Filling data gaps in urban drainage networks: An automated graph theory framework for data collection and reconstruction

Abstract

Missing network data, such as sewer diameters and slopes, pose significant challenges for modeling urban drainage networks (UDNs). However, efficient and fully automated approaches for addressing such data gaps remain limited in the literature. This study proposes an integrated framework for reconstructing key network attributes and supporting informed data collection. The method systematically infers missing sewer diameter and slope (invert elevation) information while identifying optimal locations for their targeted collection (e.g., in critical regions with high data gaps). Built on graph theory, the framework leverages topological features of UDNs, such as connectivity and hierarchical patterns between sewers with available information, while incorporating the hydrodynamic model assembly to ensure the basic functional validity of the reconstructed network. To validate the framework, data gaps were artificially created in increments from 10 % to 90 % for two real-world stormwater networks with complete datasets, resulting in 1800 networks with incomplete data. The reconstructed networks were evaluated against the original networks by comparing physical (diameters, invert elevations) and hydrodynamic attributes (flow rates, flood volumes, flooded nodes). Results demonstrated high accuracy in data reconstruction, especially for data gaps up to 70 %. For higher data gaps, the data collection framework significantly improved reconstruction accuracy by identifying optimal locations for targeted data collection. In addition, the proposed approach was applied to a large-scale separate wastewater network, where it outperformed existing methods in the literature in terms of reconstruction accuracy. The proposed framework offers a fast and practical solution for retrieving information, ensuring a reliable basis for any hydrodynamic analysis.

Topics & Concepts