Physicochemical conditions and properties of particles in urban runoff and rivers: Implications for runoff pollution

Chemosphere 173 (2017) 318e325

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Physicochemical conditions and properties of particles in urban runoff and rivers: Implications for runoff pollution Qian Wang a, Qionghua Zhang a, b, *, Yaketon Wu a, Xiaochang C. Wang a, ** a b

Key Lab of Northwest Water Resource, Environment and Ecology, MOE, Xi’an University of Architecture and Technology, Xi’an 710055, China International Science & Technology Cooperation Center for Urban Alternative Water Resources Development, Xi’an 710055, China

h i g h l i g h t s
The physicochemical conditions varied greatly from rainwater to runoff, and to rivers.
Higher proportion of nano-scale particles was in runoff than in rivers.
The ratio of turbidity and TSS indicated the size and settleability of particles.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 28 June 2016 Received in revised form 4 January 2017 Accepted 12 January 2017 Available online 16 January 2017

In this study, to gain an improved understanding of the fate and fractionation of particle-bound pollutants, we evaluated the physicochemical conditions and the properties of particles in rainwater, urban runoff, and rivers of Yixing, a city with a large drainage density in the Taihu Lake Basin, China. Road runoff and river samples were collected during the wet and dry seasons in 2015 and 2016. There were significant differences between the physicochemical conditions (pH, oxidation-reduction potential (ORP), and electroconductivity (EC)) of rainwater, runoff, and rivers. The lowest pH and highest ORP values of rainwater provide the optimal conditions for leaching of particle-bound pollutants such as heavy metals. The differences in the physicochemical conditions of the runoff and rivers may contribute to the redistribution of pollutants between particulate and dissolved phases after runoff is discharged into waterways. Runoff and river particles were mainly composed of silt and clay (<63 mm, 88.3%e90.7%), and runoff particles contained a higher proportion of nano-scale particles (<1 mm) but a lower proportion of submicron-scale particles (1e16 mm) than rivers. The ratio of turbidity to TSS increased with the proportion of fine particles and was associated with the accumulation of pollutants and settling ability of particles, which shows that it can be used as an index when monitoring runoff pollution. © 2017 Elsevier Ltd. All rights reserved.

Handling Editor: X. Cao Keywords: Physicochemical conditions Particle characteristics Fractionation Urban runoff Urban river

1. Introduction Runoff volumes, runoff coefficients, and pollutant loads have tended to increase in recent decades, as a result of the increases in impervious surfaces caused by rapid rates of urbanization worldwide (Sun et al., 2015). Increases in these indexes exacerbate the risk of urban floods. Therefore, drainage systems are generally designed to ensure rapid discharge of runoff into adjacent

* Corresponding author. Key Lab of Northwest Water Resource, Environment and Ecology, MOE, Xi’an University of Architecture and Technology, Xi’an 710055, China. ** Corresponding author. E-mail addresses: [email protected] (Q. Zhang), [email protected] (X.C. Wang). http://dx.doi.org/10.1016/j.chemosphere.2017.01.066 0045-6535/© 2017 Elsevier Ltd. All rights reserved.

waterways. It has been reported that more than 90% of runoff is discharged directly into rivers in districts that have implemented rain and sewage diversion. This means that the increasing number of non-point source pollutants are more widely distributed, and are difficult to control with centralized measures (Loperfido et al., 2014). Non-point source pollution, recognized as a pertinacious illness of urban rivers because of its ongoing contribution to their pollution, has received considerable attention in previous studies (Reidsma et al., 2012). The concentrations, loads, and first flush effects of runoff pollutants have been investigated, and first flushes of pollutants with high concentrations and loads have also been observed during rainfall events (Kim and Sansalone, 2008; Chow et al., 2013; Zhang et al., 2013; Chow and Yusop, 2014; Gasperi et al., 2014). The processes that control the build-up and wash-off

Q. Wang et al. / Chemosphere 173 (2017) 318e325

of road-deposited sediment (RDS) have been explored to investigate the fate and fractionation of runoff pollutants, including their accumulation, mobility, and fractionation (Sansalone and Ying, 2008; Joshi and Balasubramanian, 2010; Yuen et al., 2012). Studies have shown that particle size plays an important role in the accumulation and mobility of pollutants (Gunawardana et al., 2012; Zhao and Li, 2013; Li et al., 2015). Also, pollutants tend to accumulate in fine particles that can be transported easily by rainfall (Gunawardana et al., 2015; Zhao et al., 2016). Other studies have reported that the fractionation of pollutants between the particulate and dissolved phases is determined by their binding states with particles and the physicochemical conditions (pH, oxidationreduction potential (ORP) and electroconductivity (EC)) in the water environment (Morselli et al., 2003; Duong and Lee, 2009; Zhang et al., 2016). For example, water environments with low pH, low ORP, and high ORP favor the release of heavy metals (HMs) associated with carbonates, Fe and Mn oxides, and organic matter, respectively (Sutherland et al., 2012; Maniquiz-Redillas and Kim, 2014). These studies provide valuable information about runoff pollution and its control. However, little is known about the fate and fractionation of pollutants contained in runoff once they are discharged into waterways. The fate and fractionation of runoff pollutants are closely related to their potential impacts on waterways; dissolved pollutants are directly bioavailable, while the sediments may clog waterway beds, smother biota, damage the respiratory systems of organisms, attenuate light, and act as vectors of hydrophobic pollutants (Helmreich et al., 2010; Zuo et al., 2012; McKee and Gilbreath, 2015). Therefore, it is important to analyze the physicochemical conditions and particle characteristics of the main receptors of RDS during runoff events, namely rainwater, runoff, and waterways, to evaluate the fate and fractionation of runoff pollutants. However, previous studies have tended to examine the particle properties and fractionation of particle-bound pollutants in runoff or rivers separately, and have rarely considered the physicochemical conditions. Consequently, we have little appreciation of whether or how the properties of particles in, and physicochemical conditions of, rainwater, runoff, and rivers differ from each other, nor do we have a good understanding of the potential fate of particle-bound runoff pollutants once they are discharged into waterways. A clearer understanding of these processes will facilitate improved management of runoff pollution. The objectives of this study therefore were to 1) detect the differences in the physicochemical conditions and particle characteristics of rainwater, urban runoff, and rivers, and 2) analyze the potential fate and fractionation of runoff pollutants during rainfall events. We hope that the information produced will support management of runoff pollution and contribute to the preservation of urban waterways.

319

discharging runoff rapidly in most conditions. When runoff is generated, untreated runoff flows directly into the nearest waterways. The catchment of the Nanhe River (blue lines in Fig. 1A), one of the three largest rivers in the Taihu Basin, is to the west of Taihu Lake. The rivers flow from west to east and ultimately feed into Taihu Lake after traversing Xijiu, Tuanjiu, and Dongjiu Lakes and other interconnected rivers (Fig. 1A and B). The water input to Taihu Lake from the Nanhe River system accounts for around 25% of the total inputs to the lake. Agriculture, aquaculture, forestry, and urban areas dominate land use in the Nanhe catchment. The central urban area is in the middle reaches and comprises residential, commercial, and industrial areas. River and runoff samples were collected from August 2015 until May 2016. A total of 34 sampling sites, comprising 3 stagnant rivers (blue triangles, Fig. 1B) and 31 flowing rivers (red triangles, Fig. 1B), distributed throughout the Nanhe system were monitored. We collected samples of urban runoff from three road sites (green circles, Fig. 1B) in residential, commercial, and industrial areas. 2.2. Sampling strategy Six different types of samples, namely rainwater, urban runoff (UR), stagnant rivers during the wet season (SR-W), stagnant rivers during the dry season (SR-D), flowing rivers during the wet season (FR-W), and flowing rivers during the dry season (FR-D), were collected. Individual samples of local rainwater were collected in clean polyethylene vessels in an open area. Runoff samples were collected from the inlet grating at the roadside (Fig. 1C) at 5-min intervals during five rainfall events. The rainfall amounts, rainfall durations, and antecedent dry periods ranged from 5.2 to 26.8 mm, from 72 to 318 min, and from 23.8 to 359 h, respectively. We collected surface water samples from the middle of the river channels (Fig. 1D). River samples were collected once or twice a month during the dry season when the number of antecedent dry days ranged from 2 to 8 d. River samples were collected through two rainfall events in the wet season at time intervals of 0.5, 1, 2, 4, and 6 h. Samples were collected in pre-washed 1-L polyethylene bottles. It should be noted that the sampling program was not designed to identify the cause-and-effect linkages between urban runoff and rivers, but to understand the physicochemical

2. Material and methods 2.1. Study area and sampling sites We chose Yixing, a city in the Taihu Lake Basin in Eastern China, as our study area. This city is characteristic of the area and, with its extensive waterway network, has a large drainage density. Yixing has a population of approximately 1.24 million and covers a total area of 1996.6 km2, 16.8% of which is occupied by water bodies. The urban area of 66.3 km2 is crisscrossed by rivers, with a river density of up to 2.27 km km2. Yixing has a subtropical monsoon climate. The average annual temperature, rainfall, number of rain days, and evaporation are 15.7  C, 1177 mm, 136.6 d, and 849 mm, respectively. The drainage is via a rain and sewage diversion system that comprises only main pipes or a few branch pipes that are capable of

Fig. 1. (A) Geographical position. The blue lines and red virtual frame represent the Nanhe River system and the study area, respectively. (B) River sampling sites. The red and blue triangles represent the sampling sites in the flowing and stagnant rivers, respectively. The green circles represent the road runoff sampling sites in the urban area. (C) and (D) are photographs of runoff sampling and river sampling, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

320

Q. Wang et al. / Chemosphere 173 (2017) 318e325

conditions and sediment properties. We collected and analyzed a total of 246 river samples and 151 runoff samples.

analysis to fit the settling curves. We used IBM SPSS statistics version 19 (IBM Corporation, Armonk, NY, USA) for the statistical analysis.

2.3. Samples analysis 3. Results The physicochemical parameters, including pH, ORP, and EC, were measured for rainwater, runoff and river samples either in situ or in the laboratory as soon as possible after collection using a multi-parameter analyzer (Hach HQ 40d) that was calibrated before use. The particle characteristics, including turbidity, total suspended solids (TSS), organic matter content (OM), surface morphology, size distribution, and settleability of particles, were also determined. 2.3.1. Particle content and composition Turbidity was measured using a laboratory turbidimeter (Hach 2100 N). Total suspended solids and OM were determined by the gravimetric and loss-on-ignition methods, respectively. As is already well known, TSS is an indicator of the mass of suspended solids and turbidity characterizes the optical properties of suspended solids. We might speculate therefore that fine particles contribute less to TSS because of their lower density and contribute more to turbidity because of their large specific surface areas that can effectively hinder light. In other words, turbidity and TSS are indicators of particle size that cannot be expressed separately. Hence, the ratio of turbidity to TSS (Tur/TSS) can be used to characterize the relative composition of particle sizes, such that a higher Tur/TSS value indicates mainly fine particles and vice versa. 2.3.2. Morphological analysis We carried out various sediment analyses to determine if the Tur/TSS ratio could be used to characterize the relative composition of particle sizes. First, we observed the surface morphology of particles with a scanning electron microscope (Hitachi S-3400 N Ⅱ, Japan). We prepared dry solids of five representative runoff samples, the Tur/TSS of which ranged from 0.35 to 1.74. We were not able to analyze river particles for morphological properties because the TSS concentrations were very low (median 37 mg L1), which meant that it was very difficult to separate enough particles. 2.3.3. Particle size distribution (PSD) We determined the PSD of runoff samples collected at runoff times of between 0 and 260 min. The turbidities and TSS concentrations of these samples ranged from 41.6 to 728 NTU and from 70 to 1456 mg L1, respectively. We also determined the PSD for river samples with turbidities and TSS concentrations that ranged from 24.4 to 159 NTU and from 29 to 207 mg L1, respectively. We used a laser particle size analyzer for the PSD analysis (Mastersizer, 2000, Malvern Instruments Ltd.). 2.3.4. Settleability We used the precipitation method in the laboratory to investigate the settleability of runoff and river particles. Runoff or river samples were poured into a specially designed settling column with a volume of 15 L. The turbidity and TSS were measured on samples collected from the middle of the column at time intervals of 0.5, 1, 2, 4, 8, and 12 h. We analyzed six river samples and three runoff samples. 2.4. Statistical analysis Because the variables were not normally distributed, we used a nonparametric test (independent-samples t-test) to detect significant differences between the physicochemical and particle properties among the different sample types. We used regression

3.1. Differences in the physicochemical conditions 3.1.1. pH The pH value reflects the acidity of a solution. The pH of rainwater, urban runoff, and rivers ranged from 3.99 to 5.19, from 7.33 to 9.11, and from 6.99 to 8.16, respectively. The average pH decreased in the following order: UR > SR-W > FR-W > FR-D > SRD > rainwater (Table 1). Nonparametric tests indicated that the pH of urban runoff was significantly higher than that of river samples. Moreover, the pH of the stagnant rivers differed significantly between the wet and dry seasons. Runoff discharges had more influence on the pH of stagnant rivers than of flowing rivers. The acidity of the urban runoff and rivers was significantly lower than the acidity of the local rainwater. 3.1.2. ORP ORP represents the overall oxidation-reduction potential of a solution. A high positive value indicates a strong oxidizing ability. In sedimentary geology, ORP ranges have been classified into 4 groups, so that þ400eþ650 mV, þ200eþ400 mV, 0eþ200 mV, and 200e0 represent oxidation, weak oxidation, weak reduction, and reduction, respectively (Guo and Jia, 2008). The ORP values of rainwater, urban runoff, and rivers ranged from 242.5 to 376.7, from 157.2 to 245.9, and from 211.2 to 291.4 mV, respectively. In contrast to pH, the ORP was lowest for urban runoff. The ORP of rainwater and rivers indicated weak oxidation, while the ORP of urban runoff extended from weak reduction to weak oxidation. The average ORP decreased in the following order: rainwater > FR-D > SR-D > SRW > FR-W > UR (Table 1). Results from nonparametric tests indicated that the ORP values for both stagnant and flowing rivers were significantly higher than the values for urban runoff. The ORP values for flowing rivers differed significantly during the wet and dry seasons. Runoff discharge had more influence on the ORP of flowing rivers than of stagnant rivers. The ORP of both urban runoff and rivers was significantly lower than that of local rainwater. 3.1.3. EC Electroconductivity is associated with the salinity, ion concentrations and impurity of a solution. Low EC values indicate pure water. The EC of rainwater, urban runoff, and rivers ranged from 27.7 to 70.8, from 46.3 to 506.0, and from 153.4 to 656 ms cm1, respectively. The average EC was ranked as follows: FR-D > FRW > SR-D > SR-W > UR > rainwater (Table 1). Nonparametric tests indicated that the EC values of UR, SR-W, SR-D, FR-W, and FR-W were significantly different. Runoff discharges caused obvious decreases in the EC of urban rivers. The EC of the rivers was significantly higher than the EC of local rainwater. However, while the EC values of runoff were higher than those of rainwater, the differences were hard to define because of the large variations in runoff. 3.2. Particle characteristics of urban runoff and rivers 3.2.1. Particle content and composition Road deposited sediment and associated pollutants are partly washed off and transported to waterways as runoff particles. The turbidity of urban runoff and rivers ranged from 10.1 to 746 and from 4.2 to 201 NTU, respectively. The average turbidity values were ranked as follows: UR > FR-W > SR-W > FR-D > SR-D (Table 1). Runoff discharges caused the turbidity of the stagnant

Q. Wang et al. / Chemosphere 173 (2017) 318e325

321

Table 1 Physicochemical properties of rainwater, runoff, and rivers in Yixing, China. Parameters

Sample type SR-W

pH ORP (mV) EC (ms cm1) Turbidity (NTU) TSS (mg L1) OM (%)

SR-D a

FR-W b

7.56 ± 0.34 258.0 ± 21.3ab 454.7 ± 49.5b 12.2 ± 5.2b 15.7 ± 8.8b 56.3 ± 20.1a

7.87 ± 0.14 251.9 ± 12.9a 249.1 ± 86.2a 44.7 ± 28.6a 95.1 ± 77.0a 34.1 ± 12.5ab

FR-D bc

7.74 ± 0.09 250.2 ± 17.1a 455.6 ± 117.2bc 49.4 ± 27.5a 72.2 ± 58.5a 27.7 ± 7.6b

UR b

7.70 ± 0.11 263.8 ± 8.1b 614.4 ± 20.9d 41.2 ± 30.8a 51.7 ± 39.7c 22.5 ± 13.2a

Rainwater d

8.42 ± 0.43 190.9 ± 20.1c 131.4 ± 107.1e 129.3 ± 118.5c 267.3 ± 236.5a 29.8 ± 10.4b

4.49 ± 0.50e 342.9 ± 56.9d 52.4 ± 15.8e / / /

The superscript characters ae demonstrated the results of nonparametric tests, and different characters indicated that there was significant difference existed.

rivers to increase significantly. The TSS of urban runoff and rivers ranged from 16 to 1728 and from 6 to 259 mg L1, respectively. The average TSS was ranked as follows: UR > SR-W > FR-W > FRD > SR-D (Table 1). Runoff discharges caused the TSS of both flowing and stagnant rivers to increase significantly. Furthermore, the OM was highest for SR-D and lowest for FR-D, but there were no other differences. Runoff particles make a significant contribution to the impairment of waterways, and should be studied in more detail. The Tur/TSS plotted with the TSS concentrations (Fig. 2A and B) varied considerably for both runoff and rivers, and ranged from 0.24 to 1.91 and from 0.39 to 3.82, respectively. The values of the Tur/TSS ratio were noticeably different for runoff and rivers. Values of the Tur/TSS ratio were uniformly low for runoff when TSS concentrations were higher than 500 mg L1 (Fig. 2A) but there was no trend when TSS concentrations were less than 500 mg L1. This may be attributed to the TSS source, which was determined by the quantity of particles stored on the road surface, the kinetic energy of rainfall, and other factors such as the RDS load, rainfall intensity, runoff time, and antecedent dry days (Wijesiri et al., 2015; Zhao et al., 2016). Since less energy is needed to remove fine particles from the road surface, runoff with higher Tur/TSS values is possibly generated under low rainfall intensities while runoff with lower

2.0

A

1.5 1.0

Tur/TSS

.5 0.0 0

500

1000

1500

4

2000

B

3 2 1 0 0

60

120

180

240

TSS (mg/L) Fig. 2. Relationship between Tur/TSS and TSS concentrations for runoff (A) and rivers (B).

Tur/TSS values is generated under high rainfall intensities. Nevertheless, there was a clear declining trend for rivers (Fig. 2B). The Tur/TSS values were around one under most conditions (78%), and some extreme values were found when the TSS concentrations were lower than 50 mg L1, which again may be attributed to the TSS source. Rivers with higher TSS concentrations contain more sediment and sediment particles that are larger, derived from the shipping disturbances, that make more of a contribution to TSS than to turbidity. However, particles in rivers with lower TSS concentrations tend to be finer, and make less of a contribution to TSS than to turbidity. 3.2.2. Morphology of particle surfaces To validate the relationship between Tur/TSS and particle size, particles in runoff samples with Tur/TSS ratios ranging from 0.35 to 1.74 were separated and dried to facilitate examination of the surface morphology (Fig. 3AeC). The majority of particles were irregularly shaped and had variable dimensions. As we expected, the particle size decreased as the Tur/TSS increased. For example, the Tur/TSS ratio of C3 was significantly higher than that of the other samples and many nano-scale particles were visible in the SEM micrograph; also, there were more micron- and submicronscale particles in samples with lower Tur/TSS values. This shows that the Tur/TSS ratio can be used to characterize the relative composition of the particle sizes. 3.2.3. Particle size distribution (PSD) In sedimentary geology, particles with diameters <4, between 4 and 63, 63 and 125, 125 and 250, 250 and 500, and 500 and 1000 are classified as silt, clay, very fine sand, fine sand, medium sand, and coarse sand, respectively (Zhao et al., 2010). We differentiated the particle sizes of sediments in runoff and rivers into these seven fractions (Table 2). Runoff and river sediments were mainly comprising silt and clay (<63 mm), which accounted for 90.7 ± 12.5% and 88.3 ± 9.2% of their total volumes, respectively. The clay fraction dominated (4e63 mm), and accounted for 67.1 ± 6.8% and 76.6 ± 14.1% of the sediment particles in rivers and runoff, respectively. Furthermore, the proportion of <4 mm particles in rivers was nearly twice the corresponding proportion in runoff, indicating that rivers transported finer particles. Nano- and submicron-scale particles were the main contributors to the differences in PSD between runoff and rivers. Previous studies have shown that pollutant concentrations may be one order of magnitude higher in nano-scale (<0.3 mm and 0.3e1 mm) particles than in submicron- and micron-scale particles (Fedotov et al., 2014). Because particles <16 mm are acknowledged to be closely associated with decreases in water clarity, we carried out further comparisons of the <16 mm particles (Table 2). The proportions of <16 mm particles in different runoff and river samples varied considerably and ranged from 22.8% to 86.7% and from 45.0% to 83.5%, respectively, with averages of 59.4 ± 17.4% and 69.1 ± 10.8%, respectively. The proportion of <16 mm particles was slightly higher

322

Q. Wang et al. / Chemosphere 173 (2017) 318e325

A Tur/TSS=0.35

B Tur/TSS=0.48

C2 Tur/TSS=1.03

C1 Tur/TSS=0.90

C3 Tur/TSS=1.74

Fig. 3. SEM micrographs of runoff particles with different Tur/SS values. The magnifications were 6000  , 8000  , and 10000  for groups A, B, and C, respectively.

Table 2 The PSD of urban runoff and rivers in Yixing, China. Ranges (mm)

Rivers %

Urban runoff %

<4 4e63 63e125 125e250 250e500 500e1000 Nano- and submicron-scale <1 1e4 4e10 10e16

21.2 ± 8.7 67.1 ± 6.8 5.7 ± 4.3 2.1 ± 2.1 3.9 ± 3.0 /

14.1 ± 8.5 76.6 ± 14.1 6.2 ± 8.1 2.4 ± 4.4 0.6 ± 0.9 0 ± 0.1

2.1 ± 2.2 20.1 ± 6.2 32.5 ± 4.1 20.3 ± 5.1

2.4 ± 2.4 11.7 ± 6.7 25.0 ± 16.1 14.5 ± 2.2

in rivers than in runoff, because of the higher proportion of submicron-scale (1e4, 4e10, and 10e16 mm) particles. It is worth noting that there were more nano-scale (<1 mm) particles in runoff than in rivers. These very fine particles should receive more attention because of their high pollutant concentrations, mobility, and bioavailability. 3.2.4. Settleability The variations in turbidity with settling time are illustrated in Fig. 4. The settling of runoff and river particles were a good fit for the exponential distribution. Goodness of fit values ranged from 0.743 to 0.977, which indicates that the settling speed gradually decreased with time (Zhao et al., 2010; Kayhanian et al., 2012). This is because larger and denser particles settled rapidly and smaller and lighter particles settled more slowly. The Tur/TSS also increased as the settling time increased (not shown in the paper). The settling process can be divided into two stages: the initial period with rapid settling and a steady period with homogeneous settling. The runoff particles settled more rapidly than the river particles in the initial period, but they settled at comparable speeds in the steady period. This result was somewhat expected because runoff particles are generally denser than river particles. Moreover, river samples with lower turbidity fitted better than samples with higher turbidity, whereas there was no trend between the fitting and turbidity of runoff samples. The better fit was attributed to the homogeneous composition of the particle sizes. As discussed in

section 3.2, the particle sizes of river samples with low turbidity were more homogeneous than those with high turbidity. However, rainfall intensity rather than turbidity determined the particle size composition of runoff samples. Therefore, it was the particle composition that actually determined the settling process, and the settleability of samples with lower Tur/TSS values was better despite their weaker fitting.

4. Discussion 4.1. The fate and fractionation of runoff pollutants The pollutants associated with particles in different binding states can leach into aqueous phases from particle surfaces under specific pH, ORP, or EC conditions (Kartal et al., 2006). As shown in Table 3, HMs are usually partitioned into four fractions by the modified BCR three step sequential extraction procedure, namely acid extractable (F1), reducible (F2), oxidizable (F3), and residual (F4) fractions that associate with carbonates, Fe and Mn oxides, organic matter, and minerals, respectively (Bacon et al., 2006). The first three fractions (F1þF2þF3) comprise the mobile part and may be released to the aqueous phase through desorption or dissolution (Zhang and Wang, 2009). As mentioned by various researchers, the F1 fraction is sensitive to low pH, the F2 fraction is sensitive to either low pH or low ORP, while the F3 fraction is sensitive to high ORP values (Li et al., 2013). Acid rain (with a pH of 3.99e5.19) in the present study provides conditions that are conducive to the release of the F1 and F2 fractions. The significantly higher pH of urban runoff indicates that the acidity of the rainwater was neutralized by the alkalinity provided by RDS (Huber et al., 2016). Although the pH of the rivers increased slightly following runoff discharges, the near-neutral environment did not facilitate the release of particle-bound pollutants. The ORP values of rainwater, runoff, and rivers were within the weak reduction and weak oxidation categories. When the wash-off commences, there is leaching of the F3 fraction of RDS pollutants because of the high oxidizing ability of rainwater, and the ORP in runoff subsequently decreases. The slight increase in the ORP of rivers after receiving runoff may further enhance the release of the F3 fraction. Furthermore, the EC increased from

Q. Wang et al. / Chemosphere 173 (2017) 318e325

200

250 initial period

A

steady period

initial period

B

steady period

160

200 UR(222 NTU) R2=0.856 UR(84.5 NTU) R2=0.929 UR(66.1 NTU) R2=0.743

150

Turbidity (NTU)

Turbidity (NTU)

323

100

FR (160 NTU) FR (136 NTU) FR (59.4 NTU) FR (47 NTU)

120

R2=0.896 R2=0.884 R2=0.940 R2=0.977

80

40

50

0

0 0

30

60

90

120

150

180

0

30

60

90

Time (h)

120

150

180

Time(h)

Fig. 4. Variations in turbidity with settling time for runoff (A) and river (B) samples.

Table 3 The characteristics of different fractions of HMs. Items

Fractions F1 (acid extractable)

F2 (reducible)

F3 (oxidizable)

F4 (residual)

Binding state Sensitive to environment Proportion (%)

Carbonates Low pH 0.7e24.7 0e58.5 8.7e54.7

Fe and Mn oxides Low pH, low ORP 3.4e59.0 7.8e85.4 22.8e65.5

Organic matter High ORP 11.1e67.8 5.2e63.0 3.8e24.5

Minerals / 1.6e62.2 5.6e25.6 10.1e38.7

Cu Pb Zn

The proportion of Cu, Pb, Zn in different fractions are obtained from Oahu, Hawaii, US; Hangzhou and Nanjing, China; Kayseri, Turkey; Nerima city, Tokyo; Dresden, Germany.

rainwater, to runoff, and then to rivers, indicating the increasingly complicated physical, chemical, and biological transformations of pollutants. The TSS of runoff was approximately 5e17 times higher than that of rivers, most of which contributed to the river sediments. Because of the variations between the runoff and river environments, pollutants will inevitably be redistributed between the particulate and dissolved phases by physical, chemical, and biological processes. Worldwide, there is considerable variation in the physicochemical conditions of rainwater, runoff, and rivers. Previous studies have reported pH values ranging from 3.77 to 7.12, from 4.15 to 9.11, and from 6.95 to 9.25; ORP values ranging from 166 to 376.7 mV, from 157.2 to 245.9 mV, and from 211.2 to 291.4 mV, and EC values ranging from 3.9 to 70.8 ms cm1, from 24.1 to 2670 ms cm1, and from 153.4 to 1713 ms cm1 in rainwater, runoff, and rivers, respectively (McQueen et al., 2010; Yang et al., 2013; Zhang et al., 2013; Xu et al., 2015; Mimura et al., 2016; Morillas et al., 2016; Zhao et al., 2017; Zheng et al., 2017). The differences may be attributed to the influence of anthropogenic activities, such as emissions of acid aerosols (mainly SOx and NOx) or wastewater from industrial and traffic-related activities (Ouyang et al., 2015; Morillas et al., 2016). Furthermore, the fractions of Cu, Pb, and Zn concentrations in urban RDS (Table 3) from six cities worldwide varied greatly in all the fractions of a given metal, which may be related to the source of RDS and the associated HMs (Zhang and Wang, 2009; Sutherland et al., 2012; Kumar et al., 2013; Li et al., 2013; Maniquiz-Redillas and Kim, 2014; Zhang et al., 2015). The fractionation of pollutants will therefore vary considerably from site-to-site because of variations in the physicochemical properties of rainwater, runoff, and rivers, and the binding states of pollutants.

4.2. Implications for runoff pollution control It is generally accepted that urban runoff with high pollutant concentrations or loads is under control. However, the potential risk from runoff is determined by the bioavailability of pollutants rather than the total mass (Zhang et al., 2016). Because of the importance of fine particles in the accumulation and mobility of particle-bound pollutants, it is more bioavailable for particles with fine sizes. Therefore, particle size must be considered when controlling runoff. The PSD results indicate that both river and runoff particles were mainly comprising the <63 mm fraction (88.3%e 90.7%), and a considerable proportion of the particles were <10 mm (39.1%e54.7%). Heavy metals are mainly concentrated in nanoscale (<1 mm) particles, and <16 mm particles are acknowledged to be closely associated with the degradation of water clarity, so more attention should be paid to nano- and submicron-scale particles in RDS and runoff, and runoff control measures should be designed accordingly (Fedotov et al., 2014). Most runoff particles in Yixing were <125 mm (96.9%), which is comparable with the results from other studies (89.5%e95.0%) (F. T. Wakida et al., 2013; Zhao and Li, 2013). Hence, RDS particles in this size category are more likely to be washed from the surface during rainfall. Runoff pollution derives from the transport of mobile RDS (<125 mm) and the leaching of immobile RDS retained on the road surface. Therefore, the implementation of, and improvements in, road sweeping will help to prevent the leaching of dissolved pollutants from coarse RDS. We found that Tur/TSS was an effective indicator of the relative composition of particle sizes. Many researchers have reported that pollutants are concentrated in fine particles, so it might be expected that runoff with a high Tur/TSS value has higher pollutant

324

Q. Wang et al. / Chemosphere 173 (2017) 318e325

concentrations. Furthermore, the general occurrence of runoff with high Tur/TSS ratios under low rainfall intensities implies that runoff with these properties comprises only a small part of total runoff. Hence, even though the volume is small, runoff with high Tur/TSS values should be controlled because of the extremely high pollutant concentrations. Further, the relationship between Tur/TSS and the settleability of particles indicates that the settleability of runoff with higher Tur/TSS was weaker than runoff with lower Tur/ TSS. Hence, the Tur/TSS ratio can also be used to set appropriate and optimal hydraulic retention times (HRT) in runoff control measures that involve presettling processes to achieve most effectiveness (Maniquiz-Redillas et al., 2014). 5. Conclusion We evaluated the physicochemical conditions (pH, ORP, and EC) and particle characteristics (turbidity, TSS, size distributions, and settleability of particles) that governed the fate and fractionation of particle-bound pollutants in rainwater, urban runoff, and rivers. The results show that the physicochemical properties (pH, ORP, EC) of urban runoff, rivers, and rainwater differed significantly. The low pH and high ORP in rainwater were conducive to the leaching of particle-bound pollutants. Different environments and the more sophisticated physical, chemical, and biological processes in rivers may contribute to the redistribution of pollutants between the particulate and dissolved phases after runoff is discharged into waterways. Both runoff and river particles were predominantly composed of silt and clay (<63 mm), most of which were within the nano- and submicron-scale (<16 mm) categories. There should be concern about the higher proportion of nano-scale (<1 mm) particles in runoff because of their high pollutant concentrations, mobility, and bioavailability. The ratio of turbidity to TSS (Tur/TSS), as an indicator of the relative composition of particle sizes, was associated with the accumulation of pollutants and the settleability of particles, and so could contribute to assessments of the effectiveness of runoff pollution control measures. Acknowledgements This study is supported by the National Natural Science Foundation of China (Grant No.51508447), the National Program of Water Pollution Control in China (Grant No. 2014ZX07305-002), and Program for Innovative Research Team in Shaanxi (PIRT) (Grant No. 2013KCT-13). References Bacon, J.R., Farmer, J.G., Dunn, S.M., Graham, M.C., Vinogradoff, S.I., 2006. Sequential extraction combined with isotope analysis as a tool for the investigation of lead mobilisation in soils: application to organic-rich soils in an upland catchment in Scotland. Environ. Pollut. 141, 469e481. Chow, M.F., Yusop, Z., 2014. Sizing first flush pollutant loading of stormwater runoff in tropical urban catchments. Environ. Earth Sci. 72, 4047e4058. Chow, M.F., Yusop, Z., Shirazi, S.M., 2013. Storm runoff quality and pollutant loading from commercial, residential, and industrial catchments in the tropic. Environ. Monit. Assess. 185, 8321e8331. Duong, T.T.T., Lee, B.-K., 2009. Partitioning and mobility behavior of metals in road dusts from national-scale industrial areas in Korea. Atmos. Environ. 43, 3502e3509. Fedotov, P.S., Ermolin, M.S., Karandashev, V.K., Ladonin, D.V., 2014. Characterization of size, morphology and elemental composition of nano-, submicron, and micron particles of street dust separated using field-flow fractionation in a rotating coiled column. Talanta 130, 1e7. Gasperi, J., Sebastian, C., Ruban, V., Delamain, M., Percot, S., Wiest, L., Mirande, C., Caupos, E., Demare, D., Kessoo, M.D., Saad, M., Schwartz, J.J., Dubois, P., Fratta, C., Wolff, H., Moilleron, R., Chebbo, G., Cren, C., Millet, M., Barraud, S., Gromaire, M.C., 2014. Micropollutants in urban stormwater: occurrence, concentrations, and atmospheric contributions for a wide range of contaminants in three French catchments. Environ. Sci. Pollut. Res. Int. 21, 5267e5281. Gunawardana, C., Goonetilleke, A., Egodawatta, P., Dawes, L., Kokot, S., 2012. Source

characterisation of road dust based on chemical and mineralogical composition. Chemosphere 87, 163e170. Gunawardana, C., Egodawatta, P., Goonetilleke, A., 2015. Adsorption and mobility of metals in build-up on road surfaces. Chemosphere 119, 1391e1398. Guo, P.W.,P., Jia, S., 2008. Research progress of the environmental impact factors of internal phosphorus loading release in river systems (in Chinese). J. Nanjing For. Univ. (Nat Sci) 32, 117e121. Helmreich, B., Hilliges, R., Schriewer, A., Horn, H., 2010. Runoff pollutants of a highly trafficked urban roadecorrelation analysis and seasonal influences. Chemosphere 80, 991e997. Huber, M., Welker, A., Helmreich, B., 2016. Critical review of heavy metal pollution of traffic area runoff: occurrence, influencing factors, and partitioning. Sci. Total Environ. 541, 895e919. Joshi, U.M., Balasubramanian, R., 2010. Characteristics and environmental mobility of trace elements in urban runoff. Chemosphere 80, 310e318. Kartal, S., Aydin, Z., Tokalioglu, S., 2006. Fractionation of metals in street sediment samples by using the BCR sequential extraction procedure and multivariate statistical elucidation of the data. J. Hazard. Mater. 132, 80e89. Kayhanian, M., McKenzie, E.R., Leatherbarrow, J.E., Young, T.M., 2012. Characteristics of road sediment fractionated particles captured from paved surfaces, surface run-off and detention basins. Sci. Total Environ. 439, 172e186. Kim, J.Y., Sansalone, J.J., 2008. Event-based size distributions of particulate matter transported during urban rainfall-runoff events. Water Res. 42, 2756e2768. Kumar, M., Furumai, H., Kurisu, F., Kasuga, I., 2013. Tracing source and distribution of heavy metals in road dust, soil and soakaway sediment through speciation and isotopic fingerprinting. Geoderma 211e212, 8e17. Li, H., Qian, X., Hu, W., Wang, Y., Gao, H., 2013. Chemical speciation and human health risk of trace metals in urban street dusts from a metropolitan city, Nanjing, SE China. Sci. Total Environ. 456e457, 212e221. Li, H., Shi, A., Zhang, X., 2015. Particle size distribution and characteristics of heavy metals in road-deposited sediments from Beijing Olympic Park. J. Environ. Sci. China 32, 228e237. Loperfido, J.V., Noe, G.B., Jarnagin, S.T., Hogan, D.M., 2014. Effects of distributed and centralized stormwater best management practices and land cover on urban stream hydrology at the catchment scale. J. Hydro 519, 2584e2595. Maniquiz-Redillas, M., Kim, L.-H., 2014. Fractionation of heavy metals in runoff and discharge of a stormwater management system and its implications for treatment. J. Environ. Sci. China 26, 1214e1222. Maniquiz-Redillas, M.C., Geronimo, F.K., Kim, L.H., 2014. Investigation on the effectiveness of pretreatment in stormwater management technologies. J. Environ. Sci. China 26, 1824e1830. McKee, L.J., Gilbreath, A.N., 2015. Concentrations and loads of suspended sediment and trace element pollutants in a small semi-arid urban tributary, San Francisco Bay, California. Environ. Monit. Assess. 187, 499. McQueen, A.D., Johnson, B.M., Rodgers Jr., J.H., English, W.R., 2010. Campus parking lot stormwater runoff: physicochemical analyses and toxicity tests using Ceriodaphnia dubia and Pimephales promelas. Chemosphere 79, 561e569. Mimura, A.M.S., Almeida, J.M., Vaz, F.A.S., de Oliveira, M.A.L., Ferreira, C.C.M., Silva, J.C.J., 2016. Chemical composition monitoring of tropical rainwater during an atypical dry year. Atmos. Res. 169, 391e399. Morillas, H., Marcaida, I., Maguregui, M., Carrero, J.A., Madariaga, J.M., 2016. The influence of rainwater composition on the conservation state of cementitious building materials. Sci. Total Environ. 542, 716e727. Morselli, L., Olivieri, P., Brusori, B., Passarini, F., 2003. Soluble and insoluble fractions of heavy metals in wet and dry atmospheric depositions in Bologna. Italy. Environ. Pollut. 124, 457e469. Ouyang, W., Guo, B., Cai, G., Li, Q., Han, S., Liu, B., Liu, X., 2015. The washing effect of precipitation on particulate matter and the pollution dynamics of rainwater in downtown Beijing. Sci. Total Environ. 505, 306e314. Reidsma, P., Feng, S., van Loon, M., Luo, X., Kang, C., Lubbers, M., Kanellopoulos, A., Wolf, J., van Ittersum, M.K., Qu, F., 2012. Integrated assessment of agricultural land use policies on nutrient pollution and sustainable development in Taihu Basin, China. Environ. Sci. Policy 18, 66e76. Sansalone, J., Ying, G., 2008. Partitioning and granulometric distribution of metal leachate from urban traffic dry deposition particulate matter subject to acidic rainfall and runoff retention. Water Res. 42, 4146e4162. Sun, S., Barraud, S., Castebrunet, H., Aubin, J.B., Marmonier, P., 2015. Long-term stormwater quantity and quality analysis using continuous measurements in a French urban catchment. Water Res. 85, 432e442. Sutherland, R.A., Tack, F.M., Ziegler, A.D., 2012. Road-deposited sediments in an urban environment: a first look at sequentially extracted element loads in grain size fractions. J. Hazard. Mater. 225e226, 54e62. Wakida, F.T., S.M.H, Garcia Flores, E., Pinon Colin, T.D.J., Espinoza Gomez, H., Ames Lopez, A., 2013. Pollutant association with suspended solids in stormwater in Tijuana, Mexico. Int. J. Environ. Sci. Technol. 11, 319e326. Wijesiri, B., Egodawatta, P., McGree, J., Goonetilleke, A., 2015. Process variability of pollutant build-up on urban road surfaces. Sci. Total Environ. 518e519, 434e440. Xu, Z., Wu, Y., Liu, W.-J., Liang, C.-S., Ji, J., Zhao, T., Zhang, X., 2015. Chemical composition of rainwater and the acid neutralizing effect at Beijing and Chizhou city, China. Atmos. Res. 164e165, 278e285. Yang, Y., He, Z., Wang, Y., Fan, J., Liang, Z., Stoffella, P.J., 2013. Dissolved organic matter in relation to nutrients (N and P) and heavy metals in surface runoff water as affected by temporal variation and land uses e a case study from Indian river area, south Florida, USA. Agr. Water Manage 118, 38e49.

Q. Wang et al. / Chemosphere 173 (2017) 318e325 Yuen, J.Q., Olin, P.H., Lim, H.S., Benner, S.G., Sutherland, R.A., Ziegler, A.D., 2012. Accumulation of potentially toxic elements in road deposited sediments in residential and light industrial neighborhoods of Singapore. J. Environ. Manage 101, 151e163. Zhang, M., Wang, H., 2009. Concentrations and chemical forms of potentially toxic metals in road-deposited sediments from different zones of Hangzhou, China. J. Environ. Sci. China 21, 625e631. Zhang, Q., Wang, X., Hou, P., Wan, W., Ren, Y., Ouyang, Z., Yang, L., 2013. The temporal changes in road stormwater runoff quality and the implications to first flush control in Chongqing, China. Environ. Monit. Assess. 185, 9763e9775. Zhang, J., Hua, P., Krebs, P., 2015. The build-up dynamic and chemical fractionation of Cu, Zn and Cd in road-deposited sediment. Sci. Total Environ. 532, 723e732. Zhang, J., Hua, P., Krebs, P., 2016. The influences of dissolved organic matter and surfactant on the desorption of Cu and Zn from road-deposited sediment. Chemosphere 150, 63e70. Zhao, H., Li, X., 2013. Understanding the relationship between heavy metals in roaddeposited sediments and washoff particles in urban stormwater using

325

simulated rainfall. J. Hazard. Mater. 246e247, 267e276. Zhao, H., Li, X., Wang, X., Tian, D., 2010. Grain size distribution of road-deposited sediment and its contribution to heavy metal pollution in urban runoff in Beijing, China. J. Hazard. Mater. 183, 203e210. Zhao, H., Chen, X., Hao, S., Jiang, Y., Zhao, J., Zou, C., Xie, W., 2016. Is the wash-off process of road-deposited sediment source limited or transported limited? Sci. Total Environ. 563e564, 62e70. Zhao, K., Song, K., Pan, Y., Wang, L., Da, L., Wang, Q., 2017. Metacommunity structure of zooplankton in river networks: roles of environmental and spatial factors. Eco. Indic. 73, 96e104. Zheng, L., Liu, Z., Yan, Z., Zhang, Y., Yi, X., Zhang, J., Zheng, X., Zhou, J., Zhu, Y., 2017. pH-dependent ecological risk assessment of pentachlorophenol in Taihu lake and Liaohe river. Ecotoxicol. Environ. Saf. 135, 216e224. Zuo, X., Fu, D., Li, H., 2012. Speciation distribution and mass balance of copper and zinc in urban rain, sediments, and road runoff. Environ. Sci. Pollut. Res. Int. 19, 4042e4048.