Quantification of infiltration processes in urban areas by accounting for spatial parameter variability

Abstract

Background, Aims, and Scope

As a consequence of human living and activity, water infiltration to the urban subsurface occurs from a variety of different sources, like precipitation, irrigation, leaking pipes and sewers, septic tanks and rainwater infiltration ponds. This infiltration is strongly related with quality issues of the infiltrated water and further impact on groundwater quality. In order to set up an integrated urban water balance it becomes essential to estimate the infiltration processes, i.e. water flow and solute transport, from these different infiltration sources and to take into account the large spatial variability of sediment properties, the geometric settings of these sources and the groundwater table. For that purpose, the development of simple, physically-based quantification approaches is required in order to establish an efficiently working prediction and risk analysis tool within the framework of an integrated urban water management system. The scope of the presented work was to demonstrate the applicability of the developed approaches at urban scale.

Methods

Since a detailed, three-dimensional, numerical quantification of the infiltration processes within the entire urban area is not possible, the individual sources were considered as independent within the EU AISUWRS project. Different models were developed for balancing infiltration from areal and point sources with respect to the related flow pattern. The analytical model UL_FLOW, based on one-dimensional, steady state analytical solutions, allows the estimation of conservative tracer residence times in layered sediments under varying infiltration rates. The numerical model WSTM, based on a three-dimensional random walk approach, calculates water and solute transport from pipe leaks. Additionally, the sources were classified in accordance to the spatial distribution of the parameters determining the infiltration processes.

Results

UL_FLOW was applied to data sets from the city of Rastatt within a case study of the AISUWRS project. Each neighbourhood of water balance computation by the Urban Volume and Quality Model (UVQ) was defined as an areal infiltration source with unique parameter values for sediment depth, profile and properties, as well as infiltration rate time series. Groundwater recharge and residence time series were computed for each neighbourhood. Relevant statistical parameters obtained by time series analyses from those time series could be mapped by GIS. Point infiltration, particularly from sewers, was classified due to the sediment parameters and the distance to the groundwater table at each source location in order to reduce computational efforts. WSTM computations provided time series of groundwater recharge and tracer breakthrough for some specific cases.

Discussion

The analytical model UL_FLOW provides fast and efficient computation of groundwater recharge and residence times accounting for storage effects within the unsaturated zone of urban areas. The reliability of this model has been shown by cross validation with HYDRUS1D. Because of the high computational effort, WSTM could provide only short-term simulations for some specific parameter sets for which residence time estimates could be derived.

Conclusions

UL_FLOW provides an analytical modelling tool for balancing one-dimensional areal infiltration and estimating residence times under varying conditions including spatial parameter variability. These balances could be used for assessing the impact of those infiltration sources on groundwater quality. The tracer breakthrough from point infiltration sources computed by WSTM could also be used for such kinds of assessment. The larger spatial parameter variability associated with these sources could be handled by classification in GIS environments.

Recommendations and Perspectives

Similar to the areal sources, a simple balance approach for point sources based on analytical solutions needs to be developed for estimating residence times in order to avoid large computational efforts. Such a model would complete the balancing of all kinds of infiltration sources in urban areas efficiently. Since the approaches are based on the balance of the physical processes, they have a large predictive capability and could be included into an integrated urban water balance and management system. The mapping of the statistical values of the residence times provides a tool to compare parts of the urban areas and to visualize differences between urban water management scenarios.