Land-Cover/Land-Use Random Forest Classification Using Sentinel 1A Radar Imagery

Sumatra Journal of Disaster, Geography and Geography Education
Vol 7 No 1 (2023): Sumatra Journal of Disaster, Geography and Geography Education (June Edition)

Extraction of existing land use/land cover information can be obtained using remote sensing data. One of them is by using Sentinel 1A Imageas a Synthetic Aperature Radar (SAR) active system image with 20 meters resolution and 12 daystemporal resolution. This study aims to map land use/land cover in the city of Padang using the random forest classification (RFC) method through active system remote sensing data extraction. Accuracy test of the classification results using the confusion matrix method with overall accuracy of 85% and kappa value of 0.78. The most dominant land use in the study area was forest land use with a total area of ​​30,285 Ha reaching 43.83% of the City of ​​Padang.