Open AccessPEMANFAATAN CITRA LANDSAT 8 OPERATIONAL LAND IMAGER (OLI) UNTUK KLASIFIKASI TUTUPAN LAHAN DI TAMAN NASIONAL GUNUNG MERBABU
Author(s) -
Diyanti Isnani Siregar,
Adnin Musadri Asbi
Publication year - 2020
Publication title -
wahana forestra/wahana forestra
Language(s) - English
Resource type - Journals
eISSN - 2548-608X
pISSN - 1858-4209
DOI - 10.31849/forestra.v15i2.4731
Subject(s) - land cover , remote sensing , geography , land use , forestry , pixel , cartography , national park , environmental science , ground truth , computer science , ecology , artificial intelligence , biology , archaeology
Gunung Merbabu National Park (TNGMb) is a conservation area with a high level of biodiversity. Information on land cover is very important in making ecological management policies in conservation areas. Proven Remote Sensing technology produces precise information on land cover in a time and cost-effective manner. This study uses Landsat 8 imagery in TNGMb land cover classification process. Maximum Likelihood approach is used because it uses a probability calculation basis. A configuration matrix table between training data and reference data is made to test the accuracy of land cover classification. Reference data refers to Google Earth Pro high-resolution imagery. Results showed that the most extensive land cover type was secondary dryland forest with total of 23393 pixels classified as equivalent to 2113.54 hectares (34.5% of the total classification area. The open area, built-up area, and rice field/vegetable garden each have an area of 12.08 Ha; 11.02 Ha; and 170.96 Ha, of which part of the area is in enclaved areas within the TNGMb area. The accuracy test shows the Kappa Coefficient of 86.25%, User's Accuracy Average, Ground Truth Average, and Overall Accuracy respectively 89.62%; 85.42%; and 88.33%. Overall Accuracy shows that 88.33% of the total pixels represent each classification correctly.