GISdevelopment.net ---> AARS ---> ACRS 1989 ---> Geology

Tectonic fabric mapping of the Sungei Jelai area, Pahang, Malaysia by Remote Sensing and digital processing techniques

Lai Kok Hoohu
Geological Survey of Malayisa


Abstract
A tectonic fabric map of the Sungai Jelai area, a heavily vegetated area was produced by interpretation of digitally enhanced landsat MSS images. The image: were found to be affected by aspect ratio distortion which was rectified by registering the at infected images to digitized map data. A large amount of structural geology information., including previously unmapped lineaments and circular features were revealed by complementary sets of directional edge and contrast enhancements of Landsat 4 MSS band 4 (infrared) images. The structural information derived from the images was integrated with overlays of digitised geological map data which included mineral occurrences and geochemical anomalies to produce an 'image map' and a hard copy obtained by computer-aided mapping. The integrated data revealed that some in the study area. In this respect, N-S NNW-SSI, and NNE-SSW trending linears and some circular features were the most significant.

Introduction
Major linear features have long been used as guides to the regional search for one deposits. Some lineaments represent zones of weakness through which mineralizing fluids may he directed and as such represent lines which mineral deposits are likely to occur. However, the mapping of lineaments (tectonic fabric) by application of remote sensing in an equatorial contry like Malaysia has many problems. Notwithstanding cloud and naze problems, vegetation cover imposes serious constraints on interpretation of remotely sensed imagery. Traditional image interpretation is limited by the eye's sensitivity to subtle changes of hue, texture and tone of vegetation cover. Mortunately, image enhancement by digital processing can be used to emphasize such subtle features, making the interpretator's task easier. Details which were previously imperceptible may be revealed. Purthermore, the information so derived can be integrated with other geological data for currelation or spatial analysis.

The Study Area
The Sungei Jelai area covers about 3000 sq. km in the northwestern part of the state of pahang in central Peninsular Malaysia. This area is mostly vegetated with primary and secondary forests, with parts alienated for agriculture. The main cultivation is oil palm planation and the rest are small holdings of rubber trees and fruit orchards.

The western part of the area is underlain by a narrow belt of Lower Palaeozoic rocks which form a narrow mountain range. The rocks consist mainly of sandstone and conglomerate. Immediately east of these rocks is a broad belt of low undulating to hilly terrain which is underlain by Carboniferous to Permian rocks consisting of three distinct facies; an argillaceous, a calcareous and volcanic facies. These Permo-Carboniferous rocks are intruded by several small igneous bodies of late Triassic age. They are mainly granitites, except for one body, which is composed of syenite and related rocks.

Within the study area there is significant mineralization of gold and base metals. The muneralization is reported to be structurally-controlled )Lee, et al, 1986) and deposits occur mainly as fissure-fillings.

Data sources and digital processing
The data used in this study consist of (1) three 1:63, 360 scale geological maps (richardson, 1950 and Procter, 1972) in a N-S sequence with Lopographic and drainage Features, (2) and digital LANDSAT MSS CCT, picture center 127/57, acquired on 28 June 1985, (3) a regional recognissance geochemical anomalies map (lee, et al, 1982) and (3) a file on mineral occurrences from the Geological Survey of Malaysia's Archive.

Image rectification
The area under study covered to full resolution 512x512 pixels subscene but the images were severely distorted by aspect ratio distortion, a common occurrence in LANDSAT MSS (Richards, 1986). Rectification of the images were performed by image to image registration. Briefly, the operation consisted of digitizing the three geological maps, which are 15x15 mins. square on the Malayan Rectified Skew Orthomorphic projection as a single base amp including a set of control points which were positively identified of both the maps and images.

A raw image 2048x1536 pixels covering an area in excess of that covered by the map was then warped or titled into the map to produce three full resolution sub scenes registered to the base map at a scale of 54x54 m per m per pixel. This procedure,, described in detail by Lai ( 1988), was a pixel to pixel operation. Degradation of images data was kept to a minimum for sufficient fidelity to be retained for further analysis and interpretation. Furthermore other spatial data such as geochemical anomalies, known lithology and mineral occurrences can be readily registered together, thereby creating simple Geographic information System (GIS) database.

Enhancement and study of images
The LANDSAT MSS bands, particular those of shorter wavelength, over a rain forest area like Malaysia are seriously degraded by atmospheric hase Onl the near infrared (band 4) band shows the sharpest details because haze penetration is better in this pant of the spectrum. Edge enhancement convolution operation was performed on this particular band to highlight geological structures especially the subtle features. From the varieties of convolution filters used, it was found that small directional. masks (Schowengerdt, 1983) are the most effective. Four directional filtered images were generated from vertical, horizontal and two opposite diagonal edge detectors. False color composites (FCC) were constructed from these directional edge enhanced image to accentuate edges which were common to some or all of them. FCC's were also constructed from the combination of bands 4, 3 and 2 because the lack of satisfactory band 1 images preluded the production of standard FCC's. All the images were contrast enhanced prior to interpretation.

Mapping of lineaments was carried out directly from the images displayed in the image processing system (International Image System model 75). Lineaments were drawn or trands interactively on the monitor screen first on contrast enhance band interactively on the monitor screen first on contrast enhanced band 4 images and FCC ( 4,3,2 = R, G, B) and then on the directional edge enhanced band 4 images and their FCC's. A graphic image or tectonic fabrication map was gradually constructed by adding new features to previous ones from the sequence of images. At the same time. in (Digital file was built. up. Hard copy map was written by cartographic plotter.

Integration of spatial data sets.
Digitized data from geochemical anomaly map and mineral occurrences file were converted to raster files were converted to raster files which were displayed as graphic images. The lineaments interoperated form the enhanced images were integrated with the map data by overlays and output as a hard copy (Figure 1).

Interpretation

Geological interpretation of images
Edges, which can be succinctly, defined as sudden changes in grey

Tectonic fabric map with mineral occurrences and geochemical anomales of the Jelai area, Malaysia

FIGURE 1.

Level, provide the basis from which lineaments are inferred from images. For this study, all lineaments were grouped under the broad definition of O' Leary (1976), who described a lineament as a mappable, simple or composite linear feature of a surface whose parts are aligned in a rectilinear or slightly corvilinear relationship and differs distinctly from the pattern or adjacent features and presumably reflects a subsurface phenomenon.

The images of the Sungai Jelai area show that lineaments abound in the region and that some occur in swamps (Figure 1). The dominant trends are N-S, NNW-SSE and NNE-SSW linear is probably mainly expressions of bedding trends in conformity with the regional trend. Some of these lineaments could be faults as N-S faults are prevalent in this area (Richardson, 1950), and an important trend associated with of gold mineralisation.

Amount the major lineaments, two are postulated as regional faults or faults zones, as both could be traced for more than 75 km across the map area. One of these is aligned roughly NNW-SSE and is defined by an almost continuous alignment of strong edges. The other is along the near western margin of the area, and coincides with most of the boundary of a conglomerate unit. It is curvilinear in nature, trends roughly N-S and is marked by weak or dark edges in edge enhancements and false colour composites. In aerial photographs (under stereoscopic interpretation), it is a very prominent and marked by linear scarp slopes. The NNW-SSE lineament is not prominent and is less continuous in aerial photographs. Other major lineaments recognized and postulated as faults fall into three main trends: N-S , NNW-SSE and NNE-SWW. These trends conform with those mapped on the ground (Lee, et al, 1982)

Circular features were also detected from contrast-enhanced images but less readily on directional edge enhancements. Circular features may be considered as a special case in the study of linear features, (Rowan and Latheram 1980) . They were identified by the same kind of features as for lineaments, i.e. morphology, pattern and tonal contrast, and were best enhancement. They did not seem to be arranged in regional pattern; same were associated with linear and some apparently independent of them.

Five circular features, ranging from 2 km to 10 km in diameter, from clearly obvious to faintly discernible, were. Delineated The most prominent of these is centered on the Bukit Ranjut syenite body in the centre of the map area. It is a composite circular feature, made up of at least two 'circles' overlapping indicate that the syenite body is a circular multiple intrusive complex.

Interpretation of integrated data.
The relationship between lineaments and mineralisation is analysed by patters (1978). Major linear features have long been used as regional guides to are deposits. Some lineaments represent zones of weakness through which mineralising fluids may be directed, and so such represent zones were mineral deposits are likely to occure.

Frequently, because of their breadth and continuity, many have been overlooked on the ground because the small and localised effects of a segment exposed discontinuously from one outcrop to the next were insufficient to allow identification. Satellite imagery can easily contribute to the identification and demarcation of such features because asynoptic view.

In the search for tin deposits, Hosking (1974) suggested that the system often play a major part in determining the sites of the tin provinces. He summarized the structural control of tin patterns as follows:
  • Lodes and vein systems are commonly the products of mineralisation of fractures systems.
Although to-data there have been no reported occurrences of mineralization within the Bukit Lima granitoid or its immediate surroundings, the intensively fractured area appears to be a good target for exploration as two geochemical anomalies area located at the western margin of this body. It was also discerned that the volcanic rocks, especially the largest unit in the north and a smaller area in the centre, have unique lineaments of their own trending WNW-ESE. These could be joints or sedimentary/volcanic layering.

Except for the Bukit Ranjut Syenite, circular features are located mainly in argillite rocks. These could be alteration zones or other surface expression of igneous intrusions. Since there is a known affinity of mineral deposits for near circular intrusions, the circular features are potential targets for mineral exploration.

Conclusion
Digital processing techniques has enabled extraction of more geological features of structural more geological features of structural interest from satellite images. The structural features could be registered with map data within the framework of a simpel GIS. The techniques have also enabled the production of a tectonic fabric map with integrated geological data and 'image maps' - cartographically rectified image.

Finally, the study has also demonstrated that remote sensing can be demonstrated that remote sensing can be tool in geological investigation, heavily vegetated terrain.

Acknoledgement
This study was carried out using the computer facilities of the Geography department., University College of London under training award by the British Council

References
  • Hosking, K.F.C., 1974. The search for deposit from which tin can be profitably recovered now and in the foreseeable future. Proceed, I.T.I. Fourth World Conference on Tin. V. 1 P. 21-85
  • Lai, K. H. , 1988. Tectonic fabric mapping in Peninsular Malasia by remote sensing and digital processing techniques. University of London, Unpublished M.Sc. thesis.
  • Lee, A.K., Gonzales, R.A., Tybally, Fateh Chand, & Troupe, A. 1982. Regional geochemistry of North Pahang. Geol. Survey Malaysia Geochemical Report, 2.
  • Lee, A.K. , Foo, K.Y. & Ong, W.S., 1986 Gold mineralization and prospects in North Phang Darul Makmur, Peninsular Malaysia Geol. Survey Malaysia EMR. Report.
  • O' Leary, D.W., 1976. Lineaments, linear, lineation: some proposed new standards of old terms. Geol. soc. Amer. Bull., Vol. 87, p. 1463 - 1469.
  • Peters, W.C. 1978. Exploration and mining geology. John wiley & Son. 696 p.
  • Procter , W.D., 1972. Geology and mineral resources. Benta area, Pahang. Geol. Survey Malaysia Map Bull. 4, 25 p.
  • Richardson, J.A., 1986 Remote sensing digital image analysis. Springer-Verlag, Berline Heidelgery, 281 p.
  • Richards, J.S., 1950. The geology and mineral resources of the neighbourhood of Chegar Perah, Pahang, Malaya. Geol. Survey Dept. Fed. Malaya, Mem. 5, 162, p.
  • Rowan, L.C. & Lathram, E. H., 1980. Mineral exploration. p. 553-605 in Remote Sensing in Geology, by Siegal and Gillespie. John Wiley & Son, 702, p.
  • Schowengerdt, R.A., 1983 Techniques for image processing and classification in remote sensing. Academic Press Inc. 249 p.