ABSTRAK

Ekosistem mangrove Desa Pegatan Besar, Kecamatan Takisung, Kabupaten Tanah Laut merupakan kawasan pesisir yang telah dikembangkan sebagai kawasan ekowisata sejak tahun 2016, namun tetap menghadapi tekanan alih fungsi lahan di sekitarnya. Penelitian ini bertujuan menganalisis perubahan luasan tutupan mangrove serta nilai Normalized Difference Vegetation Index (NDVI) pada bulan April 2022 dan April 2026 menggunakan citra Sentinel-2. Analisis dilakukan melalui klasifikasi terbimbing (supervised classification) menggunakan perangkat lunak ArcMap 10.8 dan Envi 5.1, uji akurasi dengan matriks konfusi berdasarkan 50 titik Ground Control Point (GCP), serta deteksi perubahan (change detection) dengan metode tumpang tindih (overlay). Hasil penelitian menunjukkan luas tutupan mangrove meningkat dari 54 ha (2022) menjadi 58 ha (2026), atau bertambah 4 ha (7,41%), di tengah dominasi kelas Perkebunan yang meningkat dari 1.436 ha menjadi 2.377 ha. Analisis NDVI menunjukkan luas kategori kerapatan lebat meningkat dari 44,70 ha menjadi 52,75 ha, sedangkan kategori sedang menurun dari 5,55 ha menjadi 2,05 ha. Uji akurasi klasifikasi tahun 2026 menghasilkan overall accuracy 84% dan kappa accuracy 0,79 (substantial agreement), dengan kelas mangrove mencapai producer's accuracy dan user's accuracy sebesar 100%. Peningkatan luasan dan kerapatan mangrove diduga dipengaruhi oleh program ekowisata berbasis masyarakat serta proses akresi sedimentasi di kawasan muara sungai.

Kata Kunci: Mangrove, NDVI, Sentinel-2, Klasifikasi Terbimbing, Perubahan Luasan

ABSTRACT

The mangrove ecosystem of Pegatan Besar Village, Takisung District, Tanah Laut Regency is a coastal area that has been developed as an ecotourism site since 2016, yet it remains under pressure from surrounding land-use change. This study aimed to analyze changes in mangrove cover area and Normalized Difference Vegetation Index (NDVI) values in April 2022 and April 2026 using Sentinel-2 imagery. The analysis was carried out through supervised classification using ArcMap 10.8 and Envi 5.1, accuracy assessment with a confusion matrix based on 50 Ground Control Point (GCP), and change detection using the overlay method. The results showed that mangrove cover increased from 54 ha (2022) to 58 ha (2026), an increase of 4 ha (7.41%), amid the dominance of the Plantation class which expanded from 1,436 ha to 2,377 ha. NDVI analysis showed that the dense category increased from 44.70 ha to 52.75 ha, while the moderate category decreased from 5.55 ha to 2.05 ha. The 2026 classification accuracy assessment produced an overall accuracy of 84% and a kappa accuracy of 0.79 (substantial agreement), with the mangrove class achieving a producer's and user's accuracy of 100%. The increase in mangrove area and density is presumed to be influenced by a community-based ecotourism program and estuarine sedimentation accretion.

Keywords: Mangrove, NDVI, Sentinel-2, Supervised Classification, Area Change