Geyser eruption in urban drainage is a high-velocity discharge of an air–water mixture driven by intense air–water interactions. It is sudden, destructive and strongly non-linear, and is often associated with air entrapment, surcharge flows during intense rainfall and transient hydraulic processes, posing risks to drainage assets and surface infrastructure. A representative horizontal main pipe–vertical riser configuration is modelled using a volume of fluid (VOF) two-phase framework and a systematic numerical study is conducted of the inlet pressure head, dimensionless initial entrapped air index, Ia0, and the initial water-column height. Results identify a distinct eruption threshold: under the fixed geometry and boundary conditions considered in this study and within the tested parameter range, geyser eruption is triggered when Hin/H0 ≥ 1.33 andIa0 ≥ 3.20. As the free surface enters the upper riser, pronounced pressure spikes at PT4 and PT5 can serve as precursors. For a given inlet pressure head and the initial entrapped air condition, increasing the initial water-column height leads to higher eruption velocities, highlighting its control over eruption intensity.
Water shortage in arid and semi-arid regions has made groundwater the primary source of water supply. However, the performance of water wells gradually declines due to physical, chemical or biological clogging. Identifying wells with the highest rehabilitation potential is of significant practical and economic importance. This study proposes advanced environmental impact assessment methodologies to prioritise water well rehabilitation. Specifically, remote sensing-based (RSB) and observation-based (OB) assessment approaches are developed to determine high-priority well revival cases. The proposed RSB methodology selects wells based on land-subsidence values derived from the differential interferometric synthetic-aperture radar technique. These values are used as indicators to classify well rehabilitation potential, considering aquifer stratification and sand pumping characteristics. In the OB assessment stage, dynamic water level, well depth, submersible pump type and well videometry data are incorporated to evaluate well conditions. A theoretical model is introduced to predict dynamic water levels and identify wells exhibiting the most significant variations in the well performance index. The results show that the proposed model effectively captures the relationships among dimensionless parameters, explaining 91.2% of the total variance, while cross-validation confirms its robustness and applicability to new data sets. Furthermore, different well clogging types were identified, and appropriate rehabilitation strategies were recommended.
Accurate short-term streamflow forecasting is essential for mitigating urban flood risks and ensuring sustainable water resource management. This study presents a comparative analysis of advanced data-driven techniques for streamflow prediction in the Gandak River basin, India, evaluating their efficacy in handling rapid hydrological fluctuations. Three sophisticated machine learning models – random forest (RF), extreme gradient boosting (XGBoost) and long short-term memory (LSTM) networks – are employed to forecast streamflow across lead times ranging from 1 to 30 days. The findings reveal that all models exhibit strong predictive capabilities for 1-day lead times, achieving high coefficients of determination (R2): 0.976 (LSTM), 0.970 (RF) and 0.964 (XGBoost). However, predictive accuracy diminishes significantly as the forecasting horizon extends. By the 30-day lead time, the LSTM network consistently outperforms the tree-based models (R2 = 0.759 compared to RF’s 0.564 and XGBoost’s 0.569). LSTM’s superior performance for medium- to long-term forecasts is attributed to its advanced architecture, which effectively captures long-term dependencies and non-linear sequential patterns in hydrological time-series data. This study demonstrates the added value of deep learning over traditional ensemble methods for extended forecasting, offering actionable insights into optimising predictive frameworks for urban flood preparedness.
Four machine learning () techniques (artificial neural network with particle swarm optimisation (ANN-PSO), adaptive neuro-fuzzy inference system (ANFIS), multivariate adaptive regression splines (MARS) and M5 Tree) were used to model scour depth around bridge piers under clear-water scour () conditions. A total of 912 datasets collected under experimental and field conditions were compiled from published literature. The most influential input parameter combinations were identified using the gamma test. Among 37 combinations involving the pier width to flow depth ratio (b/y), approach velocity to critical velocity ratio (V/V-c), critical Froude number (Fr-c), pier width to median sediment size ratio (b/d(50)) and sediment gradation (sigma(g)), the optimal combination was selected based on minimum gamma and V-ratio values. The results indicate that the MARS model achieved superior performance, yielding a coefficient of determination (R-2) of 0.942 and a mean absolute percentage error () of 0.086, outperforming the ANN-PSO, ANFIS and M5 Tree models. The trained models were further validated using independent field datasets from Chemung, Honey Creek, Tanana and Clarks Fork Rivers and compared with ten existing scour prediction equations. The MARS and ANFIS models consistently showed higher accuracy, with R-2 > 0.90 and MAPE < 33%. The findings demonstrate the superiority of approaches over empirical models and identify key parameters governing CWS, supporting reliable bridge scour assessment.
The focus of this research was on prioritising strategies to mitigate the adverse effects associated with the construction of urban wastewater treatment plants (WWTPs), specifically during the critical study phase. Topsis (technique for order preference by similarity to ideal solution) - a multi-criteria decision making () methodology - was employed. A significant contribution of this work is the integration of neutrosophic concepts with to effectively manage imprecise information, uncertainties stemming from human consciousness and potential unforeseen changes crucial for future planning. The first part of the study was an environmental impact assessment of the wastewater treatment system, establishing nine criteria categorised into natural resources, human health impacts and economic conditions. To address negative consequences during construction and operation of the WWTP, three strategic alternatives were proposed: advanced sludge management, green space development and stringent wastewater leakage control. Neutrosophic Topsis analysis highlighted sludge management as the most pivotal strategy for minimising detrimental effects in the foreseeable future. Furthermore, an evaluation of criteria importance identified specific factors with the most significant influence: control of unpleasant odours, mitigation of pathogen dispersal and the economic valorisation of both treated wastewater and derived sludge. This work offers a valuable decision making tool for sustainable urban wastewater management.
