00
30
imagery also supplementsthe level II data
by specific tone and association. Built-
up areas where concentration is less, are
found to be having reddish pink back-
ground due to its association with vegeta-
tion. The habitats are easily distingui-
shable in FCC by the particular colour
and rough texture. Also, habitats are
found to be associated with river/stream
or other linear features presumed to be
roads and/or railways. The roadsand rail-
ways are mostly distinguishable in Band 2
imagery. Though, tea-gardens are found
at the edge of forest, they are found
near habitats in some placés.
RESULT AND DISCUSSION
The land cover classes which could be
identified and mapped are shown in fig l.
In preparing the map, we referred to the
topographical sheets (Table 1) as needed.
But there is a time lapse of about 18
years between the survey conducted for
topographical sheet and the date of acqu-
isition of the imagery, 1989 which might
result to some misclassification. But
due care and attention was given on this
point all throughout during the prepara-
tion of map from imagery. Again, due to
very low scale (1:1 million) of the imag-
ery, delineation was not that perfect as
could be achieved by use of imagery of
higher scale. Also, the higher level of
discrimination of features could not be
done due to this reason. The forest
cover could only be classified into dense
and medium types,not the tree species. It
is observed that density of vegetation is
largely controlled by altitude. Although
altitude details have been taken from
topographical maps, this aspect can be
solved by classifying vegetation using
digital satellite data overlayed with
Digital Terrain Modelling (Andrew, 1989).
CONCLUS ION
The IRS imageries can be used effecti-
vely for mapping different land cover ty-
pes to be used for macro level planning,
management and conservation. The False
Colour Composite synthesised by combining
bands2,3,4 provides an added advantage in
carrying out this work. However, multi-
date multi-seasonal imageries of higher
Scale would make this a very important
tool for quick mapping and updating the
land cover of vast area.
ACKNOWLEDGMENT
This study was supported by Regional
Engineering College, Silchar, India under
undergraduate course program. Authors
are thankful to Dr M.C. Borgohain, then
Principal of R.E.C. Silchar and Dr B.U.A.
Barbhuinya, Head, Civil Engineering Dep-
artment, R.E.C. Silchar for making nece-
ssary financial arrangement.
REFERENCES
o Andrew K. Skidmore, 1989. An expert
system classifies eucalyptus forest
types using Thematic Mapper data and
a digital terrain model, Photogramme-
tric Engineering and Remote Sensing,
Vol.55(10) Oct 1989 (pp 1449-1464).
o Bowden L.W and Pruitt E.L, 1983 Editor.
Manual of Remote Sensing. American
Society of Photogrammetry. Falls Ch-
urch, Virginia Vol.2 (pp 882-891).
o Ghosh, Rekha and Ghosh, Dipen N, 1990.
Land use Map of Jharia Coal-field
Eastern India Aided by Remote Sensing.
Journal of Indian Society of Remote
Sensing,Vol.18,No.1 and 2 (pp 23-28).
o Gupta, D.M. Bgd and Munshi, M.K, 1983.
Land use and Forestry studies of Hima-
chal Pradesh. National Natural Resou-
rces Management, National Seminar pro-
ceeding, Hyderabad.
o Mikihiro, Ioke and Masatd Koda, 1986.
Performance of Landsat-5 Thematic
Mapper data in Land cover classifica-
tion. International Journal of Remote
Sensing, Vol.7,No.12 (pp 1715-1728).
o Pant, D.N. and Roy, P.S. 1990, Vegeta-
tion and Land use Analysis of Aglar
Watershed using satellite Remote Sens-
ing Technique Journal of the Indian
Society of Remote Sensing.Vol.18, No.4.
o Welch, R, H.C. Lo and C.W. Pannel1,1979.
Mapping China's New Agricultural Lands.
Photogrammetric Engineering and Remote
Sensing, Vol.XLIV, No.9.
439
