This study compares Artificial Neural Networks (ANN) and Elastic Net regression for predicting surface runoff in urban stormwater catchments. Both models were trained on a data set derived from the Stormwater Management Model that included parameters such as imperviousness, flow path width, slope, Manning coefficients, and depression storage. ANN exhibited greater predictive accuracy and stability, especially when modeling nonlinear hydrologic interactions, while Elastic Net offered faster inference and clearer interpretability, but showed reduced accuracy in low-flow conditions. Validation on real-world data revealed the sensitivity of the models to scenarios not fully represented during training. Despite higher computational demands, the ANN proved more adaptable, while the more resource-efficient Elastic Net remains suitable for time-critical or large-scale applications. These findings provide practical insights for urban water resource management, indicating when each approach can be most effectively used in flood risk assessment and stormwater infrastructure planning.

Urban stormwater systems face escalating challenges due to climate change, aging infrastructure, and increasing impervious surfaces, necessitating holistic frameworks that integrate hydraulic, structural, and operational factors. This study proposes a fuzzy logic-based controller to evaluate the performance of stormwater drainage systems through three linguistic variables: Hydraulic Performance, Technical Condition, and Operational Condition. The model synthesizes expert knowledge into 125 inference rules, enabling a unified assessment of system reliability. Validated against empirical datasets from European and Chinese drainage networks, the framework demonstrates robust performance across diverse geospatial and operational contexts. Unlike traditional deterministic methods or single-criterion fuzzy systems, the controller addresses interdependencies between hydraulic efficiency, material degradation and external stressors. By translating multi-dimensional uncertainties into actionable maintenance priorities—from “Immediate Replacement” to “No Action Required”—the model enhances decision-making for utilities balancing flood resilience and infrastructure longevity.

Parameterization of SWMM subcatchments is labor-intensive and a major source of model uncertainty. This study presents the Rapid Catchment Generator (RCG), a fuzzy logic framework that derives hydraulic width, average slope, and impervious fraction from three easily accessible descriptors—area, landform type, and land cover type—and inserts them directly into SWMM input files. A sensitivity analysis of 116,640 synthetic simulations confirmed that width, slope, and imperviousness are the dominant controls on runoff and infiltration. Their relationships are encoded in triangular membership functions covering nine geomorphic classes and twelve imperviousness classes, linked through expert-calibrated Mamdani rules. Validation on a calibrated 37-subcatchment, 10-hectare urban basin in Wrocław, Poland, showed Mean Absolute Percentage Errors of 15.9–16.0% for total runoff, 19% for infiltration, and 29–37% for peak flow, while preserving hydrograph shape. RCG thus reduces model setup time and provides a transparent, reproducible starting point for rapid scenario screening and subsequent fine-scale calibration.

Liquid Time-Constant networks (LNNs), recurrent models with adaptive, input-dependent time constants, were evaluated for water level prediction in an urban drainage system. The architecture was assessed on a ten-year measurement dataset from the Bellinge catchment and benchmarked against gated recurrent unit (GRU), long short-term memory (LSTM), temporal convolutional network (TCN), and multilayer perceptron (MLP) baselines. The best-performing LNN variant achieved the highest mean predictive accuracy in the benchmark (Nash–Sutcliffe efficiency, NSE = 0.849), with particularly accurate representation of the continuous response of the gravity-driven part of the network without substantial degradation during flash-flood events. The analysis showed that the adaptive time constants help distinguish system-wide hydraulic processes from local control actions while providing interpretable diagnostics of the model’s internal response scale. The principal challenge for the LNN was the intermittent operation of the pumping station; isolating the pump pathway in the dual-branch DB-LNN variant mitigated this performance degradation while preserving predictive accuracy at the remaining sensors. The ablation analysis indicated that architectural separation of the processing pathways reduced interference between the continuous hydraulic dynamics and the local switching dynamics of the threshold-controlled facility.