High Heterogeneity LULC Classification in Ujung Kulon National Park, Indonesia: A Study Testing 11 Indices, Random Forest, Sentinel-2 MSI, and GEE-based Cloud Computing

Research reveals that water bodies constitute the largest land use-land cover (LULC) class in Ujung Kulon National Park, Indonesia, covering 45.44% of the park's area, while primary forest accounts for approximately 19.53%. The study employed Sentinel-2 MSI imagery, classified using a Random Forest algorithm and tested against 11 spectral indices on Google Earth Engine (GEE) cloud computing platform. Despite achieving 10 LULC classes, discrepancies in spatial data were noted, underscoring the need for further evaluation of index combinations, training data, and classification algorithms to refine mapping accuracy in high-heterogeneity landscapes like UKNT.

Asy’Ari, R., Ranti, A., Rahmawati, A. D., Zulfajrin, M., Nurazizah, L. L., Putra, M. C. A., Khairunnisa, Z. N., Prameswari, F. A., Pramulya, R., Zamani, N. P., Setiawan, Y., Sudrajat, A., & Anggodo. (2023). High Heterogeneity LULC Classification in Ujung Kulon National Park, Indonesia: A Study Testing 11 Indices, Random Forest, Sentinel-2 MSI, and GEE-based Cloud Computing. CELEBES Agricultural, 3(2), 82–99. https://doi.org/10.52045/jca.v3i2.381