air quality model. Zhang et al. (2003) applied a two-big-leaf model that classified the
canopy as sunlit and shaded and calculated their stomatal resistances separately. This
model also parameterizes the dry deposition of gaseous pollutants at the
non-stomatal surface of the vegetation canopy and has been applied in the Canadian
air quality model.
With the development of numerical models, some complicated processes have
been added to dry deposition schemes, e.g. coupling plant physiological processes
such as plant photosynthesis into deposition parameterization. Previous studies
generally calculated canopy stomatal resistances using the Jarvis empirical equation,
which was expressed as a fitted stress equation of photosynthetically active radiation
and environmental factors including temperature, humidity and soil moisture (Jarvis
1976). Recent studies (Charusombat et al. 2010; Wu et al. 2012) attempted to
estimate the canopy stomatal resistance using the Ball-Berry semiempirical equation
(Ball et al. 1987) in which the photosynthesis rate, CO 2 concentration at the leaf
surface, leaf surface humidity and the responses of stomatal resistance to photosynthesis for C3 and C4 plants were considered in details. Such photosynthesisdeposition models require the observed or modelled photosynthetic rate of the
canopy; both need accurate laboratory measurements. Therefore, the
photosynthesis-deposition model is mostly applied to ecological sites with flux
towers (Pleim and Ran 2011).
Previous studies have found that simultaneous deposition of NH 3 and SO 2 (also
called co-deposition) over higher-humidity regions is more efficient than the deposition of NH 3 and SO 2 alone (Erisman and Wyers 1993). The SO 2 and NH 3
co-deposition phenomenon has been implemented by Erisman et al. (1994) in the
calculation of R c based on in situ observations and has also been applied to the
WRF-Chem model. Furthermore, some dry deposition models have been developed
for specific underlying surfaces. For example, the Noll and Fang scheme (Fang et al.
1999) was designed to better simulate particle deposition in urban traffic jams and
the Chamberlain leaf surface retention model (Chamberlain 2004) which is applicable to some grasslands.
4.2.3 Statistical Methods
In additional to chemical transport models, statistical models have also been applied
to estimate N deposition by fitting the key factors that influence N deposition to
observations of deposition fluxes. The factors are typically activity levels related to
Nr emissions and meteorological parameters affecting deposition velocities. An
early statistical model developed by Lin et al. (2000) estimated N deposition to
global natural ecosystems as a linear regression function of precipitation. Observational analyses of N deposition to China showed strong relationships with fertilizer
use and fuel consumption (Jia et al. 2014; Zhu et al. 2015). These two factors have
been included in recent statistical studies (e.g. Gu et al. 2015) to quantify the longterm trend of N deposition to China.
4 Modelling Atmospheric Nitrogen Deposition in China
73
canopy as sunlit and shaded and calculated their stomatal resistances separately. This
model also parameterizes the dry deposition of gaseous pollutants at the
non-stomatal surface of the vegetation canopy and has been applied in the Canadian
air quality model.
With the development of numerical models, some complicated processes have
been added to dry deposition schemes, e.g. coupling plant physiological processes
such as plant photosynthesis into deposition parameterization. Previous studies
generally calculated canopy stomatal resistances using the Jarvis empirical equation,
which was expressed as a fitted stress equation of photosynthetically active radiation
and environmental factors including temperature, humidity and soil moisture (Jarvis
1976). Recent studies (Charusombat et al. 2010; Wu et al. 2012) attempted to
estimate the canopy stomatal resistance using the Ball-Berry semiempirical equation
(Ball et al. 1987) in which the photosynthesis rate, CO 2 concentration at the leaf
surface, leaf surface humidity and the responses of stomatal resistance to photosynthesis for C3 and C4 plants were considered in details. Such photosynthesisdeposition models require the observed or modelled photosynthetic rate of the
canopy; both need accurate laboratory measurements. Therefore, the
photosynthesis-deposition model is mostly applied to ecological sites with flux
towers (Pleim and Ran 2011).
Previous studies have found that simultaneous deposition of NH 3 and SO 2 (also
called co-deposition) over higher-humidity regions is more efficient than the deposition of NH 3 and SO 2 alone (Erisman and Wyers 1993). The SO 2 and NH 3
co-deposition phenomenon has been implemented by Erisman et al. (1994) in the
calculation of R c based on in situ observations and has also been applied to the
WRF-Chem model. Furthermore, some dry deposition models have been developed
for specific underlying surfaces. For example, the Noll and Fang scheme (Fang et al.
1999) was designed to better simulate particle deposition in urban traffic jams and
the Chamberlain leaf surface retention model (Chamberlain 2004) which is applicable to some grasslands.
4.2.3 Statistical Methods
In additional to chemical transport models, statistical models have also been applied
to estimate N deposition by fitting the key factors that influence N deposition to
observations of deposition fluxes. The factors are typically activity levels related to
Nr emissions and meteorological parameters affecting deposition velocities. An
early statistical model developed by Lin et al. (2000) estimated N deposition to
global natural ecosystems as a linear regression function of precipitation. Observational analyses of N deposition to China showed strong relationships with fertilizer
use and fuel consumption (Jia et al. 2014; Zhu et al. 2015). These two factors have
been included in recent statistical studies (e.g. Gu et al. 2015) to quantify the longterm trend of N deposition to China.
4 Modelling Atmospheric Nitrogen Deposition in China
73