Implementation and evaluation of 3D cadastral databases in on-premises and cloud environments

Traditional cadastral databases are unable to effectively manage the complexity and scale of modern 3D land administration data. Despite significant advances in conceptual data modelling, physical implementations continue to rely on conventional technologies that do not fully exploit contemporary computing environments. This study proposes a component-based architecture for implementing and evaluating various physical 3D cadastral database systems. The architecture is developed across multiple computing environments, encompassing on-premises and cloud-based deployments. It incorporates exchange file formats, relational databases, and non-relational databases to address the diverse structured, semi-structured, and unstructured data inherent in 3D cadastral systems. Cloud deployments are also investigated to evaluate the potential of distributed computing platforms for large-scale 3D cadastral data management. Emerging storage paradigms, including Data Warehouses, Data Lakes, and Lakehouse architectures, are further examined for their suitability within 3D cadastral systems. The evaluation phase formulates 11 normative 3D cadastral-specific metrics, including schema compliance, 3D spatial unit representation, geometry data types, coordinate reference systems, topological structures, spatial indexing and clustering, and performance tuning. The results demonstrate that relational databases combined with modern cloud architectures deliver robust, scalable, and high-performance solutions, while NoSQL document databases exhibit notable limitations. Furthermore, query optimisation strategies can significantly improve spatial database performance.