A satellite remote sensing survey was conducted in a disturbed logged-over hill dipterocarp forest of Ulu Tembeling in northern Pahang, Malaysia to identify and quantify the site disturbance classes due to road construction and logging activities. The merged SPOT data path/row K271-5341 was taken in October 3, 1988. The SPOT scene was obtained in a computer-compatible tape format and manual analysis was initiated by selecting a representative subsection of the scene that covered the study area. Ground truthing/field work was carried out and parameters such as causal factors, forms, sizes, shapes and patterns of soil disturbance were recorded, measured and correlated with image classification. Image interpretation, registration and classification were conducted for visual image analysis. The results showed that forest soil disturbance could be easily detected and monitored with 93% accuracy. Six classes of soil disturbance were quantified and recognized, namely (i) primary forest road, (ii) secondary forest road, (iii) skid road, (iv) skid trail, (v) secondary landing, and (vi) primary landing. By reference to maps of licensing applications and logging permits, legal and unpermitted logging activity can be well identified. However it is emphasized that interpreters should have a prior knowledge of the topography and the soil disturbance pattern of a logged-over forest area to be quantified and identified before actually beginning with image interpretation.
