KLASIFIKASI TUTUPAN LAHAN DATA LANDSAT-8 OLI MENGGUNAKAN METODE RANDOM FOREST

Land Cover Classification of Landsat-8 OLI Data Using Random Forest Method The latest information of land cover and land use is needed in regional development planning and environmental monitoring. One of the method to obtain this information is through remote sensing satellite image data processing. Landsat-8 OLI is one of the remote sensing satellite images with a multispectral spatial resolution of 30 m and a temporal resolution of 16 days. This study aims to classify the land cover in most areas of Pidie Regency using the random forest method based on Landsat-8 OLI imagery and to calculate the accuracy level of classification results. Information extraction of land cover was carried out using the random forest method with a proportion of 70% for training data and 30% for testing data. Then the accuracy assessment of the classification results uses confusion matrix method. The results of land cover mapping in most areas of Pidie Regency show that the land cover class of rice fields dominates the study location with an area of 22,598.20 ha (29.22% of the total study area). The results of the land cover classification showed an overall accuracy value of 89.53% and a kappa value of 0.91.