4.2.2 Dry Deposition Parameterization
Dry deposition fluxes (F d ) of air pollutants at height z 1 near surface are calculated as
the product of the number density concentration (n) and the dry deposition velocity
(V d ):
F d ¼ n z 1
ð Þ Â V d z 1
ð Þ
ð4:4Þ
At present, most models use the resistance-in-series parameterization to calculate V d
(Wesely and Hicks 2000):
V d z 1
ð Þ ¼
1
R a z 1 ; z 0
ð
ÞþR b þ R c
ð
Þ
ð4:5Þ
Here R a (z 1 , z 0 ) is the aerodynamic resistance, considering turbulent transfer from z 1
to the zero momentum point z 0 (roughness height). R b is the quasi-laminar layer
resistance accounting for transfer from z 0 to the surface via molecular diffusion.
Generally R b is much smaller than R a and is neglected in some models. R c is the
surface resistance depending on land use type and meteorological conditions such as
temperature and solar radiation.
The surface resistance R c is related to the physical, chemical and biological
conditions of the Earth’s surface. The intricacy of R c exacerbates the difficulty of
calculations. Current models of dry deposition often lack descriptions of some
important processes in the ecosystem or oversimplify the processes, leading to
deviations in the V d calculation (Wesely and Hicks 2000; Cornell et al. 2003;
Byun and Schere 2006; Petroff et al. 2008; Wu et al. 2015; Zhang et al. 2017a).
The canopy processes in different deposition parameterizations also vary in their
choice of parameters and calculation methods (e.g. vertical variation of leaf area
index, canopy resistance coefficient, wind profile attenuation coefficient, higherorder closure coefficient, etc.), causing large uncertainty in the simulation results
(Baldocchi et al. 1987; Droppo 2006). Recent improvements in dry deposition
parameterization mainly focused on improving the calculation of R c .
Table 4.1 summarizes the dry deposition schemes used in models. Early schemes,
e.g. the Sehmel and Hodgson scheme (Sehmel 1980), usually parameterize dry
deposition by fitting meteorological conditions without considering land surface
properties. These models are inapplicable for complex surfaces such as urban area
and uneven vegetation. The Slinn scheme (Slinn 1982) parameterizes the effects of
the vegetation canopy by considering it as a collection cylinder with a certain
collection efficiency and has been applied to simple ecosystem models and environmental assessment models. However, the Slinn scheme only considers canopy
turbulence and gravity deposition processes and ignores molecular diffusion and
entrapment or impact in the quasi-laminar layer. The Wesely scheme (Wesely 1989)
calculates deposition processes in the top vegetation (leaf stomata and stratum
corneum), low canopy (branch, bark) and ground (soil, litter) over different land
70
L. Zhang et al.
Dry deposition fluxes (F d ) of air pollutants at height z 1 near surface are calculated as
the product of the number density concentration (n) and the dry deposition velocity
(V d ):
F d ¼ n z 1
ð Þ Â V d z 1
ð Þ
ð4:4Þ
At present, most models use the resistance-in-series parameterization to calculate V d
(Wesely and Hicks 2000):
V d z 1
ð Þ ¼
1
R a z 1 ; z 0
ð
ÞþR b þ R c
ð
Þ
ð4:5Þ
Here R a (z 1 , z 0 ) is the aerodynamic resistance, considering turbulent transfer from z 1
to the zero momentum point z 0 (roughness height). R b is the quasi-laminar layer
resistance accounting for transfer from z 0 to the surface via molecular diffusion.
Generally R b is much smaller than R a and is neglected in some models. R c is the
surface resistance depending on land use type and meteorological conditions such as
temperature and solar radiation.
The surface resistance R c is related to the physical, chemical and biological
conditions of the Earth’s surface. The intricacy of R c exacerbates the difficulty of
calculations. Current models of dry deposition often lack descriptions of some
important processes in the ecosystem or oversimplify the processes, leading to
deviations in the V d calculation (Wesely and Hicks 2000; Cornell et al. 2003;
Byun and Schere 2006; Petroff et al. 2008; Wu et al. 2015; Zhang et al. 2017a).
The canopy processes in different deposition parameterizations also vary in their
choice of parameters and calculation methods (e.g. vertical variation of leaf area
index, canopy resistance coefficient, wind profile attenuation coefficient, higherorder closure coefficient, etc.), causing large uncertainty in the simulation results
(Baldocchi et al. 1987; Droppo 2006). Recent improvements in dry deposition
parameterization mainly focused on improving the calculation of R c .
Table 4.1 summarizes the dry deposition schemes used in models. Early schemes,
e.g. the Sehmel and Hodgson scheme (Sehmel 1980), usually parameterize dry
deposition by fitting meteorological conditions without considering land surface
properties. These models are inapplicable for complex surfaces such as urban area
and uneven vegetation. The Slinn scheme (Slinn 1982) parameterizes the effects of
the vegetation canopy by considering it as a collection cylinder with a certain
collection efficiency and has been applied to simple ecosystem models and environmental assessment models. However, the Slinn scheme only considers canopy
turbulence and gravity deposition processes and ignores molecular diffusion and
entrapment or impact in the quasi-laminar layer. The Wesely scheme (Wesely 1989)
calculates deposition processes in the top vegetation (leaf stomata and stratum
corneum), low canopy (branch, bark) and ground (soil, litter) over different land
70
L. Zhang et al.