Simulation of stratospheric ozone in global forecast model using linear photochemistry parameterization

The two types of ozone, the simulation with interactive (prognostic) ozone using linear photochemistry parameterization (LPP) (INTR) and the simulation with non-interactive ozone using ozone climatology (CLIM), were used in the global forecast model. These two types of ozone were compared with ozone observations from the Aura Microwave Lim Sounder (MLS) and ozonesondes from 16-30 September 2008. The INTR is sensitive to LPP schemes while less sensitive to the time average of initial ozone data. Among three LPP schemes, CARIOLLE, COPCAT, and LINOZ, the COPCAT produces ozone profiles with least differences from MLS and ozonesondes. CLIM overestimates MLS at 200-20 hPa while INTR with COPCAT scheme underestimates MLS ozone above 5 hPa. Over the Antarctic in the lower stratosphere CLIM overestimates MLS and ozonesondes whereas INTR underestimates MLS but overestimates the ozonesonde data. Thus, COPCAT agrees better with ozonesonde data than any other LPP schemes and CLIM. Changing the ozone distribution from CLIM to INTR affects temperature profiles mainly through the modification of differential radiative fluxes. The correlations between ozone, differential radiative fluxes, and temperature are distinguished by altitude (or pressure levels). The correlations are strong or moderate between 3-1000 hPa (lower atmosphere) and weak above 3 hPa (upper atmosphere). This study demonstrates that the simulation of ozone using an appropriate LPP scheme is excellent in overcoming the drawbacks of using climatological ozone profiles that poorly agree with observations in extreme ozone hole events.