Inﬂuence of plant ecophysiology on ozone dry deposition: comparing between multiplicative and photosynthesis-based dry deposition schemes and their responses to rising CO 2 level

Abstract. Dry deposition is a key process for surface ozone (O3) removal. Stomatal resistance is a major component of O3 dry deposition, which is parameterized differently in current land surface models and chemical transport models. We developed and used a standalone terrestrial biosphere model, driven by a unified set of prescribed meteorology, to evaluate two widely used dry deposition modeling frameworks, Wesely (1989) and Zhang et al. (2003), with different configurations of stomatal resistance: 1) the default multiplicative method in each deposition scheme; 2) the traditional photosynthesis-based Farquhar-Ball-Berry (FBB) stomatal algorithm; 3) the Medlyn stomatal algorithm based on an optimization theory. We found that using the FBB stomatal approach that captures ecophysiological responses to environmental factors, especially to water stress, can generally improve the simulated dry deposition velocities compared with multiplicative schemes. The Medlyn stomatal approach produces higher stomatal conductance (reverse of stomatal resistance) than FBB and is likely to overestimate dry deposition velocities for major vegetation types, but its performance is greatly improved when spatially varying slope parameters based on annual mean precipitation are used. Large discrepancies were also found in simulated stomatal responses to rising CO2 levels, and that multiplicative stomatal method with an empirical CO2 response function produces reduction (−35 %) in global stomatal conductance, which is much larger than that with photosynthesis-based stomatal method (−14–19 %) when atmospheric CO2 level increases from 390 ppm to 550 ppm. Our results show the potential biases in O3 sink caused by errors in model structure especially in the Wesely dry deposition scheme, and the importance of using photosynthesis-based representation of stomatal resistance in dry deposition schemes under a changing climate and rising CO2 concentration.



Text S2 45
The stomatal resistance parameterization for W89 is calculated as described in Wesely (1989) and Wang et al. (1998) where G is solar radiation, Ts is surface air temperature. Di and Dv are molecular diffusivities for water and the pollutant gas respectively. 50 The stomatal resistance parameterization for Z03 is calculated as described in Zhang et al. (2003) and Zhang et al. (2002).
The expressions to calculate stomatal conductance implemented in TEMIR are also represented here.
where f(T), f(VPD) and f(ψ) are dimensionless stress functions for temperature (T), vapor pressure deficit (VPD), and water 55 stress (ψ) respectively as described in Brook et al. (1999). Gs(PAR) is the unstressed canopy stomatal conductance. Gs is calculated as weighted sum of sunlit and shaded leaves. where α is the angle between the leaf and the sun, θ is the solar zenith angle, Rdiff and Rdir are the downward visible radiation fluxes from diffuse and direct-beam radiation above the canopy.
where Tmin, Tmax, Topt are minimum, maximum and optimum temperature respectively.
where bvpd and D are vapour pressure constant and vapour pressure deficit.
For the photosynthesis-stomatal conductance module in TEMIR, we follow the description by the Community Land Model 4.5 (CLM4.5) (Oleson et al., 2013). A brief summary is also represented here. Photosynthesis in C3 and C4 plants is 80 computed as follows based on Collatz et al. (1992): The Rubisco-limited photosynthetic rate (Ac, μmol m -2 s -1 ) is: The RuBP-limited photosynthetic rate (Aj, μmol m -2 s -1 ) is: 85 The product-limited photosynthetic rate (Ap, μmol m -2 s -1 ) is: The dark respiration (Rd, μmol m -2 s -1 ), which is adjusted by the water stress factor βt, is given by: In the equations above, ci is the intercellular CO2 partial pressure (Pa). Kc and Ko are the Michaelis-Menten constants for carboxylation and oxygenation (Pa). oi is the intercellular oxygen partial pressure (Pa). Γ* is the CO2 compensation point (Pa). Vcmax is the maximum rate of carboxylation (μmol m -2 s -1 ). φ is the absorbed PAR (μmol m -2 s -1 ). J is the electron transport rate (μmol m -2 s -1 ). Tp is the triose phosphate utilization rate (μmol m -2 s -1 ), Patm is the ambient atmospheric pressure (Pa), kp is the initial slope of CO2 response curve for C4 plants (Pa / Pa). The function βt ranges from one when soil is wet 95 and to zero when soil is dry.
The stomatal conductance of water gs (μmol m -2 s -1 ) for FBB and MED is then calculated as in Eq. (4) and Eq. (5) in the main text.

100
Text S3 We use evaporative-resistance form of Penman-Monteith method to keep consistent with SynFlux stomatal conductance.
The leaf stomatal conductance is: where ε is mass ratio between water and dry air, p is air pressure, E is surface moisture flux, Tf is leaf temperature, es(Tf) is 105 the saturation vapor pressure at leaf surface. ra is aerodynamic resistance, rb,w is quasi-laminar layer resistance to water vapor. Tf is estimated as follows: where T is air temperature, H is sensitive heat, cp is specific heat of air, ρ is the mass density of air, rb,H is quasi-laminar layer resistance to heat. 110 Stomatal conductance of O3 is calculated with molecular diffusion coefficient ratio 0.6 between O3 and water vapor:    145 Figure S2. Average JJA diurnal aerodynamic resistance (Ra) and boundary layer resistance (Rb) at long-term measurement sites.