Back to Search
Start Over
Quantifying urban groundwater complexity: A high parameterization modelling approach.
- Source :
-
Journal of Hydrology . Jul2024, Vol. 638, pN.PAG-N.PAG. 1p. - Publication Year :
- 2024
-
Abstract
- • Groundwater modelling in urban settings is complicated by altered recharge and infrastructure interactions with groundwater. • Scripted based model development, iterative ensemble smother (IES) methods, and cloud computing, can facilitate representation of the complexity of urban groundwater systems using extremely highly parameterized approaches. • This study presents an applied consulting project where new methods were used to better represent urban groundwater in numerical models. Modelling of groundwater flow and transport in urban environments is complicated by changes in recharge due to urbanization, and interactions between groundwater, water supply networks and sewerage pipe infrastructure. This study presents an applied groundwater modelling project that attempts to capture the complexity and quantify the uncertainty of groundwater flow and mass transport in urban environments using recently developed methods. A python scripting approach was used for model development and preparation of a calibration constrained parameter uncertainty analysis using the iterative ensemble smoother method (IES). Water supply and sewer infrastructure locations were explicitly included in the numerical model as boundary conditions with adjustable parameters during the calibration. The model calibration, and predictive uncertainty analysis considered more than 100,000 independent parameters, which were applied at zones, pilot points and directly on the model grid to define model properties, boundary conditions, and initial conditions. Calibration observations included both water level and concentration measurements at 18 boreholes over the 12-year operational life of a remediation system. The IES method was applied to calibrate 120 alternative and equally plausible parameter realizations. Initial parameter values were produced from initial statistical distributions defined by parameter bounds and based on the conceptual hydrogeological model, literature review and spatial correlations of hydraulic properties observed in well tests. The Amazon EC2 cloud computing service was used to provide the parallel computing capacity necessary to complete the project within a reasonable timeframe. Predictions of long-term mass transport towards a creek after the termination of active remediation were made with 47 of the resulting IES calibrated parameter sets. The IES method provided a distribution of potential impacts and allowed decisions about the future of the current remediation system to be made based on a quantitative, statistical, risk-based analysis. The extremely high degree of model parameterization allowed the complexity and uncertainty of an urban groundwater systems to be represented using adjustable model parameters, which reduced the likelihood of structural model error and biased predictions. This study found that fully python scripted model development and calibration setup, combined with IES, successfully allowed for improved representation of urban groundwater systems in numerical models. The post calibration groundwater flow budget of the model showed that groundwater interactions with sewer and water supply infrastructure are of similar magnitude to distributed recharge, and the omission of representing infrastructure in urban groundwater models could result in misleading predictions. [ABSTRACT FROM AUTHOR]
Details
- Language :
- English
- ISSN :
- 00221694
- Volume :
- 638
- Database :
- Academic Search Index
- Journal :
- Journal of Hydrology
- Publication Type :
- Academic Journal
- Accession number :
- 178233115
- Full Text :
- https://doi.org/10.1016/j.jhydrol.2024.131416