Supporting Information (SI) for
Modeling the air-soil exchange, secondary emissions and residues in soil of polychlorinated biphenyls in China
Authors: Song Cui1,2, Qiang Fu1, Yi-Fan Li2,3,1*, Jianmin Ma4, Chongguo Tian5 , Liyan Liu2& Leiming Zhang6
Affiliations:
1 International Joint Research Center for Persistent Toxic Substances (IJRC-PTS), School of Water Conservancy and Civil Engineering, Northeast Agricultural University, Harbin, Heilongjiang, 150030, China
2 IJRC-PTS, State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology, Harbin, Heilongjiang, 150090, China
3 IJRC-PTS-NA, Toronto, M2N 6X9, Canada
4 Key Laboratory of Western China’s Environmental System, Ministry of Education, College of Earth and Environment Sciences, Lanzhou University, Lanzhou, Gansu, 730000, China
5 Yantai Institute of Coastal Zone Research, Chinese Academy of Sciences, Yantai, Shandong, 264003, China
6 Air Quality Research Division, Science and Technology Branch, Environment Canada, 4905 Dufferin Street, Toronto, Ontario, M3H 5T4, Canada
* Corresponding author: Yi-Fan Li
E-mail address:
Number of pages: 24
Number of tables: 3
Number of figures: 10
Contents
A1. Model region and grid system·················································································2
A2. Model framework····································································································2
A 2.1 Transfer module·····························································································4
A 2.2 Transport module·························································································11
A3. Model input data····································································································13
A4. Model evaluation···································································································17
A 4.1 Comparison between modeled and measured soil concentrations···············17
A 4.2 Comparison between modeled and measured air concentrations················18
A5. Air-soil exchange and secondary emissions and residues·····································19
A5.1 Air-soil exchange·································································································19
A5.2 Secondary emissions and residues·······································································21
References····················································································································22
A1. Model region and grid system
Chinese Gridded Industrial Pollutants Emission and Residue Model (ChnGIPERM) used in this study has been developed to describe transport, transfer, and environmental behavior of PCB28 in China from 1965 to 2010. Model domain and grids adopted in the present study are illustrated in Fig.A1.
Figure A1. Model domain and grid system (1 cell=1/4°×1/6° latitude/longitude, 1 cell = approximately 24×24 km). The map was drawn by the software of Surfer 9.0, http://www.goldensoftware.com/.
A2. Model framework
The ChnGIPERM employed in this investigation is developed based on both Chinese Gridded Pesticide Emission and Residue Model (ChnGPERM), which has been used in numerical studies of α-HCH budget and atmospheric outflow from China and environmental fate of β-HCH[1, 2], and Gridded Basin-based Pesticide Mass Balance Model (GB-PMBM), which has been applied to assess α-HCH budget in the Taihu region, China [3]. ChnGIPERM is a gridded mass balance model on a gridded system with a 1/6° latitude and 1/4° longitude resolution.
Figure A2. The modeling processes in a grid cell. The map was drawn by the software of Microsoft Visio 2010, https://products.office.com/.
The main body of the model that including both transfer and transport modules. For the transfer module, a level Ⅳ fugacity method was employed to calculate and describe the changes in industrial pollutant concentrations and inter-compartmental transfer of the modeled chemicals in the multimedia environment. The transfer module consists of 6 soil types in 4 soil vertical layers, water, sediment, and the air compartment that including two layers are the atmospheric boundary (ABL, 0~1000 m) and the atmospheric low troposphere (ALT, 1000~4000 m). There are six soil types, including urban land, dry cropland, paddy field, forestry, grassland, and uncultured land. In the soil matrixes, the model adopts four well-mixed soil layers, with depths at 0.1cm, 1.0 cm, 20 cm, and 30 cm from top to bottom [4].The transport module describes the mass exchange of chemicals between simulation grid cells driven by atmospheric transport (wind direction and speed) and water current. Detailed transfer and transport processes considered in the model at a model grid cell are displayed in Fig.A2.
A2.1 Transfer module
A dynamic fugacity theory was employed to calculate the concentrations and inter-compartmental transfer of the PCB28 between adjacent compartments at a grid cell [5]. For each compartment at a grid cell, the change rate of fugacity equals to the difference between input and output rates divided by a product of compartment volumes and fugacity capacity. The following differential equations (A1-7) describe the change in fugacity in four soil layers with various soil types, air, water, and sediment as:
(A1)
(A2)
(A3)
(A4)
(A5)
(A6)
(A7)
In Eqs (A1-7), T and D represent quantity of transfer and transfer rate coefficient, f, V, Z and t are fugacity, compartment volume, fugacity capacity and time, respectively. The subscript 1, 2, 3 and 4 represent each compartment in the sequence of soil, air, water, and sediment, respectively. The subscript I, II, III and IV are four soil layers from the upper to bottom, respectively. Subscript v, q, d, m, l, r, b, s, a, and e represent the diffusion (absorption and volatilization), dry deposition, wet deposition, rain dissolution, leaching, degradation, resuspension, sedimentation, air advection, and emission data of chemical, respectively.
The calculation of partition coefficients, Z and D in non-diffusive processes are outlined in Table A1 according to previous studies [4, 5]. The values of corresponding environmental parameters to calculate the Z values in Table A1 and corresponding physicochemical properties of PCB28 used in the model are presented in Table A2.
Table A1 Calculation of partition coefficients, Z values
Partition coefficients / EquationOctanol-air / log Koa = a + b´RRT, a=-2.0296+1310.7/T, b = -5.5305+5879.8/T
Organic matter-air / Kom-a=Koa
Soil-air / Ksa=Zs/Za
Aerosol-air (m3/μg) / logKP=logKoa+logфaom-11.91
Aerosol-air (dimensionless) / K'P=KPTSP=Z'q/Za
Air-water / Kaw= KH(T)/RT*
Mineral matter-water / logKmw=0.7logAW-10.68
Mineral matter-air / Kma=KmwKaw
Suspension-water (dimensionless) / K'sus=фsocKocρsus
Sediment-water(dimensionless) / Kse=фseocKocρse
Z values (mol m-3 Pa-1)
Air / Za=1/RT
Organic matter / Zom=ZaKoa
Aerosol / Z'q=K'pZa
Water / Zw=1/KH(T)
Mineral matter / Zmm=KmwZw
Suspension / Zsus=ZwK'sus
Sediment / Zse=ZwK'se
Bulk soil / Z1=фsaZa+фswZw+фsomZom+фsmmZmm
Bulk air / Z2=Za+Z'q
Bulk water / Z3=Zw+Zsus
Bulk sediment / Z4=фse-wZw+(1-фse-w)Zse
* log(KH(T))=11.97-3100/T [7]; R=8.314 Pa m3 mol-1K-1; T is absolute temperature.
Table A2 The physico-chemical properties parameters of the PCB28 used in the model
Parameters(unit) / Value / RefMolar mass (g mol-1) / 257.54 / [6]
Molar volume:VM (cm3 mol-1) / 289.1 / [6]
Henry’s law constant (Pa m3 mol-1 )(250C) / 31 / [6]
Melting point (0C) / 57 / [6]
Aqueous solubility (g m-3)
Vapor pressure (Pa) / 0.16
0.034 / [6]
[6]
Continuing tableA2
Parameters(unit) / Value / Ref
Log Koc (dimensionless) (250C) / 4.28 / [6]
Log Kow (dimensionless) (250C) / 5.67 / [6]
Degradation rate in air: Ka (h-1) / 3.50E-7 / [7]
Degradation rate in water: Krw (h-1) / 4.07E-5 / [8]
Degradation rate in sediment: Krse (h-1) / 1.26E-5 / [8]
Base degradation rate in the six kinds of soil: Ks0 (h-1) / 1.26E-5 / [8]
Activation energy of deg. by OH radicals (J mol-1) / 8314 / [7]
Activation energy for freshwater (J mol-1) / 30000 / [9]
Activation energy for coastal/oceanic water (J mol-1) / 30000 / [9]
Activation energy for soil (J mol-1) / 30000 / [9]
In the exchange layer, volatilization from the top soil layer to air is defined by
, (A8)
where chemical diffuse in the boundary layer is given by
, (A9)
where Ar is the area of the model cell (24 km ´ 24 km). Kv is the boundary-layer mass transfer coefficient, and
, (A10)
is the chemical capacity for air. Where the boundary-layer mass transfer coefficient can be expressed by
(A11)
In which Ba is given in Eq (A14).
The chemical diffuses in air is defined by
, (A12)
where
(A13)
in which the molecular diffusivity in air (m2 h-1) is expressed as
, (A14)
and fsa is volume fraction air in soil, fsw is volume fraction water in soil, and Vm is molar volume (cm3 mol-1).
The chemical diffuses in water is defined by
, (A15)
where Yd is the log mean diffusion distance in soil (layer 1, 2, 3, and 4), defined by Mackay [5]
(A17)
where depbot and deptop are depth of bottom and top for soil layer, and
,
In which the diffusivity in water (m2 s-1) can be expressed by
, (A18)
where m is molecular viscosity in water (= 4.96).
In the exchange layer, the volatilization from the buffer layer is described by the diffusion between this layer and buffer layer. Generally, diffusions between soil layers n and n+1 are defined by
, (A19)
where Da and Dw are given by Eqs. (A12) and (A15). Other “D” values are given below.
The dry particle deposition is
, (A20)
where Gq is the dry particle deposition flux,
, (A21)
vd (m s-1) is the dry deposition velocity, which is calculated by the formulae based on [5] result:
, (A22)
where u* is friction velocity, Z0 (m) is roughness length, and
. (A23)
In (A23) Kp is the particle-gas partition coefficient (m3 mg-1), given by [10]
, (A24)
where f’aom is the mass fraction of organic carbon (=0.2), and Koa is the octanol-air partition coefficient, given by
log Koa = a + b´RRT, a = -2.0296+1310.7/T, b = -5.5305+5879.8/T (A25)
In the above equation, RRT is the coefficient (=0.3993, for PCB28) [10].
The predicted particulate fraction φ for chemical is calculated as
(A26)
where TSP (mg m-3) is the total suspended particle matter.
The diffusion due to wet particle deposition Dd is geiven by
. (A27)
Zq is computed from Eq. (A23), Gd is the wet deposition flux:
, (A28)
where precipitation flux is defined by
, (A29)
where VR is the precipitation rate (m s-1), and the aerosol scavenging ratio Qr has a typical value of 4.0 ´ 105 as given before.
The rain dissolution rate Dm is defined by
, (A30)
where Gc and Zq are given by Eqs. (A29) and (A23), respectively.
The leaching in soil Dl is defined by,
, (A31)
where,
, Lw is the leaching rate (typical value = 0.002), and
. (A32)
Transfer rate coefficient of diffusions between water and air (D2-3,v and D3-2,v) in Eqs (A5-6) is defined by
(A33)
where Ar is the area of water surface at a model grid cell. KOL is the overall mass transfer coefficient is defined by Schwarzenbach [11]
(A34)
where kw and ka are water-side and air-side mass transfer coefficients, respectively. They are defined by
(A35)
(A36)
where Sc(a) and Sc(w) are the air (=2.9) and water (=1000) phase Schmidt number, U10 is the wind speed at 10 m height.
Transfer rate coefficient of diffusions between water and sediment (D3-4,v and D4-3,v) in Eqs (A6-7) is described by
(A37)
The chemical diffuses in water (Dw) and sediment (Dse) are defined by
(A38)
(A39)
In Eqs (A37-38), the water-side boundary-layer mass transfer coefficient can be expressed by
(A40)
where the diffusivity in water (Bw) is given by Eq (A18) and Yw is the boundary-layer thickness of water (0.0001 m).
The boundary-layer mass transfer coefficient of sediment-side is defined by
(A41)
where Bse is the diffusivity in sediment and Yse is the log mean diffusion distance in sediment which is solved by Eq (A17). Bse is given by Wu and Gschwend [12]:
(A42)
where φse-w is the volume fraction of water in sediment (=0.43) [1]. Kse is partition coefficients between sediment and water, expressed by
(A43)
where φsed is the content of organic carbon in sediment (0.033) [1]; ρsed is density of sediment (1540 g L-1), and Koc is listed in Table A2.
The change in degradation rate of PCB28 in soil, water and sediment with the temperature was calculated by
(A44)
where Krs is the temperature-modified degradation rate constant (h-1), Ks is the reference degradation rate constant at 298 K, ΔE is the activation energy (Table A2), R is the gas law constant at 8.314 J K-1mol-1, T0 is the reference temperature (298 K), and Te is the average temperature (K) over the grid cell.
In the atmosphere, degradation rate constants alter not only with temperature fluctuation, but also OH radical concentration [8]. Hence, the following relationship is adopted in this study.
(A45)
where Kra is a modified degradation rate constant in the atmosphere, Kr0 is the reference degradation rate constant associated with the concentration of hydroxyl radicals [OH] in the atmospheric compartment. The values of these parameters have been listed in Table A2.
A2.2 Transport module
A Lagrangian method was employed to calculate the mass exchange of atmospheric horizontal advection for PCB28 between inter-grid cells in the atmosphere. The transport of chemical is driven by wind direction and speed. The distance and concentration of transport are defined as following:
(1) The distance of transport
The initial-position of emission source is at S(t), which corresponding grid cell is (X1,Y1). The final position through an advection time step is S’(t+δt), which corresponding grid cell is (Xi,Yj). The average wind speed between (X1,Y1) and (Xi,Yj) is as wind speed of long-range atmospheric transport. According to the amounts of transport grid along X and Y direction affirm the transport direction and concentration.
(A46)
(A47)
where , is the velocity of longitudinal and latitudinal direction at initial-position for chemical, respectively. , is the velocity of longitudinal and latitudinal direction at final-position for chemical, respectively. , is the length of longitudinal and latitudinal direction for a grid cell, respectively. , is the transported grid amounts of longitudinal (n) and latitudinal (m) direction for chemical, respectively. is the constraint condition, which the 260 and 228 represent the grid amounts along longitudinal and latitudinal direction, respectively.
(2) The transmission concentration
In general, the transmission concentration decreases with increase of distance from emission source to final destination for chemical. According to the transported grid amounts to define the concentration change of chemical, in the process of calculation, all conditions need to be considered such as positive and negative of wind direction, equal or not equal of the transported grid, etc. Hence, there is a sample to interpret the condition of transmission concentration between source and final destination, as following: