INTEGRATED GEOPHYSICAL AND MACHINE LEARNING APPROACHES FOR MAPPING ACTIVE FAULT SYSTEMS AND SEISMIC HAZARD ZONATION IN NORTHERN PAKISTAN

Abstract

Northern Pakistan is characterized by complex tectonic interactions resulting from the ongoing convergence between the Indian and Eurasian plates, making it highly susceptible to seismic hazards. Accurate mapping of active fault systems and reliable seismic hazard zonation are therefore critical for effective risk mitigation and sustainable development. This study developed an integrated framework combining multi-source geophysical datasets with advanced machine learning techniques to enhance fault detection and seismic hazard assessment within key tectonic regions, including the Himalayan Fold and Thrust Belt and the Chaman Fault System.Geophysical data, including seismic records, remote sensing imagery, digital elevation models, and derived geomorphological parameters, were processed and analyzed within a geographic information system environment. Machine learning models—Random Forest, Support Vector Machine, and Artificial Neural Networks—were implemented to classify fault zones and predict seismic hazard levels. Model performance was evaluated using statistical metrics such as accuracy, F1-score, and AUC-ROC.The results indicated that the Random Forest model achieved the highest predictive accuracy and robustness, effectively capturing nonlinear relationships among geophysical variables. Feature importance analysis revealed that proximity to faults, lineament density, and seismicity density were the most significant factors controlling hazard distribution. The generated seismic hazard zonation maps identified high-risk areas concentrated along major tectonic structures, providing improved spatial resolution compared to conventional methods.This study demonstrates that integrating geophysical data with machine learning significantly enhances the accuracy and reliability of fault mapping and seismic hazard assessment. The findings provide valuable insights for disaster risk reduction, infrastructure planning, and policy development in seismically active regions.