3. Initial and Boundary conditions.



3.1 Chemical Species

The initial conditions for the concentration of chemical species are obtained from a standard model run of four years of the date Jan 1st . Results of selected species are shown in Section 5.

Boundary conditions for the chemical species can be specified as flux or mixing ratio boundary conditions. Listed in Table 8 are the specifications of upper and lower boundary conditions for the transported species, including the dry deposition velocity (wd) at the surface. The lower boundary conditions adapted in the model are those of Hauglustaine et al. (1994) but will be described in more detail here. When fluxes are specified as boundary flux conditions ( Table 8 ), they are generally based on Hough (1991). An uptake of methane by soils is assumed to be 30 Tg-CH4 yr-1, with a latitudinal distribution adopted from Fung et al. (1991). The surface emissions of these gases are estimated based on the relative contribution of each emission category listed in Table 9 , and the latitudinal distribution of each emission category and of methane soil absorption are presented in Table 10 . The boundary fluxes of each emission category expressed in units of Tg yr-1m-2 are obtained by multiplying the values listed in Table 9 with values listed in Table 10 . The latitudinal distribution of the boundary fluxes for each species are then the addition of all emission categories. For the net surface flux of methane, the soil absorption flux is subtracted out from the emission flux. The boundary fluxes are then converted to units of cm-2s-1 and used in the chemical transport equation.

Source gases whose lower boundary conditions that are currently specified as mixing ratios according to 1990 conditions are : CO2, CFC-10, CFC-11, CFC-12, CFC-113, CFC-114, CFC-115, HCFC-22, CH3CCl3, CH3Br, CHBr3, Ha-1211, Ha-1301 and HF. For those with values varying with latitude, their surface mixing ratio are specified as follows: for CO2: [356.+b(CO2)] ppmv; for CFC-10: 108*b(CFC-10) pptv; for CFC-11: 270*b(CFC-11) pptv; for CFC-12: 465*b(CFC-12) pptv; for CFC-113: 70*b(CFC-12) pptv; for CFC-114: 10*b(CFC-12) pptv; for CFC-115: 5*b(CFC-12) pptv; for HCFC-22: 106*b(CFC-12) pptv; for CH3CCl3: 153*b(CH3CCl3) pptv; for Ha-1211: 2.9*b(CFC-12) pptv; and for Ha-1301: 1.7*b(CFC-12) pptv. The values for the b coefficients as a function of latitude are listed in Table 11 . CH3Br is specified as 9 pptv in the Southern Hemisphere, and 11.5 pptv in the Northern Hemisphere.

As listed in Table 8 , dry deposition surface velocity are considered for species NOy, NOx, HNO3, N2O5, Ox, CO, H2O2, CH2O, PAN, and H2. The velocities for different surface types are as listed in Table 2 of Hauglustaine et al. (1994). To estimate thefraction of land, ocean, snow and sea ice as a function of surface temperature, the parameterization from Robock (1983) is used.

The upper boundary conditions at 120 km for the majority of chemical species are specified as zero fluxes (see Table 8 ), except for NOy, NOx, Ox, H2 and H. Downward flux of NOx and NOy at 120 km can be specified as a function of latitude and season according to Solomon et al. (1982):

eq105

except at polar night, where FluxNOx=0. In the current version of the model, NO, NOx and NOy is specified as fixed mixing ratio of 5.e-5. Likewise, Ox and O(3P) is currently specified as mixing ratio. However, they can also be specified as flux at the upper boundary according to Kasting and Roble (1981) as:

eq106

where h1/2day is the half day hour angle in units of radius as mentioned in section 2.1.1.

3.2 Temperature, wind and stream function

3.2.1 Initial and boundary conditions for temperature.

The model integration time starts at January 1st, and the steady state model calculation of the temperature distribution at January 1st is used as the initial condition of temperature (Figure 2). The temperature at and below the lower boundary of the dynamical model (0, 1 and 2 km) is specified as a function of latitude and month according to Randel (1987).

3.2.2 Boundary conditions for zonal wind, and stream function.

The zonal wind in the 2-D model is derived by vertically intergrating the thermal wind equation (Eq. 4), knowing the temperature distribution. The zonal wind at the surface needed to start the vertical integration is specified as a function of latitude and month according to Fleming et al. (1988).

The boundary condition for the stream function at 2 km is calculated interactively with the cumulative wave forcing above the boundary. According to the 'downward control principle' of Haynes et al. (1991), the mass flow across an isentropic surface is controlled exclusively by the amount of eddy forcing above that surface. As in Garcia et al. (1992), the stream function at the lower boundary ( zb=2km) is calculated from:

eq107

This expression is derived from a linearized, steady state momentum equation , where Ftot is the combination of all types of wave forcing as calculated by the model, including gravity and planetary wave forcing, tropospheric wave forcing (Ftot=FR+FG+FT...).