Evaluating and optimizing the design and spatial allocation of green infrastructure in shallow groundwater environments

Abstract

Green infrastructure (GI) represent semi-engineered stormwater management practices that have been widely implemented to tackle urban stormwater problems, e.g., urban flooding, non-point source pollution. However, it is challenging to implement GI in shallow groundwater conditions. The impact of shallow groundwater, and the suitable engineering approaches to implement GI in these areas remain to be studied. First, although shallow groundwater is known to be a challenge of implementing GI, its interaction with GI has not been quantified. This thesis thus performed statistical analyses to evaluate the interaction between porous pavement and shallow groundwater at various temporal scales. The results showed that rainfall remained the key driver of underdrain flow. The impact of groundwater on underdrain flow outweighed that of rainfall when groundwater table was shallower than the underdrain, particularly at finer temporal scales. Second, although some design recommendations have been given to implement GI in groundwater table, they mostly appear to be empirically based, and their feasibility and effectiveness have not been thoroughly studied. This thesis thus built numerical models of individual bioretention cell and porous pavement and proposed design recommendations through investigating their hydrologic performance in shallow groundwater. The thesis showed that it is generally feasible to implement well-designed GI in shallow groundwater areas. Lower- and higher- permeable media soils are recommended for bioretention cells when groundwater contamination and runoff control are of concern, respectively. Building a thinner pavement, installing an underdrain, and installing an impermeable liner showed some benefits and drawbacks in mimicking natural hydrologic cycle and retaining the performance of porous pavement. Third, although there are many GI-enabled models, they all have limitations in simulating GI in shallow groundwater at catchment scale. This thesis thus developed a modified SWMM (named SWMM-LID-GW) which considered groundwater table condition into the calculation of GI-related hydrologic processes. Furthermore, this thesis developed a coupled surface-subsurface hydrological model (named SWMM-MODFLOW) which realized the two-way interaction between GI and groundwater at catchment scale. The thesis showed that both models are applicable; SWMM-LID-GW performed significantly better than the current SWMM particularly for shallower groundwater table conditions. Fourth, when planning GI spatially at catchment scale, maximizing the reductions in runoff and pollution are normally the main objectives. Groundwater dynamics, however, are seldom considered. Thus, this thesis performed scenario-based simulations and algorithm-based optimizations to quantify the impact of spatial allocation of bioretention cells on groundwater table dynamics using SWMM-MODFLOW. It also provided recommendations on the optimal spatial allocation of bioretention cells when their interaction with shallow groundwater is considered. The thesis demonstrated the importance of considering groundwater table condition in GI planning. Apart from the areal coverage of GI, their aggregation level and locations are also vital when groundwater management is of concern. In conclusion, this thesis demonstrated that the impact of shallow groundwater on the hydrologic performance of GI needs to be carefully considered in the design and planning of GI. A better design and planning strategy of GI can mitigate the impact of shallow groundwater and facilitate the GI implementation in shallow groundwater environments.