Previous: Model and experiments Next: Relationships to the Arctic Oscillation Up: Ext. Abst.

Results

Seasonal marches and their interannual variations

Figure 1 shows seasonal marches of zonal mean dynamical fields and ozone UV heating at several high latitudes in the Northern Hemisphere and levels. In each panel, green curve denotes the average for the ozone depletion experiment, while red curve denotes that for the control experiment. Vertical bars show standard deviations for each calendars day. It is found that zonal mean zonal winds and temperatures at 1 hPa for the both experiments show similar seasonal marches, whereas in the lower stratosphere they are largely different each other, especially for the temperature field. In the ozone depletion experiment, the temperature at 86o N and 54 hPa are kept cold well below the threshold value of the ozone loss term, 198 K, until the end of April, which is delayed by about 2 weeks comparing to the control experiment. This is connected with decreased ozone UV heating due to the Arctic ozone depletion. For a period from the beginning of the sunlit period (mid-March at 86o N) to the end of April, ozone UV heating in the ozone depletion experiment is smaller than that in the control experiment by a factor of 2. This is due mainly to the fact that the Arctic ozone losses occur almost every spring in the ozone depletion experiment.

Figure 1. (a) Time series of the simulated zonal mean zonal wind at 69o N and 1 hPa. Green, blue and red lines denote averages for "ozone depletion experiment", "D.long years" (see the text) and "control experiment", respectively. Vertical bars for the green and red curves show standard deviations. (b) Same as (a) except for the zonal mean temperature at 86o N and 11 hPa. (c) Same as (a) except for the zonal mean zonal wind at 69o N and 11 hPa. Broken line indicates 10 ms-1. (d) Same as (a) except for the zonal mean temperature at 86o N and 54 hPa. Broken line indicates 198K. (e) Same as (a) except for the vertical component of E-P flux averaged over north of 58o N at 120 hPa. (f) Same as (a) except for the zonal mean ozone UV heating at 86o N and 54 hPa.

It is also noted that the interannual variation, expressed by the vertical bars, in the ozone depletion experiment is relatively large throughout the period, which is closely connected with the Arctic ozone depletion. Figure 2 shows the interannual variations of the date of polar vortex breakdown at 11 hPa, which is defined by the date when zonal mean zonal wind becomes smaller than 10 ms-1. Green circles show the breakdown date in the ozone depletion experiment, and red ones show those in the control experiment. The occurrence of final warmings in the ozone depletion experiment widely distributes for the period from mid-March to late June. As a result, in seven years denoted by blue circles, i.e. 2, 12, 15, 31, 35, 37, 40th years, the strong polar night jet and the Arctic ozone depletion continue harmonically until June. We call these seven years as 'D.long years' hereafter.

Figure 2. Interannual variation of date of breakdown of the polar vortex at 11 hPa. Green, blue and red circles denote the ozone depletion experiment, the D.long years (later than beginning of June) and the control experiment, respectively.

The blue curve in Fig. 1 shows the average of each field in the D.long years. Seasonal marches of the polar night jet, the polar temperature and the ozone UV heating in the lower stratosphere depart from each other from the latter half of April. In that period, the vertical component of Eliassen-Palm (E-P) flux in the lower stratosphere is smaller than other time series. Both the easterly acceleration due to E-P flux divergence in the upper stratosphere and the adiabatic heating related to descending motion are small in high latitudes in the stratosphere (not shown). After the period, final warmings occur in late May in the upper stratosphere, but relatively strong westerlies, low temperature and small ozone UV heating are maintained beyond the end of June in the lower stratosphere. Moreover, the strong polar night jet and the low temperature extend downward to the upper troposphere (not shown). It is also found that decrease of UV heating due to the ozone depletion in the polar lower stratosphere is a primary cause of the thermal structure, which leads to keeping the polar night jet strong until late spring. Similar to our former results [Hirooka et al., 1999a, b], this causes strengthening and continuation of the ozone depletion itself, through chemical destruction within the polar vortex and interfering dynamical transport of ozone-rich air from low latitudes. Hence, it is considered that the dynamically calm period in late April is a precursor for these positive feedback processes well shown in this experiment.