Identifying and monitoring rockfall- and landslide-prone areas is crucial for effective risk mitigation, as these geohazard events present significant challenges and hazards to railway operations and safety. Current geohazard prediction models do not identify specific areas where an event could initiate under favorable conditions with triggering event. This study used satellite data and an SBAS threshold stacking method to improve geohazard prediction models for railway systems. The research demonstrates the successful application of the SBAS threshold stacking method in regions characterized by high and low coherence.