Future changes in the stratosphere-to-troposphere ozone mass flux and the contribution 1 from climate change and ozone recovery

10 Model simulations consistently project an increase in the stratosphere-troposphere exchange 11 (STE) of ozone in the future. Both, a strengthened circulation and ozone recovery in the 12 stratosphere contribute to the increased mass flux. In our study, we investigate with a state-of- 13 the-art chemistry-climate model the drivers of future STE change as well as the change in the 14 distribution of stratospheric ozone in the troposphere. Our focus is on the investigation of the 15 changes on the monthly scale. The global mean influx of stratospheric ozone into the 16 troposphere is projected to increase between the years 2000 and 2100 by 53% under the RCP8.5 17 greenhouse gas scenario. We find the largest increase of STE in the NH in June due to 18 increasing greenhouse gas (GHG) concentrations. In the southern hemisphere (SH) the GHG 19 effect is dominating in the winter months, while decreasing levels of ozone depleting substances 20 (ODS) and increasing GHG concentrations contribute nearly equally to the increase in SH 21 summer. A large ODS-related ozone increase in the SH stratosphere leads to a change in the 22 seasonal breathing term which results in a future decrease of the ozone mass flux into the 23 troposphere in the SH in September and October. We find that the GHG effect on the STE 24 2 change is due to circulation and stratospheric ozone changes, whereas the ODS effect is 25 dominated by the increased ozone abundance in the stratosphere. The resulting distributions of 26 stratospheric ozone in the troposphere for the GHG and ODS changes differ because of the 27 different regions of ozone input (GHG: midlatitudes; ODS: high latitudes) and the larger 28 increase of tropospheric ozone loss rates due to GHG increase. Thus, the model simulations 29 indicate that stratospheric ozone is more efficiently mixed to lower levels if only ODS levels 30 are changed. The increase of the stratospheric ozone column in the troposphere explains more 31 than 80 % of the tropospheric ozone trend in NH spring and in the SH except for the summer 32 months. The importance of the future stratospheric ozone contribution to tropospheric ozone 33 burdens therefore depends on the season. with a chemistry-climate model (CCM) under the most extreme RCP8.5 scenario for the annual 105 and monthly means. We identify the changes in the seasonal cycle of STE due to the projected 106 increase in GHGs and decline in ODS, i.e. the associated stratospheric ozone recovery. the future, the ozone mass flux is clearly larger in the timeslice simulations than in the 345 the upper and middle troposphere, the regions where cross-tropopause changes in stratosphere-to-troposphere transport of 511 ozone in timeslice and transient simulations with the CCM EMAC to address the questions 512

change is due to circulation and stratospheric ozone changes, whereas the ODS effect is  Ozone (O3) in the troposphere has two sources: photochemical production involving ozone 36 precursor species such as nitrogen oxides (NOx), carbon monoxide (CO) and hydrocarbons 37 (e.g., methane (CH4)) and the transport of ozone from the stratosphere into the troposphere (i.e.  In addition to a changing amount of stratospheric ozone in the troposphere, changing future 85 emissions of ozone precursor species will affect the local ozone production in the troposphere. showed also the annual cycle of the ozone mass flux derived from a boxmodel approach 95 introduced by Appenzeller et al. (1996). In their model simulation, the maximum ozone flux 96 occurs in spring in the SH and Northern Hemisphere (NH) for the 1960 to 1970 mean. In the 97 future (2090-2100), the peak is shifted towards late spring/early summer in the NH and towards 98 winter in the SH. As Roelofs and Lelieveld (1997) reported, the seasonal timing of the input of 99 stratospheric ozone into the troposphere is relevant for potential mixing of stratospheric ozone 100 towards the surface, since in summer the ozone loss rate is larger than in winter. This means Furthermore, we analyse the resulting changes in the distribution of stratospheric ozone in the 108 troposphere, using comprehensive stratospheric and tropospheric chemistry and therefore 109 considering the full range of changes in chemical loss and production caused by GHG or ODS  In this study we want to address the following research questions: 114 (1) How will the stratosphere-to-troposphere ozone mass flux change in the future? 115 (2) What are the major drivers of the future changes in the stratosphere-to-troposphere ozone 116 mass flux? 117 (3) Will the seasonality of the STE change in future?

118
(4) How will the GHG emission scenarios affect the ozone mass flux into the troposphere?

119
(5) How is the ratio of stratospheric ozone in the troposphere changed in the future?

120
The study is structured as followed: First the model and the experimental set-up used for the 121 simulations are described as well as the methodology for calculating the ozone mass flux from 122 the stratosphere to the troposphere (Section 2). In Section 3 we show the climatological mean   An overview of the boundary conditions in the four simulations is given in Table 1.  Table 1.

173
More detailed information of this simulation can be found in Jöckel et al. (2016). It has to be 174 noted, that the stratospheric ozone loss in the past is underestimated in this simulation, which 175 affects the trends in ozone mass flux.

177
To quantify the net ozone mass flux from the stratosphere into the troposphere we apply the 178 boxmodel approach described by Appenzeller et al. (1996). They described the hemispheric net 179 mass transport in a simple model which consists of three regions (i.e. boxes), the troposphere,  It has to be noted that with this methodology it is not possible to study the transport pathways.    NH and SH respectively. The ozone mass flux into the NH is larger than into the SH and has 246 its peak in early summer, whereas in the SH the maximum ozone mass flux is found in spring.

247
The annual cycle in the EMAC REF2000 simulation is comparable to the results of Hegglin influx from the stratosphere is smaller than in summer.

278
In the next section we analyse the abundance of stratospheric ozone in the troposphere for June    Table   358 1).

359
The    Table 1) leading to different tropopause heights and therefore to in ozone mass flux (see Table 2 and 3).

541
The future spatial distribution of the tropospheric O3s column in the troposphere is determined higher than in winter and mixing is less efficient anyway.

572
In summary, this study shows that GHG and ODS changes have different effects on the future 573 ozone mass flux, the seasonality and the resulting abundances of stratospheric ozone in the 574 troposphere. Moreover, it shows that both forcings are projected to cause an increased amount 575 of stratospheric ozone in the troposphere, which will not only contribute to the radiative forcing 576 and global warming but will also affect the air quality at the surface.