<![CDATA[Surface deformation is a geodynamic phenomenon that reflects continuous and episodic tectonic activity and has significant implications for environmental conditions, including increased susceptibility to landslides, land degradation, and infrastructure instability. Although Synthetic Aperture Radar (SAR) technology enables high-resolution deformation monitoring, most studies still focus primarily on deformation analysis without comprehensively integrating environmental indicators for spatial planning. This study aims to perform active deformation zoning by integrating Sentinel-1 deformation data with vegetation conditions represented by the Normalized Difference Vegetation Index (NDVI) derived from Sentinel-2 imagery. Deformation data were processed using time-series analysis with LiCSBAS to obtain cumulative displacement and subsequently integrated with NDVI using unsupervised machine learning approaches based on anomaly detection, namely Isolation Forest and Local Outlier Factor. This method was applied to the western region of Java Island. The results demonstrate that the proposed method effectively classifies the study area into zones with varying levels of potential tectonic risk. Areas characterized by high anomaly scores and low NDVI values are interpreted as highly potential risk zones with potential environmental degradation, whereas areas with low anomaly scores and high NDVI values are relatively stable. Spatially, the highest risk zones are identified in western Java Island, particularly in North Jakarta, Sukabumi, and Cianjur, where active deformation is dominant.]]>
Copyrights © 2026
