Model study of the tropospheric chemistry of ozone

A photochemical box model with CO-CH4-NOy-H 2O chemistry is used to calculate the diurnally averaged net photochemical rate of ozone production and destruction in the troposphere. A detailed sensitivity study of these rates to the input variables (NOx, O3, CO, H2O, temperature, solar flux) is conducted. The variations of the production and destruction with altitude, latitude and season are analyzed. The magnitude of the net ozone chemical tendency (the difference between production and destruction) decreases rapidly with height to less than 2ppbv/day at 190mb mostly as a result of lower absolute humidity and temperature. Lower temperature and specific humidity are also to a large extent responsible for lower magnitudes of the net tendency at higher latitudes and in winter and they cause a shift of NOx balance point to lower NOx.;Rates of ozone production and destruction calculated with the box model for 17 latitudes, 6 months, 7 vertical levels, and for a range of mixing ratios of O3, NOx, CO and H2O are stored in look-up tables that are used to interpolate the chemistry of ozone in a single tracer simulation conducted with the 3-D Global Chemistry Transport Model (GCTM). The interpolation routine calculates the ozone destruction and production rates in every grid box and at every model time step for current O3, observed H2O, and NOx and CO from previous simulations. In the boundary layer a NOx dependent empirical relation is used to account for NMHC driven O3 production. Monthly averaged chemical ozone tendency fields show a strong correlation with the NOx fields. In contrast with the lower and middle remote troposphere where the tendencies are negative, in the upper troposphere the tendencies are generally small but positive.;Ozone sonde observations are used to evaluate the results of the ozone simulation. The agreement was found to be good for both vertical profiles and seasonal trends. Scatter plots show that 73% of the points are within +/-25% from the observed mean. The greatest discrepancies occur in lower tropical troposphere in biomass burning regions.;A set of GCTM experiments testing sensitivity of ozone in the upper troposphere to O3 rates of production and destruction are analyzed. The impact of chemistry in the upper and middle troposphere is examined by doubling ozone chemical production at the top three tropospheric levels of the model. Raising production at 190mb leads to a maximum ozone change of only 15%. Increasing production at 315 and 500mb has a stronger effect (up to 40--50%). The sensitivity of ozone and OH fields to acetone and water vapor in the upper troposphere is studied. Adding acetone increases ozone by up to 10% in the upper troposphere in the tropics, and less than 5% elsewhere.;Similar interpolation tables are used to approximate the on-line chemistry for a simulation of NOy species (NOx, HNO3 and PAN). The tables with five rates necessary to describe interconversion rates between the NOy species are pre-calculated with the box model. Relatively good agreement with observations is obtained for NOx and PAN (over 60% of the points within 50% of the observed median) when the model is run with a global NOx lightning source of 10TgN/year.