ARCHITECTURE OF A DECISION SUPPORT SYSTEM WITH SECURE DATA PROCESSING IN FLOODED AREAS

Abstract

This paper introduces a decision-support information technology system aimed at improving the accuracy and efficiency of flood risk prediction using a combination of satellite data and hydrological modelling. The system architecture incorporates various components, including cloud computing services, digital elevation models, hydrological simulation tools, and big data analytics. These elements work together to monitor comprehensively and precisely forecast flood-prone areas. By integrating data from the Sentinel-2 satellite and the Shuttle Radar Topography Mission  digital elevation model, the system significantly improves the identification of flood zones. This approach achieves a spatial resolution that minimizes the prediction error to within 5%, critical for accurate flood risk management and disaster preparedness.

A vital advantage of the developed system lies in its ability to handle real-time data processing while maintaining operational stability under high loads. The system's cloud-based infrastructure is designed for scalability, allowing for rapidly expanding processing capacity as the volume of incoming data increases. This ensures that the system can accommodate large datasets without experiencing significant delays. In testing, the average response time for query processing has remained between 400 and 500 milliseconds, even when subjected to a load of up to 20 simultaneous computational units. It demonstrates the system's robustness and capability to operate under conditions of high demand, which is essential for timely decision-making during flood events.

One of the system's core functionalities is the generation of interactive flood risk maps, which provide detailed visualizations of potential flood zones and develop an interactive flood hazard map for the city of Zaporizhzhia, Ukraine. The map demonstrated 90-95% prediction accuracy, making it a valuable tool for local authorities and emergency services to manage flood risks more effectively. Visualizing and forecasting flood zones in such detail allows for quicker and more informed decision-making in emergencies, potentially saving lives and reducing the economic impact of floods.

The proposed system significantly impacts disaster risk reduction, particularly in areas susceptible to flooding. The technology can support proactive flood management strategies by enhancing the precision of flood forecasts and enabling real-time data processing. This system improves the accuracy of flood zone detection and facilitates more efficient management of emergency responses and mitigation efforts. Future applications could extend beyond flood prediction to include other types of natural disasters, further expanding its utility in disaster management. The results of this study underscore the potential for this system to be adopted as a critical tool for mitigating the effects of natural disasters and optimizing crisis management operations.