information to differentiate between various forest 
or other agricultural and cultural cover types. 
Radar data, which utilizes the active microwave 
portion of the spectrum, can provide important 
additional information about terrain surfaces and 
forest and agricultural plants canopies. Satellite 
remote sensing often used to obtain information, 
on actual land cover/land use and to monitor 
changes in the regions with often cloudy 
conditions such as indonesia. However, radar data 
can be independent of the atmospheric conditions 
and the time of the day. Therefore, data supplied 
by Synthetic Aperture Radar (SAR) such as the 
European Remote Sensing Satellite (ERS-1) has 
potential for environmental applications. The 
main objective of this research was to detect 
changes in the land use of the tropical rain 
forested area, including the forest land, agricultural 
land and other land resources near the forest using 
multi-temporal optical data (e.g., Landsat MSS and 
TM and Spot XS data) and microwave (e.g. ERS-1 
and JERS-1 radar data) satellite images in two 
Indonesian test site (i.e. Bengkulu, South Sumatra 
and Jambi, Central Sumatra). 
2. MATERIALS AND METHODS 
2.1 Study Areas 
Two test sites were used for this research. The 
first study area is located on the south west coast 
of Sumatra, Indonesia. It is situated between 
3°35’00" and 3°51°00" Latitude South and between 
102°12°00" and 102°45°00" Longitude East. The 
area lies in the Bengkulu province and covers 
about 2,100 km?. In general, the topography of 
this study area can be considered undulating 
through rolling to mountainous. The altitude 
varies from O to 1800 m asl. The settlements are 
located along the main and secondary roads. 
Bengkulu is one of the nine major cities in 
sumatra and it is the seaport city on the south- 
western coast of Sumatra. The main river in the 
area is the Air-Bengkulu river. Most of the 
irrigation systems are of the conventional gravity 
type. 
The major land use/cover classes exist in the area 
are: cultivated land which consist of rain fed 
agriculture land and irrigated agriculture land; 
agro-forestry, which is a combination of forest 
trees and crops; forest plantations; shifting 
cultivation, is characterized by scattered arable 
plots in the forest area which involves the 
cutting/burning of the natural vegetation (forest, 
scrub or grass); natural forest is consist of 
highland and lowland rain forest; mixed cover of 
bushes Grass/Scrub/shrubland; and urban or 
settlements which include villages, towns and 
cities. 
The second area of study lies approximately from 
316 
latitude 1° 15 to 1° 45 (S) and from longitude 
102° 15 to 102° 45 (E). This area is located in the 
Bungotebo county, northwestern part of Jampi 
province, Central Sumatra. The river Batang Har 
divided the area into two part. The form of 
topography in this site is relatively flat land with 
a very small part of mountains. The altitude varie 
between 25 to 160 m. Verstappen (1973), used the 
peneplain to refer to this type of terrain. Part of 
the peneplain are covered with acid volcanic tuff 
particularly near the Bukit Barisan mountain 
Podzolic red yellow soil type is dominant in the 
study area. The land cover in this test site is 
dominated by forests. Plantations of rubber, oi 
palm are found in the area. Most of rubber 
plantations were planted by the indigenous people 
and normally located at the boundary of or inside 
forest. New settlements that cause by 
transmigration is located in the south and southeast 
of the study area. Other agriculture cover types 
like rice field also found in the area. 
2.2. Data Used 
For the first test site, the remote sensing data were 
acquired by both passive and active systems, 
Landsat-1 MSS data of 13 June 1973 were used to 
classify land use in 1973 and to compare with land 
use data of 1990. The spatial resolution was 80 x 
80 m square and spectral bands were representing 
the green, red, and two in infrared part of the 
spectrum. The Landsat-5 TM data, acquired on 29 
December 1990 were used to study recent land 
use. The spectral bands used were blue, green, 
red, NIR, two MIR, and one thermal IR. The 
spatial resolution was 30 x 30 m square. 
ERS-1 Synthetic Aperture Radar (SAR) data 
acquired on 20 July 1993 in C-Band (5.6 cm 
wavelength) and VV polarization were used. The 
spatial resolution was 12.5 m with a swath width 
of 100 km. The incidence angle was 23°. During 
1994, two other images were acquired of the study 
area. Unfortunately, because the new images wer 
obtained in a different acquisition mode, there was 
a significant shift from the image acquired in the 
preceding year. Two JERS-1 Synthetic Aperturt 
Radar data sets acquired on 6 March 1993 and 20 
May 1994 were also used for the study. The data 
were obtained in the L-Band (23 cm wavelength) 
HH polarization and an incidence angle of 35. 
The spatial resolution was 12.5 m and the swath 
width was 75 km. 
For the second study area the following data wer 
available: Landsat-5 TM data of September 13 
1993, Spot XS data of March 21, 1993, ERS! 
images of October 17, 1993, June 6, 1994, and 
July 7, 1994, and JERS-1 of August 16, 1993. 
2.3 Method Used 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996