f resource
raphs and
lc content
rtographic
8.
parate map
r with the
main data
, prevailing
station and
ant of data
. It will be
pnever they
itories and
}.
a statistical
particular"'.
1dsats which
ery in areas
This stage will result in photomaps of selected regions at the scale of 1 : 250,000.
Eventually, the whole country will be covered with thematic photomaps at this scale
to facilitate planning at the national and provincial levels.
Progress of the reconnaissance remote sensing depends, to a large extent, on the
availability of good quality Landsat imagery. However, a systematic Landsat
coverage can be expected only after construction of a Landsat ground receiving
station and data processing facility in Indonesia.
Shortage of Landsat imagery with quality suitable for photomapping allowed
only limited number of cartographic applications in Indonesia. Recently was
completed a resource photomap at 1 : 250,000 of Lombok Island in the Nusa
Tenggara archipelago. Eight vegetation and land use categories were delineated by
computer-aided classification of Landsat multispectral images recorded on computer
compatible tapes. Thematic data were overprinted on the base photomap, which was
mosaicked from enlargements of Landsat film images.
The map was favourably accepted and the same procedure will be used in other
regions when suitable Landsat imagery becomes available. Methodology of the
Lombok Island mapping was described in detail by Kalensky et.al. (1978).
Reconnaissance resource surveys based on conventional interpretation of
Landsat film enlargements are in progress in Southern Sumatra and in Central and
South-East Sulawesi. This technique enables, with relatively modest outlay of
money, quick categorization of extensive areas, for which little recent data have
been available, into broad vegetation and land use classes. Conventional
interpretation techniques were helped by improved reproduction quality of Landsat
images, particularly when processed with edge enhancement and printed by a laser
beam recorder.
Further increases of classification level and accuracy in resource surveys from
satellite imagery depend on improvements of their remote sensing payloads. New
design of two Return Beam Vidicon (RBV) cameras on board of Landsat-3 yields
nominal ground resolution of 40m, in comparison with the 80m yielded by RBV
cameras in Landsats-1 & 2. Planned lower orbital altitude of Landsat-D, projected
for launch in 1981, combined with a new design of multispectral scanner (Thematic
Mapper), will yield nominal ground resolution of 30m. (Kalensky et.al., 1977).
However, the breakthrough for cartographic applications of satellite imagery may be
provided by a ZEISS (Oberkochen) metric (photogrammetric) camera, which will be
part of the Spacelab-1 remote sensing payload in 1980.