Assessment of Forest Cover Changes Using Landsat TM for Langkawi Islands, Malaysia

One of the major environmental issues today is the rapid conversion of tropical forest to agriculture, pasture, human settlement, urban area, and many other land uses. Under these circumstances, the need for conservation and effective management of forest is imperative. One of the present technologies being used in the monitoring of environment changes is remote sensing. This study was undertaken to verify the suitability and capability of LANDSAT TM in monitoring forest changes on langkawi Islands. Detection of forest cover change was performed using multi temporal LANDSAT data taken in 1992 and 1996, with the support of existing land use, topographic, and forest resource maps. The data were classified using maximum likelihood classifier (MLC) and overlay to generate forest change. Principal component analysis (PCA) was also used to detect changes where the multi temporal of six bands data were combined and treated as a single 12-dimensional data. A new set of images were obtained from a PCA color composites of PC2, PC3 and PC4 and were then classified using supervised classification to detect forest changes. It was found that MLC and PCA gave high overall accuracy of 90 per cent. However, MLC was found to be more accurate because of its better delineation along the forest cover changes in multi temporal data. The study quantified that the rate of deforestation of Langkawi Islands is about 235.18 halyr with a n accuracy of 93 percent. Factors causing forest cover changes include the expansion of tourism industry, encroachment of forest areas by local people and development of socio-economic activities such as residential areas, road network, quarries, jetties and agriculture.

PDF FH_1999_12_A.pdf Download (1MB)

Actions (login required)

View Item