Comparison of infiltration models to describe infiltration characteristics of bioretention

Highlights

• The fitting accuracy of selected models was investigated under different experiment conditions.

• An optimal infiltration model was proposed for bioretention.

• Evaluating the optimal model with simulated runoff process for its flexibility and reliability.

Abstract

Bioretention is one of low-impact development measures, which widely used not only because it can reduce stormwater runoff total volume, decrease peak flow rate and delay peak flow time, but also can remove the runoff pollutants. Infiltration is an important hydrological process for bioretention to evaluate its runoff total volume reduction and pollutants removal. So, it is important to find an optimal infiltration model that can well describe the infiltration performance of bioretention. The Horton, Philip and Kostiakov infiltration models were selected to compare their accuracy when using for describe the infiltration characteristics of bioretention, and the errors between the different models simulate results and experiment results were assessed via the maximum absolute error (MAE), bias and coefficient of determination (R²). The experimental results showed that Horton model is fitting well and flexible under different experiment conditions, especially when the hydraulic head was 10 cm, with MAE of 0.50–0.81 cm/h, bias of 0.1–0.23 cm/h and R² of 0.98–0.99. R² of the Philip and Kostiakov models were all over than 0.87 at the initial infiltration period, but the model fitting accuracy decreased significantly with infiltration time elapse. Furthermore, the total runoff volume capture ratio and emptying time were advanced used to evaluate the flexibility of Horton model, and the Nash-Sutcliffe efficiency coefficients of them were over than 0.61 and 0.58, respectively. Therefore, the Horton model can be optimal selected to describe the infiltration process of bioretention and for its hydrological evaluation.

Introduction

With rapid urbanization, the interactions between humans and the surrounding environment have increased, and this has led to serious environmental issues such as flooding, ecological destruction, heat island effects and so on (Su et al., 2012, Ahiablame et al., 2012, Vu and Ranzi, 2017). Urbanization also results in impervious areas increased markedly, which destroy the natural geological features and hydrological properties of the underlying surface and cause numerous urban water environmental issues, such as increased flooding risk, decreased rainwater infiltration volume and reduced recharge capacity to groundwater (Dong et al., 2019, Leimgruber et al., 2018, Shao et al., 2016). Eshtawi et al. (2016) found that a 1% increase in impervious area would decrease the total infiltration volume from rainfall (range from 0 to −0.41 of Urban-Percolation Index) and increase the runoff volume (range from 0 to around 1 of Urban-Surface Index). Olivera and DeFee (2007) found that when the impervious area reached 10% of the total area, the annual runoff depth of the catchment increased by 146%. In order to alleviate these environmental problems caused by urbanization, low impact development have been proposed to reduce the runoff volume via many measures decentralized at the source of the catchment (Hamel et al., 2013, Dagenais et al., 2017), such as permeable pavement, bioretention, green roof, grass swale and so on. Among these measures, bioretention not only can removal pollutants from the runoff, but also has an ecological function and positive effects on the landscape, and it can reduce rainwater runoff volumes and peak flow rates and delay peak times (Xiang et al., 2019, Liu et al., 2016).

Infiltration is one of the important process that water entry into soil during hydrological recycling (Ren et al., 2020). The moisture content, media types, water suction head, temperature, and rainfall intensity were all important roles in influencing the infiltration rate (Ren and Santamarina, 2018). So it is necessary to investigate the infiltration performance of bioretention under different condition to evaluation its hydrological characteristics. Many studies have been implemented to find the factors that affect the infiltration process of bioretention. Le Coustumer et al. (2012) found that the infiltration rate of the bioretention gradually decreased with infiltration time elapse, and which were affected by the plants. Lee et al. (2016) found that the infiltration characteristics of bioretention are influenced by construction practices, which in turn influence the hydrological performance. Most research on the infiltration characteristics of bioretention adopt a soil infiltration monitoring method, which based on Darcy’s law (Xu et al., 2004, Li et al., 2018). Darcy (1856) found that the infiltration rate (Q) is positive correlation to the hydraulic gradient and the cross-sectional area (A), which is perpendicular to the infiltration direction, and is inversely proportional to the infiltration distant (L) when water passing through saturated sand. Darcy’s law assumes that the relationship of infiltration rate to hydraulic gradient in moisture saturated soil is linearly, so it is also called the linear infiltration law. The assumption of Darcy’s law include (1) constant temperature kept in the whole infiltration process, (2) constant laminar flow kept in the whole infiltration process, (3) the media porous is homogeneous and isotropic, (4) the relationship between infiltration rate and hydraulic gradient kept in linear (Zhang et al., 2018). However, bioretention has a multi-layer structure with inhomogeneous media, the infiltration process was not submitted to the laminar flow, and the infiltration rate changes over time. Furthermore, Fassman and Blackbourn (2010) found that if the Reynolds number increases, the relationship between the infiltration rate and the hydraulic gradient gradually deviates from a linear, and infiltration process cannot be described by Darcy’s law. Therefore, there are knowledge gaps that if the Darcy’s law is suitable for describing infiltration characteristics of bioretention or not.

Infiltration performance is important for bioretention to evaluate its runoff volume reduction and pollutants removal. So, it is important to found an optimal infiltration model to describe the infiltration characteristics for bioretention. Several infiltration models were selected which application condition was flexible and assimilate to infiltration condition of bioretention. The Kostiakov (Parhi et al., 2007) and Horton (Horton, 1941) models are empirical models, and neither has a physical meaning. The Green-Ampt (Green and Ampt, 1911) and Philip (Philip, 1957) models are physical models, and are often used to analyze the relationship between the infiltration rate and soil characteristics. They are all working well for different types of soil. So the empirical model (Horton and Kostiakov model) and physical model (Philip model) were selected to found an optimal model to describe the infiltration process of bioretention.

The objectives of this study were to (1) investigating the fitting accuracy of selected models under different hydraulic gradients and different types of bioretention to found an optimal model, (2) further analyzing the fitting accuracy of selected models under different infiltration times to found an optimal moldel (3) investigating the flexibility and reliability of the optimal model via R_v and T_p, which were compared between calculated results by optimal model and experimental results.