Abstract
This article discusses a GIS-based method for predicting the impact zones of launch vehicle separating stages based on the integration of deterministic trajectory modeling, stochastic uncertainty analysis, and spatial risk assessment. The proposed approach combines stage performance parameters, meteorological data, and geospatial information within a single computational framework that enables the calculation of a nominal trajectory, the generation of a probabilistic impact zone, and the construction of digital risk maps. To verify the approach, three scenarios were considered, differing in stage mass, wind speed, and separation angle. A comparison was made between a classical ballistic model, a stochastic Monte Carlo model, and a model with elements of artificial intelligence and machine learning. It is demonstrated that the use of modern methods allows for a reduction in the average prediction error to 60 m, a decrease in the standard deviation to 25 m, and a reduction in the percentage of overshoots to 1%. Spatial risk analysis revealed that the most vulnerable areas are those with high population density and critical infrastructure. The results confirm that the integration of trajectory modeling, uncertainty analysis, and GIS tools improves forecast accuracy and the practical value of the system for supporting decision-making in ensuring space launch safety.