The 11-level Geophysical Fluid Dynamics Laboratory (GFDL) global chemical transport model has been used to assess the impact of stratospheric NOx production on tropospheric reactive nitrogen (NOy) concentrtions. A temporally varying source function was constructed using specified two-dimensional, monthly average O3, N2O, temperature, and surface pressure data generated by the GFDL ''SKYHI'' model. The calculated yearly NOy production rate is 0.64 Tg N (0.64¿1012 g N). A wet removal scheme, which distinguishes between stable and convective rain based on the bulk Richardson number, is introduced. Simulations have been performed with a simplified chemical mechanism which fractionates NOy into soluble and insoluble species. The role of peroxyacetyl nitrate (PAN) in determining the impact of stratospheric injection on the tropospheric NOy budget is studied by comparing results of simulations with and without PAN chemistry. We conclude that (1) the stratospheric source is too small to account for background surface NOy concentrations observed in the remote (i.e., regions a few thousand kilometers from continental source regions) troposphere. Surface NOy mixing ratios seldom exceed 10 parts per trillion by volume (pptv) in the model northern hemisphere and are always below 20 pptv. Together, fossil fuel combustion emissions and stratospheric injection account for less than 10% of observed surface nitrate concentrations in the remote tropical Pacific. (2) The impact of the stratospheric source is comparable to that of the fossil fuel combustion source in terms of NOy mixing ratios in the northern hemisphere at the 500 mbar model level and is more important in the middle and high latitudes of the southern hemisphere.

At the 315 mbar model level the stratospheric source contribution to NOy levels is more important than that of the fossil fuel source at all latitudes, except in the tropics. However, substantial contributions from other NOy sources are needed to explain observations in the remote middle and upper troposphere. (3) Inclusion of PAN chemistry has the effect of increasing model-calculated surface NOy mixing ratios in the northern hemisphere middle and high latitudes by factors of 1.5--3 during winter/spring and by factors of 2--4 during summer/fall. Surface NOy mixing ratios in the southern hemisphere show a smaller increase due to slower rates of PAN formation. This is a direct result of lower hydrocarbon concentrations in the southern hemisphere. ¿American Geophysical Union 1991