Predictability of Total Ozone Using a Global Three-Dimensional Chemical Transport Model Coupled with the MRI/JMA98 GCM

A global three-dimensional chemical transport model is being developed for forecasting total ozone. The model includes detailed stratospheric chemistry and transport and couples with a dynamical module of the Meteorological Research Institute/Japan Meteorological Agency 1998 (MRI/JMA98) general circulation model, which can yield realistic atmospheric fields through a meteorological assimilation system. Its predictability on total ozone is investigated for up to 4 weeks from 1997 to 2000. Global root-mean-square errors (rmses) of a control run are approximately 10 DU (3% of total ozone) throughout a year; the control run results are used as initial values for hindcast experiments. Rmses of the hindcast experiments globally range from 10 to 30 DU. The anomaly correlation between the 5-day forecasts and satellite measurements is approximately 0.6 throughout a year in the mid- and high latitudes of both the Northern and Southern Hemispheres. Thus, the model has potential for utilization on total ozone forecasts up to 5 days. In the northern mid- and high latitudes, the model produces better total ozone forecasts than the persistence up to 2 weeks, indicating that the deterministic limit of the total ozone forecasts is durationally comparable to that of weather forecasts. Good correlations between changes in total ozone and 100-hPa geopotential height reveal that the predictability of the dynamical field in the lower stratosphere critically affects the predictability of total ozone.