References

We utilize a multiphase model, CON-AIR ( Con densed Phase to Air Transfer Model), to show that the photochemistry of nitrate (NO 3 − ) in and on ice and snow surfaces, specifically the quasi-liquid layer (QLL), can account for NO x volume fluxes, concentrations, and [NO]/[NO 2 ] (γ=[NO]/[NO 2 ]) measured just above the Arctic and coastal Antarctic snowpack. Maximum gas phase NO x volume fluxes, concentrations and γ simulated for spring and summer range from 5.0×10 4 to 6.4×10 5 molecules cm −3 s −1 , 5.7×10 8 to 4.8×10 9 molecules cm −3 , and ~0.8 to 2.2, respectively, which are comparable to gas phase NO x volume fluxes, concentrations and γ measured in the field. The model incorporates the appropriate actinic solar spectrum, thereby properly weighting the different rates of photolysis of NO 3 − and NO 2 − . This is important since the immediate precursor for NO, for example, NO 2 − , absorbs at wavelengths longer than nitrate itself. Finally, one-dimensional model simulations indicate that both gas phase boundary layer NO and NO 2 exhibit a negative concentration gradient as a function of height although [NO]/[NO 2 ] are approximately constant. This gradient is primarily attributed to gas phase reactions of NO x with halogens oxides (i.e. as BrO and IO), HO x , and hydrocarbons, such as CH 3 O 2 .

If nitrate depth profiles in polar ice were preserved over time, they would provide a valuable record of global paleoatmospheres.However, physical and photochemical processing of nitrate can alter its surface and near-surface concentrations, especially at low-accumulation sites (Rothlisberger et al., 2002), and possibly compromise its isotopic signatures (Blunier et al., 2005;McCabe et al., 2005;Hastings et al., 2004).While the photochemistry of nitrate in the snowpack has significant implications for tropospheric chemistry, since its photoproducts, NO and NO 2 , are intimately linked to reactions involving ozone, hydrocarbons and halogens, this process also generates OH radicals (see below), which can oxidize organic matter within snowpacks, leading to the C. S. Boxe and A. Saiz-Lopez: Multiphase modeling of nitrate photochemistry in the QLL formation of oxidized hydrocarbons (e.g.formaldehyde, acetaldehyde, acetone) (Dominé and Shepson, 1999;Sumner and Shepson, 2002;Grannas et al., 2004).In addition, HONO has been measured in the polar regions (Zhou et al., 2001;Honrath et al., 2002;Amoroso et al., 2006;Clemitshaw, 2006), where it has also been suggested as a possible byproduct of nitrate photolysis (Zhou et al., 2001).Yet, actual HONO concentrations and its source at polar sites have been debated (Chen et al., 2004;Dibb et al., 2004;Liao et al., 2006;Jacobi et al., 2007).
It is clear that overlying boundary layer chemistry is affected by photochemistry occurring at the snowpack at polar regions.A useful tool to study specified photochemical mechanisms occurring in the snowpack is to multiphase model boundary layer chemistry linked to chemistry at the snowpack surface, which requires a physicochemical understanding of ice surfaces.The ice-air interface of solids is an area that exhibits characteristics different from those of the bulk material.This is primarily due to the fact that atoms (or molecules) at the surface only encounter bonding forces with other molecules from one side; simultaneously, there is a similar imbalance at other interfaces.Furthermore, this behavior causes the dislocation of atoms from their original locations, alterations in their associated force and energy constants, and effects on layers below the ice-air demarcation.Michael Faraday in 1850 first suggested that the ice-air interface consists of a thin wet film (Faraday, 1850), variously called the quasi-liquid layer (QLL), premelting layer, liquid-like layer, or surface melting layer, by showing "that a particle of water which could retain the liquid state whilst touching ice on only one side, could not retain the liquid if it were touched by ice on both sides" (Faraday, 1850).
The fact that the boundary between the solid and vapor phase is wetted by a thin liquid film causes the free energy of the boundary to be lower than it would be if the thin liquid film were absent (Dash et al., 1995).As a result, if the surface of ice were initially dry, then it would reduce its interfacial free energy by converting a layer (e.g. the surface) of the solid to liquid.Hence, a liquid-like layer should exist over some measurable and quantifiable temperature range on the surface of ice, below its bulk normal melting temperature.The existence of the QLL is not prohibited due to its thinness and closeness to the normal melting temperature of ice and is present at a state where the free energy of the ice system is at a minimum and is governed by the competition between the free energy of the ice surface and the energy required to melt a solid layer.
The thickness of the QLL as a function of temperature has been quantified both experimentally (Doppenschmidt and Butt, 2000;Pittenger et al., 2001;Bluhm et al., 2002;Sadtchenko and Ewing, 2002) and theoretically (Ohnesorge et al., 1994;Landa et al., 1995;Wettlaufer, 1999).With the single exception of Elbaum et al. (1993), whose experiments were done on exposed horizontal facets in the prismatic orientation 10 10, these studies have shown that the QLL in-creases with increasing temperature.Additionally, impurities enhance its thickness (Doppenschmidt and Butt, 2000;Wettlaufer, 1999).The addition of impurities at constant pressure will shift the normal melting point of the bulk solid, which is directly dependent on the concentration of the impurity.As the melting point is approached, the QLL appears to be indistinguishable from the liquid phase in its uppermost layers.Concurrently, we do acknowledge that, at specified temperature regimes, below the actual melting point of pure water ice, the QLL is distinctly dissimilar than pure liquid water (e.g. its ability to take up trace gases).For instance, McNeill et al. (2006) showed for the first time that the solubility of HCl in the QLL, rather exhibiting a solubility similar that in a true-liquid matrix, exhibits a solubility that is intermediate between that in bulk ice and its respective solubility in a true-liquid matrix.
The QLL can play a pivotal role in environmental phenomena such as 1) controlling the friction of ice and snow; 2) soil freezing, permafrost formation, and frost heave; 3) sintering and sliding in glaciers, sea-ice, and snow fields; and 4) behavior of atmospheric ice (Dash et al., 1995).The QLL has also been suggested to contribute to the electrification of thunder clouds via charge transfer at the liquid-ice interface (Baker and Dash, 1994).Abbatt et al. (1992) and Molina (1994) even proposed that polar stratospheric clouds are able to accommodate HCl by dissolution in multilayerthick quasi-liquid films, where they can efficiently participate in ozone destruction during winter and spring months in Antarctica and the Arctic.These hypotheses were later confirmed by seminal work, via laboratory analyses, showing explicitly that trace gases do efficiently accommodate snow/ice surfaces through trace-gas induce QLL formation McNeill et al. (2006).
In addition, field measurements of NO x range from ∼5.7×10 8 to ∼2.9×10 9 molecules cm −3 and exhibit Honrath et al., 1999Honrath et al., , 2002;;Jones et al., 2000;Beine et al., 2002;Dibb et al., 2002;Simpson et al., 2007).In this study, we use CON-AIR to show that nitrate photochemistry in the QLL does simulate well NO x volume fluxes, concentrations, and γ measured just above the snowpack (i.e. at ∼25 cm) at various sites in the Arctic and coastal Antarctica.The implications of these findings are also discussed.

Model description
CON-AIR is a multiphase model that treats the interaction of gas phase boundary layer chemistry with condensed phase chemistry and photochemistry in and on snow and ice surfaces, specifically the QLL.As described previously, here the QLL is defined as a thin layer on the surface of snow and ice, where water molecules are not in a rigid solid structure, yet not in the random order of a liquid (Petrenko and Whitworth, 1999), which, in our model, is the demarcation between the vapor and bulk ice phase.It is structured in two main components: i) condensed phase chemistry and photochemistry regime in the QLL; and ii) gas phase chemistry scheme comprising photochemical, thermal, and heterogeneous reactions.
The exchange of nitrogen species between the QLL and the atmosphere depends on the respective Henry's law constants of species including NO and NO 2 .The Henry's law solubility constants and temperature dependences for the gas phase equilibrating species NO and NO 2 are 1.9×10 −3 ×e (1500(1/T −1/T o)) M atm −1 and 6.4×10 −3 e (2500(1/T −1/T o)) M atm −1 , respectively (Schwartz and White, 1981;Lelieveld and Crutzen, 1991).The temperature dependence of the solubility of species is taken into account by including a diurnal variation of the typical temperature profile of both the Arctic and coastal Antarctic region during spring and summertime (i.e.250≤T/K≤265).A description of the radiation and gas phase scheme, and a complete set of all gas phase reactions employed in the model are summarized in Table 1 of the supplementary material (http://www.atmos-chem-phys.net/8/4855/2008/acp-8-4855-2008-supplement.pdf).

Condensed phase scheme and QLL parameterizations
We note that the following formulation is developed within the context of a current overall shortage of physico-chemical data pertinent to the uptake and release of trace gases to snow/ice at conditions relevant for the polar snowpack.Bulk concentrations ofNO − 3 and NO − 2 (i.e. at the top few centimeters) at the Arctic and coastal Antarctic snowpack are 1≤[NO − 3 ]/µM≤17 and ∼1 nM, respectively (Stotlemyer and Toczydlowski, 1990;Jaffe and Zukowski, 1993;Li, 1993;Silvente and Legrand, 1995;De Angelis and Legrand, 1995;Dibb et al., 1998;Jones et al., 2007).A number of laboratory experiments have provided evidence that the photolysis of nitrate transpires in the QLL on the surface of ice crystals (Dubowski et al., 2001(Dubowski et al., , 2002;;Boxe et al., 2003;Chu and Anastasio, 2003).In this study, we restrict our model simulations within the context that all condensed phase reactions take place in the much smaller volume of the QLL.Typical bulk concentrations of NO − 3 and NO − 2 measured in the Arctic and coastal Antarctic snowpack were re-quantified following the formulation established by Cho et al. (2002).Cho et al. (2002) derived the following equation which relates the fraction of liquid water (ψ H 2 O ) as a function of temperature (T ) and the total solute concentration in the QLL (C 0 T ).ψ H 2 O (T ) is the fraction of water in the QLL as a function of temperature, m H 2 O is the molecular weight of water (18.01 g/mole), R is the gas constant (8.314×10 −3 kJ/K mole), H 0 f is enthalpy of fusion of water (6 kJ/mole), and T f is the freezing temperature of water (273.15K).Assuming that the total initial concentrations of NO − 3 and NO − 2 reside in the QLL, we relate their respective bulk concentrations (C bulk ) to their respective concentrations in the QLL via Eq.( 2): Substituting Eq. (1) into Eq.( 2), yields the following: Then, given the upper limit C bulk−upper−limit (=17.0001µM, , we calculate 4.54×10 −5 and 1.01×10 −5 as the mean of the upper and lower limit ψ H 2 O from 250 to 265 K, respectively, by using Eqs.( 4) and ( 5): Summing the mean of ψ H 2 O−upper−limit and themean of ψ H 2 O−lower−limit gives 5.55×10 −5 .In CON-AIR, as an approximation, we incorporate the average of this sum, 2.78×10 −5 , as the fraction of liquid water, representative for temperatures from 250 to 265 K. Taking the mean of the median of [NO − 3 ] o found both in the Arctic (i.e. from 3 to 17 µM) and coastal Antarctic (i.e. from 1 to 9 µM) yields 7.5 µM.Then, as an estimation, we take o in the QLL is 270 mM and 0.04 mM, respectively, which we incorporate in CON-AIR as their initial concentrations.
As an approximation, we use a use a conservative snow depth of 1 cm and a snow density of 0.31 g cm −3 (Michalowski et al., 2000;Sumner and Shepson, 1999).As a a Aqueous phase reaction rate constants were obtained from Mack and Bolton (1999).b QLL rate reaction rate constants were quantified by including the "volumetric" factor (Grannas et al., 2007;Takenaka et al., 1996).(Grannas et al., 2007;Takenaka et al., 1996).result, the total potential liquid content in a snow column of 1 cm 2 cross-sectional area of snowpack is: The estimated fraction of liquid water is 2.78×10 −5 ; therefore, the QLL volume at the snowpack surface: QLL volume = 0.31 cm 3 cm −2 × 2.78 × 10 −5 = 8.6 × 10 −6 cm 3 cm −2 (7) To properly express aqueous phase reaction rates to QLL reaction rates, a volumetric factor (volumetric) was estimated based on laboratory derived reaction rate enhancement factors.A volumetric factor was quantified by taking the average of the upper limit reaction rate enhancement factors obtained in the laboratory by Grannas et al. (2007) and Takenaka et al. (1996), 40 and 2.4×10 3 , respectively, yielding Therefore, the reaction rates are quantified by incorporating volumetric factor, volumetric.The rate constants for reactions taking place in the QLL are: where k are the actual literature aqueous phase rate constants in units of cm 3 molecule −1 s −1 and cm 6 molecule −2 s −1 , for second-and third-order rate constants, respectively.Table 1 lists the major reactions pertaining to nitrate photochemistry, their condensed phase reaction rates, and their QLL reaction rates.
The rate constant for the transfer of species from the QLL to the gas phase is calculated using an approximation of the first order rate constant, k t =1.25×10 −5 s −1 (Gong et al., 1997;Michalowski et al., 2000).
9.31 × 10 −6 cm 3 (QLL) 10, 000 cm 3 (atmosphere) Nevertheless, the rate of transfer of species will depend on the concentration and Henry's law constants for solubility of the corresponding species.Hence, the complete expression for the phase equilibration of species from the QLL to the atmosphere is: where H is the dimensionless Henry's law constant.H is defined as H =(H RT ), where H is a species' Henry's law constant, R is the gas constant, 0.082058 L atm K −1 mol −1 , and T is the temperature (K).

Results and discussion
The photochemistry of nitrate in the aqueous phase has been studied extensively (Mark et al., 1996;Mack and Bolton, 1999).Dissolved nitrate has two primary absorption bands in the ultraviolet (UV).The first occurs in the far UV via the strong π→π * transition, centered at 201 nm (ε max =9500 M −1 cm −1 ), and the second is a weaker absorption band that occurs via the highly forbidden n→π * transition, centered at 302 nm (ε max =7.14 M −1 cm −1 ).Furthermore, it was proposed that the weaker absorption band may occur from the combination of a singlet and triplet n→π * and σ →π * transition (Maria et al., 1973).Mack and Bolton (1999) showed that the overall stoichiometry for nitrate irradiation is In the absence of .OH scavengers this stoichiometry is maintained over the entire pH range (Wagner and Strehlow, 1980).For λ<280 nm, the major reaction pathway is through isomerization of [NO − 3 ] * , generated via Reaction (R2), to form ONOO − , peroxynitrite, and at low pH, peroxynitrous acid, HONOO (Reaction R3).HONOO can also be produced from the recombination of .OH and NO 2 within a solvent cage as shown in Reaction (R4).HONOO isomerizes rapidly back to NO − 3 (Reaction R5) (Mack and Bolton, 1999).
Yet, in the troposphere, all λ<290 nm is completely attenuated by stratospheric ozone.Therefore, λ≥290 nm are pertinent for this study.In aqueous solutions at pH<6 and λ≥290 nm, nitrate photolysis proceeds via two primary photolytic pathways as illustrated in Reactions (R6) and (R7), through the generation of nitrate in the excited state, [NO − 3 ] * , from Reaction (R2).As shown in Reaction (R8), O − reacts rapidly with water to form the hydroxyl radical.
The UV absorption spectrum of nitrite displays three absorption bands: the first involves a π→π * transition with maxima at 220 nm, and the latter two peaks are maxima at 318 nm (ε max =10.90 M −1 cm −1 ) and 354 nm (ε max =22.90 M −1 cm −1 ), both corresponding to n→π * transitions.Similar to nitrate, nitrite undergoes direct photolysis as shown in Reaction (R12) to produce NO, and it also oxidizes by reaction with .OH via Reaction (R13).
The photolysis of NO 2 also produces NO (Reaction R14).
We exclude this reaction from the chemical scheme used in CON-AIR since its photolytic lifetime during midday spring and summertime (1/J NO 2 =1/4.6×10−3 s −1 ≈ 217 s ) (Yung and DeMore, 1999) is longer than its diffusion lifetime through a ∼300 nm thick QLL (see above calculation).As calculated using the diffusion length equation (Eq.13) (Dubowski et al., 2001): L is the thickness of the QLL, D is the diffusion coefficient of NO 2 (Dubowski et al., 2001), and τ is the time it takes NO 2 to diffuse through a thickness L. Using Eq. ( 13), we show that the maximum snowpack depth, starting from the top of the ice surface, where NO 2 photolysis will not occur is ≈(9.8×10−9 cm 2 s −1 ×217 s) 1/2 ≈15 µm.Consequently, below 15 µm NO 2 will undergo photolysis to produce NO, supporting the exclusion of this photolytic pathway.This estimation is further supported by previous findings, which have shown that NO 2 produced from nitrate photolysis in the outermost layers of thin ice films are readily released to the gas phase, compared to NO 2 formed at deeper depths, which undergoes further chemical and photolytic processing (Dubowski et al., 2001;Boxe et al., 2005Boxe et al., , 2006)).Finally, the photoproduced NO and NO 2 are readily released to the gas phase after equilibration due to their low solubility (Reaction R15).
NO 2(QLL) , NO (QLL) The protonation of nitrite to form nitrous acid (HONO (aq) ) (Reaction R16) was also not considered in the QLL reaction mechanism since model simulations yielded γ ∼1500, much larger than any reported measurements from field studies (e.g.γ ∼0.8 to ∼2.0) (Honrath et al., 1999(Honrath et al., , 2002;;Jones et al., 2000Jones et al., , 2007;;Beine et al., 2002;Dibb et al., 2002;Simpson et al., 2007).This result implies that a significant amount of HONO produced in the snowpack may be retained by matrix or solvent cage effects or may be dependent on photosensitized organic compounds, such as possible reaction cycles that may efficiently transfer electrons to NO 2 , possibly leading to the production of HONO (Beine et al., 2006).Presently, the mechanism of HONO formation from NO − 3 is not well known.
A simplified scheme illustrating the primary reactions governing NO x release from the QLL film to the gas phase from nitrate photochemistry used in CON-AIR is shown in Fig. 1.Laboratory studies have shown that the photochemistry of nitrate in ice is analogous to its aqueous phase photochemistry (Dubowski et al., 2001(Dubowski et al., , 2002;;Chu and Anastasio et al., 2003;Boxe et al., 2006) Therefore, as shown in Table 1, QLL reaction rates were quantified by scaling aqueous phase reaction rates according to the micro-scopic dimensions of the QLL.The absorption spectrum for NO 3 -.
The absorption spectrum for NO 2 -.
Normalized solar spectrum at Earth's surface.
Fig. 2. The absorption spectrum for NO − 3 and NO − 2 (Gaffney et al., 1992;Zuo and Deng, 1998) and the normalized solar spectrum at the Earth's surface from 290 to 400 nm.
Laboratory studies have shown that nitrate is a source of NO and NO 2 from ice surfaces (Honrath et al., 2000b;Dubowski et al., 2001Dubowski et al., , 2002;;Chu and Anastasio, 2003;Boxe et al., 2003Boxe et al., , 2005Boxe et al., , 2006;;Jacobi et al., 2006;Jacobi and Hilker, 2007).Only a small number of laboratory investigations of nitrate photochemistry in ice were carried out to correlate their respective NO and NO 2 fluxes with field measurements (Boxe et al., 2003(Boxe et al., , 2006).Yet, these studies were restricted by high detection limits for NO and NO 2 and the use of irradiation sources emitting at 313±20 nm (i.e.overlapping the absorption spectrum of nitrate), resulting in higher NO x concentrations than measured in the field and much lower γ (e.g.0.043 to 0.0005) than those measured over the Arctic and Antarctic snowpack (Boxe et al., 2003(Boxe et al., , 2006)).Compared to the typical initial nitrate concentrations (1 to 20 µM) and the typical actinic flux spectrum at Earth's surface for the Arctic and coastal Antarctic regions, the higher initial nitrate concentrations (50 mM) and the dissimilar actinic flux spectrum used, likely contributed to the disparity between these laboratory results and those from the field.Figure 2 illustrates this disparity to some extent by comparing the absorption spectrum for nitrate and nitrite and the actinic flux spectrum at the Earth's surface.The surface irradiance is computed using a 2-stream radiative transfer code (Thompson, 1984).We calculate the diurnal variation of J NO − 3 and J NO − 2 for snowpack summer and springtime conditions by extrapolating laboratory modeled and measured J NO − 3 and J NO − 2 for ice, snowpack, and seawater (Zuo andDeng, 1998, Qiu et al., 2002;King et al., 2005) to the radiative transfer code, coupled to CON-AIR, such that J NO − 3 and J NO − 2 vary as a function of solar zenith angle (or as a function of time of day), therefore providing a more complete representation of nitrate photochemistry.Note, the smaller and larger summer/springtime diurnal profiles of J NO − the lower and upper limits for the J NO − 3 values obtained for surface snow and sea-ice (Qiu et al., 2002;King et al., 2005), while the summer/springtime diurnal profile of J NO − 2 was derived from extrapolating the J NO − 2 value obtained for surface seawater (Zuo and Deng, 1998).Figure 3 illustrates a typical diurnal profile for NO (e.g.maximum volume fluxes of 2.3×10 4 molecules cm −3 s −1 during spring and 3.2×10 5 molecules cm −3 s −1 during summer) and NO 2 (e.g.maximum concentrations of 1.2×10 4 molecules cm −3 s −1 to 2.7×10 4 molecules cm −3 s −1 during spring and 1.8×10 5 to 3.2×10 5 molecules cm −3 s −1 during summer) over the Arctic and Antarctic snowpack.These simulated NO x volume fluxes are comparable to field measurements of Jones et al. (2007).Assuming a ∼100 m boundary height and taking the median of the concentration of molecules between 250 and 265 K at atmospheric surface pressure (1 atm or 1.01325×10 5 N m −2 ) (2.86×10 19 molecules cm −3 ), simulated maximum concentrations of NO x , ∼5.7×10 8 to ∼4.8×10 9 molecules cm −3 , agree well with maximum concentrations of NO x measured just above the snowpack by field measurements, ∼5.7×10 8 to ∼2.9×10 9 molecules cm −3 , (Honrath et al., 1999(Honrath et al., , 2002;;Jones et al., 2000;Beine et al., 2002;Dibb et al., 2002;Simpson et al., 2007).It also accounts for the range of γ measured during Arctic and Antarctic summer and springtime, where springtime maximum γ ranges from ∼0.84 to ∼1.86 and summertime maximum γ ranges from ∼0.50 to ∼2.20, which is also in good accord with measured γ over the snowpack (Honrath et al., 1999(Honrath et al., , 2002;;Jones et al., 2000;Beine et al., 2002;Dibb et al., 2002;Simpson et al., 2007).Furthermore, these model results reinforce laboratory and snow chamber results showing that the major source of NO release from snow/ice surfaces isNO − 2 , its immediate photolytic precursor that absorbs at wavelengths longer than nitrate itself (Cotter et al., 2003;Boxe et al., 2006), as shown in Fig. 2. Thus, incorporating the actinic flux at the Earth's surface shows that nitrite is more photolabile than nitrate (Cotter et al., 2003).
Furthermore, we investigate the profile of gas phase boundary layer NO and NO 2 as a function of height up to 20 m during the summertime over the snowpack using a 1-D model (Saiz-Lopez et al., 2008).Figure 4 shows that the model predicts a slight negative gradient for both [NO] and [NO 2 ], and γ remains approximately constant.The gradient is the result of gas phase reactions of NO x with halogens oxides (i.e.BrO and IO), HO x , and hydrocarbons (e.g.CH 3 O 2 ) (Saiz-Lopez et al., 2008).Atmospheric stability and wind speed may also affect the concentration gradient of NO x above the snowpack (Beine et al., 2002).However, constraining the 1-D model with the lower limit summertime NO and NO 2 volume fluxes derived from CON-AIR leads to good agreement with recent summertime observations of NO and NO 2 concentrations (13 ppt and 7 ppt as average noon values) and ratios (([NO]/[NO 2 ]∼1.8) obtained at a few meters above the coastal Antarctic snowpack (e.g.Jones et al., 2007) In addition, our model simulations suggest that, in general, nitrite photolysis (predominantly produced from nitrate photodecomposition) governs the release of NO x just above the Arctic and coastal Antarctic snowpack, which is controlled by nitrite's coincident absorption spectrum with the solar spectrum at the polar snowpack surface.Finally, our model analyses show that NO and NO 2 display a negative concentration gradient as a function of height although their concentration ratios remain constant.We attribute this effect to gas phase reactions of these species with halogen oxides, HO x , and hydrocarbons.

Fig. 1 .
Fig.1.Simplified schematic diagram illustrating the primary reactions governing NO x release from a 300 nm thick QLL film to the gas phase from nitrate photochemistry.At QLL depths <15 µm, NO 2 photolysis does not occur, while at QLL depths >15 µm NO 2 photolysis occurs.

Table 1 .
QLL reactions and rate constants.
. www.atmos-chem-phys.net/8/4855/2008/Atmos.Chem.Phys., 8, 4855-4864, 2008 We use a novel multiphase model, CON-AIR (Condensed Phase to Air Transfer Model) to show that the photochemistry of nitrate in and on snow/ice surfaces (i.e. the QLL) can account for measured NO and NO 2 volume fluxes, concentrations, and [NO]/[NO 2 ] measured just above the Artic and coastal Antarctic snowpack.Our model results produce comparable results although polar snowpack site physicochemical properties are dynamic and specific in nature.