Inventarisasi ekosistem bawah air dari Citra Landsat 7 ETM+ untuk penyiapan basis data wilayah pesisir :: Studi kasus di Kabupaten Kutai Timur Kalimantan Timur
ISTARNO, Dr.Ir. Haryono
2005 | Tesis | S2 Teknik GeomatikaWilayah pesisir dan lautan Indonesia terkenal dengan keanekaragaman hayati (biodiversity) laut terbesar di dunia. Oleh karena itu perlu pemikiran yang komprehensif dalam penanganan potensi yang berlimpah agar dapat dimanfaatkan secara berkesinambungan dalam jangka panjang. Untuk itu perlu dilakukan inventarisasi ekosistem bawah air yang mencakup wilayah pesisir yang luas. Kabupaten Kutai Timur, Propinsi Kalimantan Timur merupakan wilayah pesisir yang relatif cepat mengalami perubahan karena pembangunan dan belum tersedia basis data yang lengkap, maka perlu dilakukan inventarisasi untuk membangun basis data wilayah pesisir. Metodologi yang digunakan untuk inventarisasi berupa klasifikasi terselia (supervised classification) ekosistem bawah air secara digital. Koreksi kolom air berpengaruh pada ketelitian hasil klasifikasi dan indek invarian kedalaman dipengaruhi oleh koefisien pelemahan sinar di dalam air. Penginderaan jauh optis (optical remote sensing) digunakan untuk pengukuran kedalaman air meskipun kemampuan penetrasi air, maksimum 25 m. Metoda Juup membuat zona penetrasi kedalaman (depth of penetration = DOP) berdasarkan panjang gelombang sinar tampak dan interpolasi zona DOP mengacu pada nilai digital piksel, kemudian zona tersebut dikalibrasi. Ketelitian batimetri diperoleh dari perbandingan antara hasil prediksi dengan data lapangan. Perangkat lunak untuk pengolahan citra digital digunakan ENVI 3.6 dan penyajian hasil digunakan ArcView 3.2a . Hasil proses klasifikasi terselia dengan algoritma maximum likelihood pada mangrove, terumbu karang dan padang lamun serta pasir diperoleh ketelitian sebesar 69,25 % yang dilakukan tanpa proses koreksi kolom air (water column correction). Bila dilakukan proses koreksi kolom air dan algoritma parallelpiped, pada skema klasifikasi yang sama maka hasil ketelitian meningkat menjadi 75,37 %. Zona penetrasi kedalaman pada batimetri menghasilkan interpolasi zona DOP; zone 1, kedalaman (14,99 sampai 24,37) m; zone 2, kedalaman (4,09 sampai 14,99) m; zone 3, kedalaman (1,05 sampai 4,09) m dan zone 4, kedalaman (0 sampai 1,05) m. Uji ketelitian batimetri antara prediksi kedalaman dan data lapangan menunjukkan korelasi (R) 0,75. Hasil data spasial dan atribut disajikan secara interaktif serta basis data wilayah pesisir dapat terpenuhi dari hasil pengolahan citra digital hanya meliputi jenis habitat dan luasannya.
Indonesian oceans and coastal regions are well recognized around the world due to their biodiversities. In order to be able to make use of these resources continuously, comprehensive thinking is needed for managing this potency wisely. Therefore, under water ecosystems that cover wide coastal areas should be documented thoroughly. Sub-province of East Kutai, one of the rapidly growing regions in the Province of East Kalimantan, is located in the coastal area. This Sub-province has been changing rapidly as the impact of the urban development. Due to lack of database availability to support this development, it is necessary to conduct data inventory for building coastal area database. In this research, for the purpose of inventorying under water ecosystem database, supervised classification method was utilized digitally. In this method, accuracy of the classification results was determined by water column correction; whereas depth invariant index were influenced by under water light attenuation constant. Optical remote sensing which has ability to penetrate under water depth until 25 meters was applied to measure this depth. Juup method was employed to make Depth of Penetration (DOP) zone based on visible light wavelength. DOP zone interpolation process was then done by referring to the digital pixel values; these zones were then should be calibrated. Bathymetric accuracy was achieved by comparing these prediction results with the field data. Software ENVI 3.6 was utilized for conducting digital image processing, whereas ArcView 3.2a was for result presentation. Results show that supervised classification with Maximum Likelihood Algorithm applied for mangroves, coral reefs, sea grasses and sand has accuracy of about 69.25%. This method was conducted without utilizing water column correction. This accuracy increases significantly to 75.37% when the water column correction and Parallelpiped Algorithm were applied at the same classification scheme. Bathymetric depth penetration zones that are obtained from DOP zone interpolation are: Zone 1 ( 14,99 m to 24,37 m depth), Zone 2 ( 4,09 m to 14,99 m depth), Zone 3 ( 1,05 m to 4,09 m depth), and Zone 4 ( 0 to 1,05 m depth). Bathymetric accuracy test between depth prediction and field data shows Correlation Coefficient (R) at 0.75. The results of spatial and attribute data are presented interactively; coastal area data base that was collected sufficiently from digital image processing covers only habitat types and their coverage.
Kata Kunci : Citra Landsat 7ETM+ ,Basis Data Ekosistem Wilayah Pesisir, under water ecosystem, digital classification, bathymetry, coastal area database. xvii
