GISdevelopment.net ---> AARS ---> ACRS 1989 ---> Poster Session

Digital, image processing of Landsat MSS data for identified, geomorphic, litho logical and structural, features in P.I.Shiresh Satpuli area of Garhwal Himalaya, India

Sayed Ahmad Ali, Liagat A.K. Rao
Department of Geology, Aligarh Muslim University,
Aligarh 202002


Abstract
Landsat MSS data is digitally enhanced through VAX 11/180 computer system using VIPS-32, software with a view to select appropriate band combination and spatial filtering functions for delineating certain geomorphic, litho logical and structural features. Among the various enhancement technique of digital image processing it has been observed that the band in his 4/5, 5/1 and 4/1 are useful for delineating broad litho logical, structural elements, flood pains, pediment zone with alluvial fans and perennial streams. FCC of brightness index shows that it is good for drainage studies. The principal component analysis, PC1 and PC3 are good for interpretation of regional structure whereas PC2 proved good for geomorphic studies. It has also been observed that the directional filtering using SOBEL operators is good for delineating lineaments and good fractures.

Introduction
The present study area which forms a part of sub-Himalaya and Lesser Himalaya is bounded by latitude 29°50' to 30°15' and longitude 79915' to 78°40' E, covering an are of about 1322 sq. km., was selected as tost site for digital image processing. The analysis of the enhanced images after applying various functions were carried out by visual interpretation method using tonal and textural contrast. An attempt has been made to delineate geomorphologic, litho logical and structural features from digitally enhanced images and correlate them with field data.

Geological set up
Geologically, this study area may be divided into three broad zones each of which is separated by a major thrust. The southern part of the area is occupied by siwalik belt, whose fringing zone is occupied recent alluvium. The central portion of the area is occupied by Karol Belt which is separated from the Siwaliks by the Main Boundary thrust, whereas the chail Group belonging to Kumaun Super group is exposed on the north of the area (Fuche and Sinha 978). The two thrusts that separated the three geological untis are Main Boundary Thrust of the Himalaya on the south name and Garhwal thrust on the north. These thrust zones trend more or less in HW-SE direction almost parallel to that of the Himalaya Trend. The two major rivers of this area, viz. the Ganga and Hayar cut across there thrusts by at some places their sources appear to have been affects by these thrusts. A generalized stratigrapchic sequence of different litho units and their litho-characters encountered in teh area are given (after, Kumar and Dhaundiyal, 1979, Gairola 1980) in Table 1 and shown in Fig. 2.

Formations Age Characteristic Litho-types
Lower Siwalik Mid Niocene to
Lr. Pleistocene
Sandstone with frequent clay bands.
Subathu Eocene Sandstone sahle and argillaceous
Tal Hid Triassic
to Upper
Cretaceous
Sandstone, quartsite, pebbly grits, Shale and limestone.
Krol Mid Permian
to
Lr. Triassic
Siliceous/Dolomitic Limestone, shale
Blaini Carboniferous
to
Lr. Permian
Boulder beds of shale, slate and limestone.
Nagthat Devonian Quartzite, siltstone, slate
Chandpur Ordovician
to cilurian
Slat, phyllite with quartzite and quarts schist.
Damta Precambrian Quartizite, slates, siltstone and greywacke
------------------- ------------ Garhwal Thrust ---------------------------------
Low grade Metamorphics Phyllite. Schist, quartzite and quartzite schist.

Methodology
Digital image processing of Landsat BSS data stored in the form of computer compatible tape (Data of pass: 27.11.83; Path-row: (D146-039) of the parts of study area (Pixel No: 18co-2312; Scan line no 1150-1662) has been carried out on Super mini computer VAX 11/780 using VIPS-32 software. Pericolor-2001 display of the images. The selected enhanced images used for display of the images. The selected enhanced images was received on film negatives recorded by using the Dunn Camera connected directly processed in the photo-laboratory to produce a hard copy for visual interpretation. Interpretation was competed with help of the survey of India toposheets, existing geological and geopmorphological map and field data. The following image functions of VIPS-32, software package have been used for generating various types of enhanced images.
  1. Band combination functions:

    1. CRAT- Takes the quotient of two linear bands (Band ratio).

    2. CBRIL- Computes the brightness index on two bands (Brightness index)

    3. KLCEVE- Carries out the factorial analysis of the bands in a subscene (Principal, component analysis).

  2. Spatial filtering functions:

    1. AWECONT - Edge enhancement of a band in a sub scene.

    2. EXTCONT- Edge extraction from a band in a sub scene

Digital image proceeding an results
CRAT: The Landsat MSS data were rationed to criminate the extraneous effects like clouds and shadows because they changes the reflectance value. The following ration combinations were carried out to produce a composite color ration image.
  1. Ratio 4/5, 5/7 and 4/7 : The ratio product obtained (P1. I Fig. 1) shows a clear cut demarcation between the Siwalik Group of rocks with Pre-Territory rocks. The structural elements like Main Boundary Fault, Garhwal thrust sheet and various parallel lineaments could be identified clearly. The flood plain, piedmont zone with alluvial fans, perennial stream are well pronounced.

  2. Ratio 4/7, 5/7 and 4/5: The ratio product obtained shows almost same results as of ratio 4/5, 5/7 and 4/7 but the flood plain and piedmont zones are more brighter in this product.
CBRIL : Brightness index on bands 4-5, 5-6 and 6-7 were combined to produce a FCC (Pl. I, Fig. 2). FCC of brightness index shows that it is good for drainage studies. Broad lithological discrimination could be easily delineated. It also shows that nW-SE and NE-SW trending lineaments are more prominent.

KLOEVE: Principal component analysis was carried out by applying various principal component transformations. The products of combination of PC1, PC2 and PC3. were analysed visually. Ti was found that n addition to the broad lithological discrimination; geomorphic variation, drainage patterns and parallel lineaments have come up picely. These lineaments are not only closely related with the thrusts and lineaments are faults interpretation aerial photos and imagery but a few more new lineaments were also picked up indicating the presence of other thrust slices. PC1 and PC3 are very good for analysis of regional structure whereas PC2 is proved good for geomorphic studies especially the study of drainage pattern.

Combinations of Band functions
Functions like CRAT-KLOEVE were also carried out using ratio 5/7, PC1 and PC2 and was found suitable for demarketing major and minor lineaments. Broad litho logical discrimination between siwalik group and pre-territories, geomorphic units like flood plain, piedmont zone and geomorphic features like alluvial fan were also picked up in this FCC.

Directional Filtering
Laplacian was applied in band 7 but it did not exploits good results and was found unsuitable for geological studies. SOBEL operator was applied in band.7. It was found very useful for lineament study because it has a capacity to enhance the image features, in a particular direction on the basis of matrix/ matrix number 7,8 and 9 were applied for present study. It was found that SOBEL-7 enhanced the NE-SW trending the lineaments, SOBEL-9 enhances the lineaments trending in NW-SE direction.

Edge Enhancement
Enhancement product exhibits drainage liens with some geomorphic and structural elements. Edge extraction in particulars bands like 7 and 5 were also applied which gives only a better contrast in the image.

Conclusion
  1. Digital analysis of Landsat-MSS data using VAX-11/180 super mini computer exhibit good results. Fine litho logical discrimination could not be possible due to the presence of vegetation cover. However, vegetation plays an important role in broad litho logical discrimination and in enhancement of various subtle lineaments.

  2. Combination of ratio and principal component gave better delineation between geomorphic features like piedmont zone and hilly terrain with some lithological and structural discrimination.

  3. Directional filtering of SOBEL gave very good enhancement of lineaments in NW-SE, NE-SW and E-W direction.

  4. Enhancement product of brightness index and different ration combinations was found suitable to differentiate various structural elements.

  5. Production of PC1, PC2 adnPC3 gave additional information of day and filled channels. Detection of dry and moist conditions in piedmont zone was possibly by tonal contrast.

  6. Various enhanced data products was found suitable for drainage studies.
Acknowledgement
The authors express there thanks to Prof. S.N. Casshyap, Chairman, Department of Geology, Aligrah Muslim University, Aligarh, for providing facilities. Thanks are due to Prof. S.H. University, Aligarh for going through the manuscript.

Reference
  1. Fuchs, G. and Sinha,A.K.1978 Geology and Tectonics of Garhwal and Kumaun, Lesser Himalaya, Jahrb, der. Geol. Bund 121, Ban,Vienna 21-241.

  2. Girola and Saxena, R.K. 1980: Structural analysis of the area around satpuli, Pauri Garhwal, U.P. Him, Goel 10, 156-177

  3. Kumar, G. and Dhaundiyal, J.N. 1980: On the Stratigraphic position of the Tal formation, Garhwal symposium, Garhwal and Tehri garhwal Dsitricts,u.P. Jour, Pale, Soci. Ind., 23&24: 58-66 (for 1978-79).