[AAS07-P12] ODS and GHG dependence of total ozone at mid- and high latitudes indicated by multi-ensemble simulations using MIROC3.2 and MIROC5 chemistry-climate models
キーワード：オゾン、オゾン破壊物質、温室効果ガス、化学気候モデル、アンサンブル実験、極域
Total ozone in the Arctic shows large interannual variations. The total ozone amount during Arctic spring is larger when the meridional circulation in the wintertime is stronger, thus when the Arctic polar vortex is weaker, and vice-versa. These variations mask the dependence of ozone concentration on ozone depleting substances (ODSs) and greenhouse gases (GHGs). Therefore, it is difficult to evaluate and isolate the effects of ODS regulations and GHG emissions from natural atmospheric variability, although ODSs and GHGs may influence the interannual variation of the atmosphere.
We conducted multi-ensemble simulations using two chemistry-climate models (CCMs). We considered the variability of interannual ozone variations to be a statistical variance. We investigated the ODS and GHG dependence of total ozone for extreme values of the variance, as well as for the mean of all ensemble members. In these simulations, the ODS and GHG concentrations were fixed at observed or projected values for specific years in the past and future (e.g., for the years 1960, 2000, 2050). Each simulation or experiment is uniquely defined by ODS and GHG levels corresponding to a specific year (for example, observed ODS levels of 1980 and projected GHG levels for 2050). For one experiment, the ozone variability among the ensemble members is then considered to be the interannual variation. We examined the ODS and GHG dependence of total ozone (i.e., as a function of the experiment), based on the minimum value found for each ensemble member at latitudes 45-90°N (‘polar cap’) during Northern spring (March – May). From the single-member minimum values, we then calculated the multi-ensemble mean, the mean for the 5 ensemble members with largest minimum value (hereafter “highest 5 members”), and for the 5 ensemble members with the smallest minima (hereafter “lowest 5 members”). The results show that the ODS and GHG dependence is large for the lowest 5 ensemble members and small for the highest 5 ensemble members. For the lowest 5 members, ozone levels should be smaller in a high-ODS atmosphere and larger in a high-GHG atmosphere. Furthermore, the zonal-mean zonal wind is stronger than for the multi-ensemble mean, while the corresponding polar cap temperature in the lower stratosphere is lower than the mean. This indicates that the Arctic polar vortex is comparatively stronger in the lowest 5 members.
The dependence of total ozone between 45-90°S in September – November (Southern spring) was also investigated. We found much smaller GHG dependence and much larger ODS dependence than for the Northern Hemisphere. The differences in the ODS and GHG dependence were much smaller among the multi-ensemble mean, the highest 5 members, and the lowest 5 members. This is consistent with much smaller interannual variations in the Southern Hemisphere and, hence, more influence of atmospheric chemistry on the ozone variations.
We conducted these analyses for MIROC3.2-CCM and MIROC5-CCM and compared the results of the two models. The difference in the ODS and GHG dependence of total ozone between these two models is consistent with their transport properties.
We conducted multi-ensemble simulations using two chemistry-climate models (CCMs). We considered the variability of interannual ozone variations to be a statistical variance. We investigated the ODS and GHG dependence of total ozone for extreme values of the variance, as well as for the mean of all ensemble members. In these simulations, the ODS and GHG concentrations were fixed at observed or projected values for specific years in the past and future (e.g., for the years 1960, 2000, 2050). Each simulation or experiment is uniquely defined by ODS and GHG levels corresponding to a specific year (for example, observed ODS levels of 1980 and projected GHG levels for 2050). For one experiment, the ozone variability among the ensemble members is then considered to be the interannual variation. We examined the ODS and GHG dependence of total ozone (i.e., as a function of the experiment), based on the minimum value found for each ensemble member at latitudes 45-90°N (‘polar cap’) during Northern spring (March – May). From the single-member minimum values, we then calculated the multi-ensemble mean, the mean for the 5 ensemble members with largest minimum value (hereafter “highest 5 members”), and for the 5 ensemble members with the smallest minima (hereafter “lowest 5 members”). The results show that the ODS and GHG dependence is large for the lowest 5 ensemble members and small for the highest 5 ensemble members. For the lowest 5 members, ozone levels should be smaller in a high-ODS atmosphere and larger in a high-GHG atmosphere. Furthermore, the zonal-mean zonal wind is stronger than for the multi-ensemble mean, while the corresponding polar cap temperature in the lower stratosphere is lower than the mean. This indicates that the Arctic polar vortex is comparatively stronger in the lowest 5 members.
The dependence of total ozone between 45-90°S in September – November (Southern spring) was also investigated. We found much smaller GHG dependence and much larger ODS dependence than for the Northern Hemisphere. The differences in the ODS and GHG dependence were much smaller among the multi-ensemble mean, the highest 5 members, and the lowest 5 members. This is consistent with much smaller interannual variations in the Southern Hemisphere and, hence, more influence of atmospheric chemistry on the ozone variations.
We conducted these analyses for MIROC3.2-CCM and MIROC5-CCM and compared the results of the two models. The difference in the ODS and GHG dependence of total ozone between these two models is consistent with their transport properties.