Algoritma indeks vegetasi mangrove menggunakan sateIit Landsat ETM+
View/ Open
Date
2013-08Author
Endriani Arhatin, Risti
Ika Wahyuningrum, Prihatin
Metadata
Show full item recordAbstract
Monitoring mangrove dengan metode konvensional sangat sulit dilakukan. Sistem penginderaan jauh merupakan salah satu alternatif dalam mengatasi kendala dalam melakukan inventarisasi mangrove dengan cakupan areal yang luas. Tujuan penelitian ini melakukan validasi akurasi dari data Landsat-7 ETM+ dalam menduga kerapatan kanopi mangrove. Data spasial yang dipergunakan adalah citra Landsat ETM+ tanggal perekaman 21 Mei 2002 (path/rowt: 116/059). Data lapangan yang diperlukan adalah data kondisi fisik mangrove, pengamatan dilakukan pada beberapa lokasi yang berbeda, pada setiap lokasi dibuat transek. Setiap transek diplot dengan ukuran (30 x 30) meter2. Analisis data meliputi koreksi radiometrik dan koreksi geometrik, penajaman citra dan klasifikasi citra. Setelah itu dilakukan uji ketelitian separability transformasi divergency dan analisis komponen utama. Hasil analisis menunjukkan algoritma vegetasi indeks yang paling baik adalah Green Normalized Difference Vegetation Index (GNDVI). Analisis PC4 diperoleh hasil persamaan {2,5180 (-0522X2 - 0,497x3 -0,470X4 - 0,510x5)} + {1,3057 (-0,462X2 - 0,515x3 - 0,548x4 - 0,469x5)}. Traditional field monitoring of mangrove would be very difficult to survey. Remote sensing is a promising alternative to answer the problem for large-scale tropical mangrove management, The objectives of this research are to validate the accuracy of remote sensing data, namely Landsat-7 ETM+ images, for estimating mangrove forest canopy. The spatial data was Landsat ETM + recording on May 21, 2002 (path/ row: 116/059). Field data required physical condition of mangrove which observed using transect at seferal different locations. Transect plotted on size (30 x 30) metres. There were several step for data analysis which were radiometric correction, geometric correction, image enhancement and image classification. After that, it is analyzed by accuracy separability divergency transformation test and principal component analysis. The result of the research shows that the best vegetation index algorithm is Green Normalized Difference Vegetation Index (GNDVI), PC4 Result is {2,5180 (-0522X2 - 0,497x3 -0,470X4 - 0,510x5)} + {1,3057 (-0,462X2 - 0,515x3 - 0,548x4 - 0,469x5)}.