AI Modeling for Sustainable Management and Policy Making in the Gunungsewu Karst Landscape

• Putu Sudira

  Yogyakarta State University

  Author

• Yoga Sahria

  Yogyakarta State University

  Author

• Imsak Rochmadi

  Universitas Gunung Kidul

  Author

The Gunungsewu Karst landscape faces severe and escalating land and water management challenges, heavily driven by its unique hydrogeological vulnerability, rapid anthropogenic land-use changes, and acute seasonal water scarcity. Traditional spatial planning methods often fail to capture the complex, non-linear environmental dynamics inherent to this fragile ecosystem. To address these critical issues, this study develops a predictive Geographic Artificial Intelligence (Geo-AI) framework. This approach integrates advanced spatial machine learning algorithms, specifically Random Forest (RF) and Support Vector Machines (SVM), with multi-temporal satellite imagery, Geographic Information System (GIS) thematic layers, and detailed hydro-topographical datasets to model future land-degradation and groundwater-availability patterns across the region. The resulting predictive model demonstrates high scientific accuracy, achieving an Area Under the Curve (AUC) of 0.92 and an overall classification accuracy of 89.5% in simulating environmentally vulnerable zones. The spatial projections successfully reveal critical future hotspots of ecosystem degradation, primarily accelerated by unmanaged agricultural expansion and climate-induced fluctuating precipitation patterns. These high-fidelity spatial insights provide necessary, actionable intelligence for sustainable regional planning and resource conservation. Ultimately, this Geo-AI predictive model serves as a robust empirical decision-support tool for policymakers. It offers a scientific foundation to optimize scarce water resource allocation, enforce strict environmental zoning regulations, mitigate land degradation, and implement targeted ecological conservation strategies, thereby ensuring the long-term socio-ecological sustainability of the fragile Gunungsewu Karst landscape.

• PDF