Research papers
Assessing the impact of spatial allocation of bioretention cells on shallow groundwater – An integrated surface-subsurface catchment-scale analysis with SWMM-MODFLOW

https://doi.org/10.1016/j.jhydrol.2020.124910Get rights and content

Highlights

  • SWMM-MODFLOW is developed to study bioretention cells’ (BCs) impact on groundwater.

  • BC may not affect the spatial uniformity of groundwater levels if properly allocated.

  • More-distributed BCs produce lower peak but higher time-averaged groundwater rises.

  • Allocating BCs upstream/downstream is recommended when groundwater is deep/shallow.

  • BC planning should consider both runoff and shallow groundwater dynamics.

Abstract

Well-designed and implemented green infrastructure (GI) can help to recover the natural hydrologic regime of urban areas. A large-scale GI planning requires a good understanding of the impact of GI spatial allocation on surface-subsurface hydrologic dynamics. This study, firstly, developed a coupled surface-subsurface hydrological model (SWMM-MODFLOW) that can simulate fine-temporal-scale two-way interactions between GI and groundwater at catchment scale. The model was calibrated and validated using the monitoring data at one urban catchment within Kitsap County, WA, US. Based on the validated model, a series of hypothetical simulations was then performed to evaluate how spatial allocation of bioretention cells (BCs), one type of GI, influences and correlates with surface runoff and groundwater table dynamics. The spatial allocation was represented by implementation ratio (i.e., area), aggregation level (i.e., density) and location of BCs. The hydrologic dynamics were quantified by peak and volume reductions of surface runoff, as well as groundwater table rise and standard deviation of groundwater levels. A small number of BCs can raise groundwater table locally and regionally. However, it may not affect the spatial uniformity of groundwater levels (represented as the standard deviation of groundwater levels) if being properly allocated. Although the impact of aggregation level of BCs was relatively low compared to the implementation ratio and relative location of BCs, more-distributed BCs resulted in lower peak groundwater table rises but higher temporally-averaged groundwater table rises. Allocating BCs upstream resulted in higher groundwater table rises regionally, which is recommended for areas of deeper groundwater tables. While, allocating BCs downstream is more recommended for areas of shallower groundwater tables. BCs of greater surface runoff control efficiencies lead to higher groundwater table rises, which highlights the importance of considering the tradeoff between surface runoff control and groundwater protection in GI planning.

Introduction

Excessive urbanization has significantly deteriorated the natural hydrological, ecological and biological regimes. There is now a general consensus that a more sustainable and environmental-friendly development approach is needed (Song, 2005). Green infrastructure (GI) is therefore proposed under this cognitive revolution (Brown et al., 2009). However, different people may define “GI” differently. For people concerned about hydrology and stormwater management (e.g., hydrologists, urban drainage engineers), GI is analogous to some other terms, e.g., low impact development (LID), sustainable urban drainage system (SuDS), and water sensitive urban design (WSUD), which represent a group of semi-natural spatially-distributed stormwater management practices (Potter, 2006, Young et al., 2014, Fletcher et al., 2015). Compared to traditional drainage systems, they possess more diverse functionalities which include collecting, storing, infiltrating rainfall runoff and recovering natural hydrological cycle (Chui et al., 2016). Representative practices include bioretention cells, porous pavements, green roofs, etc. While, for other people (e.g., landscape designers, urban planners), GI can include forests or other green spaces that could be better in providing other environmental benefits such as urban heat island mitigation and biodiversity increase (EC, 2013, Zhang and Chui, 2019). The “GI” in this study follows the first definition having a main focus on hydrology and stormwater management.

Among all the benefits that GI can provide (e.g., peak runoff and non-point source pollution control), groundwater recharge is the one that attracts relatively little attention (Jefferson et al., 2017, Sohn et al., 2019). One possible reason is that recharging groundwater using GI normally comes with many challenges, which are prominent in shallow groundwater areas particularly. First, a groundwater mound can form when the recharge rate exceeds the dissipation rate of groundwater. This may slow down or inhibit surface infiltration and increase the risk of groundwater contamination due to a shorter traveling distance and more carried pollutants (Fischer et al., 2003, Datry et al., 2004, Göbel et al., 2004, Endreny and Collins, 2009, Machusick et al., 2011, Stewart et al., 2017, Zhang and Chui, 2017, Zhang and Chui, 2018a). However, recharging groundwater using GI can increase the baseflow, recover the hydrological cycle and help maintain urban water supplies (Newcomer et al., 2014, Bhaskar et al., 2016, Bhaskar et al., 2018, Bradshaw and Luthy, 2017). It should therefore be promoted if given appropriate conditions, e.g., suitable subsurface soil properties and relatively deep groundwater table (Trinh and Chui, 2013, Chui and Trinh, 2016).

For the reasons aforementioned, the objectives and constraints related to groundwater recharge should be thoroughly considered in GI planning. However, maximizing the control in surface runoff often remains the dominant objective in GI implementation. As reviewed by Zhang and Chui (2018b), many studies considered peak and volume control of surface runoff (e.g., Perez-Pedini et al., 2005, Damodaram and Zechman, 2012, Sebti et al., 2016, Giacomoni and Joseph, 2017, Lim and Welty, 2017, Voter and Loheide, 2018). While many others considered pollution mitigation of surface runoff (e.g., Maringanti et al., 2009, Rodriguez et al., 2011, Chiang et al., 2014, Chen et al., 2015, Chen et al., 2016), and some examined the two aspects together (Lee et al., 2012, Liu et al., 2016a, Liu et al., 2016b, Mao et al., 2017, Xu et al., 2018). Regarding works relating GI and groundwater, some studies assessed the response of shallow groundwater to GI (Endreny and Collins, 2009, Trinh and Chui, 2013, Chui and Trinh, 2016, Zheng et al., 2018). Also, some proposed recommendations about the suitable distance between GI and groundwater table (Locatelli et al., 2015, Zhang and Chui, 2017, Muñoz-Carpena et al., 2018, Lauvernet and Muñoz-Carpena, 2018). However, to the best of the authors’ knowledge, the impact of GI spatial allocation on shallow groundwater table dynamics remains to be evaluated.

The spatial allocation of GI is hypothesized to affect the groundwater table dynamics in a number of aspects. Based on the study of Zhang and Chui (2018b), the spatial allocation of GI can be mainly represented by implementation ratio (i.e., the area), aggregation level (i.e., density), and location of GI practices. First, the implementation ratio of GI is one of the major factors because it determines the amount of rainfall that can be infiltrated. With a higher implementation ratio, more water can be recharged, and the groundwater table should rise higher. Second, the aggregation level of GI is also influential, because more-aggregated GI practices can infiltrate more water locally and result in localized groundwater mound as reported by Endreny and Collins (2009). Third, the location of GI also matters, as the land use and geologic conditions (e.g., hydraulic properties of in-situ soil, groundwater table depth) can be very different at different locations. More specifically, in areas of higher imperviousness, more permeable soils, and shallower groundwater tables, GI can affect the groundwater table dynamics more due to more surface runoff generated, higher infiltration and recharge rate, vice versa. The concept of variable source area well explains the impacts of these factors (Miles and Band, 2015, Lim, 2016). Some studies determined the spatial allocation of GI only based on these land use and geologic factors by using spatial analysis tools without hydrological analysis (Martin-Mikle et al., 2015, Johnson and Sample, 2017).

Although there are many different numerical models that can simulate the hydrological processes of GI, they all have their limitations in simulating GI in shallow groundwater environment. As reviewed by Zhang et al. (2018), variably-saturated porous media software, e.g. COMSOL Multiphysics, VS2D, Hydrus 1D/2D/3D, have been used in some cases (He and Davis, 2010, Stewart et al., 2017, Zhang and Chui, 2017). However, they generally cannot handle or are not suitable for catchment-scale studies because they simplify or cannot simulate rainfall-runoff generation and surface runoff routing. In addition, some surface-subsurface hydrological models (e.g., MODHMS, MIKE-SHE, and VELMA) can better simulate rainfall-runoff processes and are more widely used at the catchment scale (Barron et al., 2013, Trinh and Chui, 2013, Locatelli et al., 2017, Hoghooghi et al., 2018). However, they mostly operate in relatively coarse temporal and spatial resolutions, which are beyond the normal scale of each individual GI practice. Thus, some time- and space-sensitive hydrological processes important to GI are simplified. Finally, many of them are commercial and non-open source software, which make them harder to be improved or integrated with other tools (e.g., data analysis and optimization tools). And both types of models cannot simulate urban hydraulics (e.g., storm sewer systems), which limit their usages in urban areas.

As an urban hydrologic-hydraulic model, SWMM has been widely adopted to simulate GI, including assessing the hydrological and water quality treatment performance of GI (Qin et al., 2013, Palla and Gnecco, 2015, Chui et al., 2016, Jayasooriya et al., 2016, Avellaneda et al., 2017, Kong et al., 2017;), as well as evaluating the optimal designs and allocations of GI (Elliott et al., 2009, Lucas and Sample, 2015, Giacomoni and Joseph, 2017, Macro et al., 2019, Yang and Chui, 2018a, Yang and Chui, 2018b, Zischg et al., 2018). However, SWMM is not as capable in simulating the subsurface hydrological performance of GI. First, it highly simplifies the simulation of unsaturated and saturated flows by assuming a linearized soil water retention curve. Second, it neglects the impact of groundwater on the hydrological processes of GI (e.g., exfiltration, percolation, underdrain flow) (Lee et al., 2018, Zhang et al., 2018). To partially overcome these deficiencies, Zhang et al. (2018) had improved SWMM by creating an interface to incorporate groundwater levels into the simulation of GI. And it had been tested that the modified SWMM is appropriate for simulating the performance of GI in shallow groundwater environments. However, it cannot simulate groundwater dynamics but requires the direct input of groundwater levels which greatly hinders its application.

As a follow-up study of Zhang et al. (2018), this study integrated the modified SWMM, named SWMM-LID-GW, with MODFLOW to develop a loosely-coupled surface-subsurface hydrological model (SWMM-MODFLOW). It is loosely coupled because the structures of two models were kept unchanged, and the two models were integrated through external file input & output without internal function calls. The coupling approach utilized in this study is similar to that of Zhang et al. (2018). However, the groundwater dynamics were simulated instead of being inputted, and the two-way interaction between GI and groundwater was realized. The model was calibrated and validated using the monitoring data in one urban catchment at Silverdale, Washington State of the United States. Then, using bioretention cell (BC) that allows exfiltration as a representative GI, a series of hypothetical scenarios of different spatial allocation patterns of BCs was simulated, which covered different implementation ratios, aggregation levels and locations of BCs, within the same catchment. Furthermore, the influence of spatial allocation of BCs on surface runoff and groundwater table dynamics was evaluated. Finally, the correlations between surface runoff and groundwater table dynamics were examined. It should be noted that this study focused more on groundwater rather than surface runoff dynamics because they were less studied and understood.

Section snippets

Modeling framework of SWMM-MODFLOW

A two-way coupled surface-subsurface hydrological model, SWMM-MODFLOW, was developed and utilized. It is a loosely-coupled model linking SWMM and MODFLOW and its structure is shown in Fig. 1. Retaining the main structures of SWMM and MODFLOW, the coupling was performed through file input and output without internal function calls between the source codes of two models. More specifically, surface infiltration rate in non-GI pervious areas (or exfiltration rate at the bottom of GI) is sent from

Model calibration and validation

Fig. 6 shows the time series of different datasets during both calibration and validation periods. The light grey and dark grey sections in Figs. 6a-6f correspond to the calibration and validation periods, respectively. One event on 23 December 2011 is shown specifically in Figs. 6g-6l. This event is considered representative given its medium rainfall intensity (i.e., 12 mm/h), runoff amount (i.e., 23 mm), and groundwater table fluctuation (i.e., 0.3 m) during the period, which can be observed

Concluding remarks

A coupled surface-subsurface hydrological model, SWMM-MODFLOW, was developed to evaluate the surface runoff and groundwater table dynamics of green infrastructure of different spatial allocations at catchment scale. The model was calibrated and validated using the monitoring data at one urban catchment at Kitsap County, Washington State of the United States.

Using bioretention cells as the representative green infrastructure, a series of hypothetical simulations was performed. The influence of

CRediT authorship contribution statement

Kun Zhang: Conceptualization, Methodology, Software, Formal analysis, Investigation, Resources, Data curation, Writing - original draft, Visualization. Ting Fong May Chui: Validation, Writing - review & editing, Supervision, Project administration, Funding acquisition.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was funded by the Seed Funding Programme for Basic Research of The University of Hong Kong (Project code: 201611159011). The authors are grateful to the Stormwater Division of the Kitsap County Department of Public Works (Washington, U.S.A.) for providing the monitoring data that made this study possible.

References (86)

  • F. Kong et al.

    Modeling stormwater management at the city district level in response to changes in land use and low impact development

    Environ. Model. Softw.

    (2017)
  • J.G. Lee et al.

    A watershed-scale design optimization model for stormwater best management practices

    Environ. Model. & Softw.

    (2012)
  • Y. Liu et al.

    Optimal selection and placement of BMPs and LID practices with a rainfall-runoff model

    Environ. Model. Softw.

    (2016)
  • Y. Liu et al.

    Optimal selection and placement of green infrastructure to reduce impacts of land use change and climate change on hydrology and water quality: an application to the Trail Creek Watershed, Indiana

    Sci. Total Environ.

    (2016)
  • L. Locatelli et al.

    Hydrologic impact of urbanization with extensive stormwater infiltration

    J. Hydrol.

    (2017)
  • L. Locatelli et al.

    Determining the extent of groundwater interference on the performance of infiltration trenches

    J. Hydrol.

    (2015)
  • W.C. Lucas et al.

    Reducing combined sewer overflows by using outlet controls for green stormwater infrastructure: case study in Richmond, Virgina

    J. Hydrol.

    (2015)
  • K. Macro et al.

    OSTRICH-SWMM: a new multi-objective optimization tool for green infrastructure planning with SWMM

    Environ. Model. & Softw.

    (2019)
  • X. Mao et al.

    Assessing the ecological benefits of aggregate LID-BMPs through modelling

    Ecol. Model.

    (2017)
  • C.J. Martin-Mikle et al.

    Identifying priority sites for low impact development (LID) in a mixed-use watershed

    Landsc. Urban Plan.

    (2015)
  • A. Palla et al.

    Hydrologic modeling of low impact development systems at the urban catchment scale

    J. Hydrol.

    (2015)
  • H. Qin et al.

    The effects of low impact development on urban flooding under different rainfall characteristics

    J. Environ. Manage.

    (2013)
  • A. Ritter et al.

    Performance evaluation of hydrological models: statistical significance for reducing subjectivity in goodness-of-fit assessments

    J. Hydrol.

    (2013)
  • W. Sohn et al.

    The influence of climate on the effectiveness of low impact development: a systematic review

    J. Environ. Manage.

    (2019)
  • T. Xu et al.

    Marginal-cost-based greedy strategy (MCGS): fast and reliable optimization of low impact development (LID) layout

    Sci. Total Environ.

    (2018)
  • Y. Yang et al.

    Optimizing surface and contributing areas of bioretention cells for stormwater runoff quality and quantity management

    J. Environ. Manage.

    (2018)
  • R. Young et al.

    A comprehensive typology for mainstreaming urban green infrastructure

    J. Hydrol.

    (2014)
  • K. Zhang et al.

    A comprehensive review of spatial allocation of LID-BMP-GI practices: strategies and optimization tools

    Sci. Total Environ.

    (2018)
  • K. Zhang et al.

    Linking hydrological and bioecological benefits of green infrastructures across spatial scales–a literature review

    Sci. Total Environ.

    (2019)
  • K. Zhang et al.

    Simulating the hydrological performance of low impact development in shallow groundwater via a modified SWMM

    J. Hydrol.

    (2018)
  • P.M. Avellaneda et al.

    Simulation of the cumulative hydrological response to green infrastructure

    Water Resour. Res.

    (2017)
  • A.S. Bhaskar et al.

    Urban base flow with low impact development

    Hydrol. Process.

    (2016)
  • A.S. Bhaskar et al.

    Groundwater recharge amidst focused stormwater infiltration

    Hydrol. Process.

    (2018)
  • J.L. Bradshaw et al.

    Modeling and optimization of recycled water systems to augment urban groundwater recharge through underutilized stormwater spreading basins

    Environ. Sci. & Technol.

    (2017)
  • R.R. Brown et al.

    Urban water management in cities: historical, current and future regimes

    Water Sci. Technol.

    (2009)
  • Y. Cai et al.

    Spatial variations of river–groundwater interactions from upstream mountain to midstream oasis and downstream desert in Heihe River basin, China

    Hydrol. Res.

    (2016)
  • L.C. Chiang et al.

    Comparing the selection and placement of best management practices in improving water quality using a multiobjective optimization and targeting method

    Int. J. Environ. Res. Public Health

    (2014)
  • T.F.M. Chui et al.

    Modelling infiltration enhancement in a tropical urban catchment for improved stormwater management

    Hydrol. Process.

    (2016)
  • C. Damodaram et al.

    Simulation-optimization approach to design low impact development for managing peak flow alterations in urbanizing watersheds

    J. Water Resour. Plan. Manage.

    (2012)
  • S. Dhalla C. Zimmer Dhalla, S. and Zimmer, C., 2010. Low Impact Development Stormwater Management Planning and Design...
  • A.R. Dussaillant et al.

    Richards equation model of a rain garden

    J. Hydrol. Eng.

    (2004)
  • EC–European Commission, Green Infrastructure (GI)—Enhancing Europe's Natural Capital. European Commission, 2013,...
  • A.H. Elliott et al.

    Effect of aggregation of on-site storm-water control devices in an urban catchment model

    J. Hydrol. Eng.

    (2009)
  • Cited by (26)

    View all citing articles on Scopus
    View full text