Selected Presentations

Home > Presentations > Abstract details

Title

TIME SERIES LANDSAT IMAGES CLASSIFICATION IN GRASS GIS CASE STUDY IN UPPER SERAYU WATERSHED, INDONESIA

Abstract

The increasing of cultivation practices has reported causing rapid land use/land cover changes in the upper Serayu watershed during 1989 to 2009. However, the spatial information of the time series land use/land cover of the study area is limited. This information is important to see what, where and when the land use/land cover has changes including the extent. Fortunately the series of Landsat imagery from USGS free source can be used to monitor the land use/land cover changes for four to six years intervals. The image classification was done by using sequential maximum a posteriori (SMAP) classifier for the most present image. The training sample was obtained from the field observation. For the previous images classification process, the spectral angle and magnitude signatures from the most present image classification result was extracted and used to obtain training samples. According to this, topographic and radiometric correction should be done beforehand. The final classification process of the previous images was done by SMAP classifier. Finally, the post classification change detection method was used to determine the change types including the extent. All the operations were done by employing GRASS GIS software.

The classification results obtained eight classes of land use/land cover of the study area: a. built up area; b. paddy field; c. dry land cultivation; d. forest; e. shrub; f. plantation; g. grassland; h. water body. The plantation in the area is relatively sparse and can be differentiated from the forest cover. The time series monitoring showed that the forest cover was decreasing 1547 ha per year or 74 % during the period. On the other hand plantation was increasing 938 ha per year or increasing 119%.

Authors

Andry Rustanto - Centre for Applied Geography Research, Department of Geography, University of Indonesia
Dhruba Pikha Shrestha - Institute for Geoinformation Science and Earth Observation