmont fans and river terraces, etc. The Siwalik sedimentary
of the Upper Tertiary age is being squeezed between the very active Main Boundary
Thrust (MBT) and the HFF. Here, eYcacy of remote sensing data can be utilized
eVectively and it proves an immensely powerful tool for geologists in various geo-
logical investigations (Saraf 1998). Realizing its several advantage s over conventional
geological mapping methods, the present study demonstrates successful application
of remote sensing data from IRS LISS-III sensor for up-to-date and accurate geo-
logical mapping in a highly rugged, tectonically active and almost fully vegetated
(covered ) terrain.
For many years remote sensing and Geographic Information System (GIS) tech-
niques have been recognized to delineate structural features like faults, folds, lineaments,
etc. In satellite images, steeply dipping faults show prominent linear features and are
unaVected by topography . However, there are many other ways in which obvious linear
features may be formed. They may be the trace of axial plane of a fold, joints, dykes,
steep to vertical strata, and cultural features, such as roads and boundaries between
areas of diVerent landuse. It may also be possible to link various elements of scenes by
lines, such as isolated boundaries between surfaces of diVerent tone, straight segments
of river, notches in ridge, etc. These may be purely fortuitous connections , but could
have some genetic relation to subtle or deeply buried fault (Drury 1987).
Due to gentle dip, expressions of thrusts in satellite images are strongly controlled
by topography , and their expression is often identical with bedding surfaces.
Occasionally, the oor of a thrust, which may be coated by tough mylonite, stands
out at as a prominent dip slope, and can be seen as lineament in satellite images.
Sometimes the tones, patterns and texture of rocks forming the roof of a thrust
clearly indicate that they are diVerent from the rocks forming the oor.
In a hilly terrain, like the Himalayas, it is not possible to map or collect data
from Želd throughout any structural feature. The data are collected from accessible
places and geologists then interpolate those data to obtain continuity. In this context,
image interpretation of structural features is more reliable than their interpolation
from Želd data. Changes in vegetation and surface texture related to fault are diYcult
to see at close range. The synoptic view of images enables widely separated pieces
of evidence to be linked as sharp and semi-continuous linear features or lineament.
That satellite sensor data, e.g. IRS LISS-II, improves geological mapping in the
Himalayas is a long-recognized fact (Saraf and Das 1997, Saraf 1998). In the present
study a portion of the Sub-Himalayas and Lesser Himalayas in the Nahan Salient
has been taken into account to prepare a new map by improving existing maps of
the area using satellite images (IRS LISS-III) and Želd checks, and also to demarcate
some areas for future Želd study.

2. Study area and geological setup

The study area is situated between latitudes 30° 30ê to 31° 10ê N and longitudes
76° 45ê to 77° 40ê E, which is a part of Himachal Pradesh and Haryana states (Žgure 1).
 T echnical note 2529

Figure 1. Location map of the study area showing thrusts interpreted from IRS-1C
LISS-III standard FCC and thrusts mapped from Želd data.

The Himalayan mountain range is about 250–350 km wide and extends for about
3000 km from Afghanistan to Burma. It is undergoing rapid uplift at rates between
0.5 and 4 mm per annum. It is consequently experiencing rapid erosion, with depos-
ition in the foreland. The extensive apron of Quaternary sequences constituting the
Sub-Himalayas was deposited in such a foreland. These sediments include the foot-
hills rising to north of the Indo-Gangetic plain and are deformed by contemporary
orogenic processes.
The Lesser Himalayas occupy elevations from 1500 to 3000 m and are thrust
over the Sub-Himalayas . They consist of Precambrian to Cambrian sedimentary and
very low-grade metasedimentary rocks.
Tectonically, the Himalayan orogen can be divided into a frontal fold-thrust belt
(FTB), bounded to the rear by an allochthonous terrain ( hinterland) and to the
front by an autochthonou s continental interior (foreland or foredeep). The frontal
FTB can be divided into the Sub-Himalayan Zone (SHZ) and the Lesser Himalayan
Zone (LHZ). The hinterland to the north is an internally complex belt of allochthon-
ous mostly Higher Himalayan crystalline rocks (HHCZ) and sedimentary rocks
(HHSZ). The foreland to the south consists of Indo-Gangetic foreland basin (IGFB).
The frontal FTB is separated from the IGFB by the Main Frontal Thrust (MFT).
Within the frontal FTB, the MBT (Valdiya 1980) separates the SHZ in the south
from the LHZ in the north. The MCT separates the LHZ in the south from the
HHCZ in the north.

3. Data used in the present study

In the present study the following types of datasets from diVerent sources were
used:
1. Remote sensing digital data of IRS-1C LISS-III (Path/Row 95/49) of
12 October 1998.
2530 A. K. Saraf et al.

2. Existing geological maps: (a) Pilgrim and West, 1928 (Scale: 1 inch=2 miles),
(b) Rupke, 1974 (Scale: 1 inch=4 miles).
3. Field data (structural and GPS).
4. Survey of India (SOI ), topographica l maps (scale: 1:50 000).

4. Methodology
The methodology adopted for the present study consists of Žve steps:
1. In the Žrst step all the existing geological maps (as mentioned above) (including
thrusts, axial traces of folds, lithological contacts and drainage lines) were converted
to digital format by means of digitization (Žgure 2 ). Remote sensing data were
acquired from the National Remote Sensing Agency (NRSA), Hyderabad in digital
format.
2. Georeferencing of the remote sensing data was done to bring the image into
the same coordinate system as the base map (SOI 1:50 000 topographica l maps).
Since all geological maps had diVerent scales and coordinate systems, these were
also brought to same coordinate system as base map through georeferencing
technique.
3. The image was enhanced after applying various linear and non-linear contrast
and edge-enhancing techniques, standard false colour composites (FCC: RGB: 321)
to improve the quality for better interpretation and used to prepare a lineament
map by visual interpretation.
4. A geological/structural map prepared by Mukhopadhya y and Mishra (unpub-
lished) was transferred on the enhanced image. An attempt was made to correlate
structural features with the above map.
5. Structural features were modiŽed and/or extended as required using the linear

Figure 2. Flowchart depicting the various steps adopted to obtain the Žnal modiŽed
geological map.
 T echnical note 2531

elements as a guide, and a Žnal complete geological/structural map was then

prepared, compiling all the data.

4.1. Hardware and software used

In the present study a PC-based digital image processing and GIS software (Arc
View 3.2 with Spatial Analyst, Image Analyst extensions) was used to achieve the
objectives. ILWIS 1.4 has been used extensively to digitize all the geological maps
using an A0-size Summagraphics digitizer.

5. Image interpretation and discussion

In the present study, the interpretation of various structural features involves the
following steps.

5.1. T racing pre-existing structural features

As discussed earlier, the study area is traversed by a series of thrust faults
generally trending NW to SE and a few transverse faults trending north to south,
all of which are not traceable from the image. This is because lineament tracing from
the image (Žgure 3) is done with the help of tonal contrast along an isolated line,
straight segment of river, line of vegetation and other linear features. As thrusts are
gently dipping planes, their expressions in the images are also aVected by topography.
So here colour contrasts along a linear or curvilinear feature play the major role.
Change in colour, however, can take place due to various reasons. This is where
Želd data and existing maps become indispensable.
Considering existing maps and Želd data, thrusts and axial traces of folds are
traced from the satellite image. The thrusts traced from the image show a very good
match with the existing maps and Želd data at many places, except in the case of
Paonta Thrust. For the Paonta Thrust, the image trace and the existing map do not
match. Field Global Positioning System (GPS) locations, however, match well with

Figure 3. Lineaments interpreted from IRS-1C LISS-III FCC with rose diagram showing
prominent lineaments directions.
2532 A. K. Saraf et al.

the image trace. Based on the above observations, modiŽcation has been done for
the location of Paonta Thrust (Žgure 4).
Traces of axial planes of folds from the image also show very good match with
the existing maps and at many places traces of the axial planes are extended with the
help of the image (Žgure 4).

5.2. Inferred thrusts and traces of fold axial planes f rom the image
5.2.1. Inferred thrusts
In the image there is a distinct continuous linear feature present north of the
MBT which is marked by tonal contrasts (Žgure 4). But in the maps of Pilgrim and
West (1928) or Rupke (1974) no such feature is marked. In the map of Pilgrim and
West (1928) a discontinuous thrust is marked in the eastern, central and western
parts of the Krol belt discontinuously (Žgure 5).
Field mapping (Mukhopadhyay and Mishra, unpublished ) in this area and
attempts to extend the thrust marked discontinuousl y by Pilgrim and West (1928)
proved unsuccessful westward due to inaccessibility. The thrust was mappable in the
central sector around the village Parara. These entire locations match with the linear
feature marked on the image, and a new thrust named as the Parara Thrust is
inferred along the line of tonal contrast (Žgure 5). The thrust seems to be a connecting
splay in between the MBT and Giri Thrust.
The same is the case north-east of the Jaunsar Thrust. Here also a curvilinear
feature can be traced discontinuously from the image that is not present in the
existing maps (Žgure 6). Field mapping in and around the village Sangrah had been
carried out in this area also (Mukhopadhyay and Mishra, unpublished), and a
forethrust and back thrust are marked at the same locations, which match with the

Figure 4. Inferred thrust and axial traces of folds north of MBT detected from IRS-1C
LISS-III FCC (zoomed part of image shown in Žgure 1).
 T echnical note 2533

Figure 5. VeriŽcation of the inferred thrust north of MBT, traced from the image, from Želd
checking and existing map (Pilgrim and West 1928).

Figure 6. Inferred thrust and axial trace of fold marked from IRS-1C LISS-III FCC north
of Jaunsar thrust.
2534 A. K. Saraf et al.

image traces. By interpolating Želd data and the image trace another new thrust
(named as the Sangrah Thrust) is inferred in this area (Žgure 7).
5.2.2. Inferred traces of fold axial planes
North of MBT, where the Parara Thrust has been inferred, two parallel linear
features were traced from the image, which are oVset by the thrust (Žgure 4). Field
data conŽrms that these two are traces of axial planes of folds (Žgure 8), the northern
being a synformal trace and the southern one an antiformal trace. The former Žtted
well with the synformal fold closure shown by both Pilgrim and West (1928) and
Rupke (1974) in the south-east part of the area within the Krol belt (Žgure 8).
Another linear feature, slightly oVset by the Sangrah Thrust north-east of Jaunsar
Thrust, has been traced from the image (Žgure 8). Field mapping (Mukhopadhya y
and Mishra, unpublished) conŽrms that this represents the axial trace of a regional
fold known as the Nigali Dhar syncline (Žgure 8), having the Tal Formation at
its core.

5.3. ModiŽed geological map

After the modiŽcation and addition done in the fold–thrust belt for both, the
thrusts and fold axial planes, as discussed in §5.1 and 5.2, a new modiŽed map of
this area was prepared (Žgure 9). The map was prepared compiling all the data,
namely the maps of Pilgrim and West (1928) and Rupke (1974), Želd data
(Mukhopadhya y and Mishra, unpublished) and satellite images.

6. Conclusions
From the present study it can be concluded that satellite images, if employed
judiciously, can be a very eVective tool for geological/structural mapping. Especially

Figure 7. The thrust inferred north of Jaunsar thrust, mapped by compiling Želd data and
trace of thrust from image.
 T echnical note 2535

Figure 8. VeriŽcation of inferred axial traces of folds marked from image with Želd data.

Figure 9. ModiŽed geological map prepared by compiling data from IRS-1C LISS-III FCC,
Želd data and existing map (Pilgrim and West 1928) .
2536 T echnical note

in hilly terrain like the Himalayas, where poor accessibility and dense vegetation
cover hinder Želdwork, satellite images can play a great role in interpolating discon-
tinuous Želd data. Undoubtedly linear features of geological importance do occur,
which are only visible at small-scale images, as they form broad, subtle and very
long connection. The geological signiŽcance of these features can be attributed only
after Želd checking. Detection of such features from images, however, makes the
Želd/ground check task in geological mapping minimal. With the aid of remote
sensing data, a Želd-mapping programme can be planned in advance based on prior
knowledge of outcrops or expected locations of interesting structural features.

Acknowledgments
The authors would like to acknowledge the kind permission of the Director of
the Alternate Hydro Energy Centre, University of Roorkee to use remote sensing
data in the present study.

References
Drury, S. A., 1987, Image Interpretation in Geology (London: Allen & Unwin).
Kumar, R., Ghosh, S. K., Virdi, N. S., and Phadtare, N. R., 1991, Excursion Guide: T he
Siwalik Foreland Basin (Dehra Dun-Nahan Sector), Special Publication No. 1, Dehra
Dun: Wadia Institute of Himalayan Geology.
Mitra, S., 2000, Satellite data interpreation of geological structures for improvements of
existing maps, MSc Project Report, Department of Earth Sciences, University of
Roorkee (unpublished).
Pilgrim, G. E., and West, W. D., 1928, The structure and correlation on the Simla rocks.
Memoirs of the Geological Survey of India, 53, 1–140.
Routela, P., and Sati, D., 1996, Recent crustal adjustment in Dehra Dun valley, western
Uttar Predesh, India. Current Science, 71, 776–780.
Rupke, J., 1974, Stratigraphic and structural evolution of the Kumaon Lesser Himalaya.
Sedimentary Geology, 11, 81–265.
Saraf, A. K., 1998, IRS-LISS-II improves geological mapping n the Himalayas. International
Journal of Remote Sensing, 19, 2239–2243.
Saraf, A. K., and Das J. D., 1997, Neogene deformation of Siwaliks aVected by the Delhi–
Hardwar ridge as seen in satellite data, India. Current Science, 73, 877–880.
Valdiya, K. S., 1980, The two intracrustal boundary thrusts of the Himalayas. T ectonophysics,
66, 323–348.