Analisis Prediksi Pola Aktivitas Ilegal Menggunakan Model Maximum Entropy di Taman Nasional Gunung Leuser

Abstract

Gunung Leuser National Park (GLNP) is one of the most important conservation areas in Indonesia that faces various threats from illegal activities such as illegal logging, wildlife hunting, non-timber forest product (NTFP) extraction, and unauthorized land use. This study aims to predict spatial patterns of illegal activities in GLNP using the Maximum Entropy (MaxEnt) method based on R Studio software. The data used include occurrence points of illegal activities and environmental variables such as distance from roads, elevation, slope, distance from rivers, settlements, forest edges, area trails, area access, and vegetation index (NDVI). The model was developed by splitting training and testing data at a 70,30 ratio and validated using the Area Under the Curve (AUC) value. The modeling results indicate that illegal activities tend to be concentrated in areas with high accessibility, particularly near roads and at low elevations. The evaluation results demonstrate that the model achieved AUC values of 0.92 for logging, 0.84 for hunting, 0.90 for NTFP extraction, and 0.94 for unauthorized land use, indicating good predictive accuracy. These findings are expected to serve as a foundation for spatial-based policy formulation in efforts to mitigate threats and manage conservation areas effectively and adaptively.