Plant diseases pose a severe threat to the economic sustainability of rubber plantations (Hevea brasiliensis), requiring advanced analytical models for effective management. Traditional monitoring often relies on uniform, reactive treatments that fail to capture spatial-temporal dynamics. This study aims to identify spatio-temporal distribution patterns and evaluate disease risk levels of White Root Rot (Rigidoporus microporus) at the PT. Bridgestone Sumatera plantation using longitudinal observation data from 2021 to 2025. To achieve this, the research implements the Spatio-Temporal DBSCAN (ST-DBSCAN) algorithm, integrating geographical coordinates, temporal duration, and attribute severity to cluster infection incidents. The experimental results successfully identified three distinct disease hotspots with an exceptional spatial validation performance, yielding a high Silhouette Score of 0.846. Based on calibrated temporal and severity thresholds, all identified clusters were classified as High-Risk Persistent zones, remaining active across the entire observation period with average severity scores exceeding 2.0. In conclusion, the proposed ST-DBSCAN framework successfully fulfills research expectations by providing a robust, data-driven decision-support tool. Practically, these findings enable estate managers to prioritize targeted, proactive interventions - such as intensive soil treatment and structural root isolation - rather than applying uniform fungicide treatments. Future research should integrate environmental variables and conduct parameter sensitivity analyses to capture smaller, emerging outbreak patterns.
Copyrights © 2026