Title: Evolution of stratospheric ozone in the mid-latitudes in connection with the abundances of halogen compounds
Abstract: This thesis addresses the issue of the long-term evolution of stratospheric ozone in relation to the halogen loading. To that aim, long-term records of satellite and ground-based (GB) ozone profile measurements at six lidar stations, of the Network for the Detection of Stratospheric Change, are examined to find the bias and drift in the measurements. The stratospheric ozone trends are then estimated from the ozone profile and total column measurements using the Equivalent Effective Stratospheric Chlorine time series and two linear trend functions (before and after 1997) called as piecewise linear trends (PWLTs), to account for the change in the trends of ozone depleting substances, at Northern midlatitude stations. The analysis uses GB measurements from lidar, Umkehr, ozonesondes and the Dobson and SAOZ spectrometers, and satellite observations from SBUV(/2), SAGE II, HALOE, UARS MLS, Aura MLS and GOMOS. First of all, a sensitivity analysis is performed to diagnose the effect of using different ozone absorption cross-section data sets (Bass and Paur and Brion-Daumont-Malicet) on the retrieved lidar ozone profiles. The relative ozone differences computed using those two cross-section data are less than ±1% from 10 to 35km at all latitudes, except a −1.5% deviation at 15km in the tropics. Above 35 km, the deviations increase with a maximum of 1.7% in the tropics and a minimum of 1.4% in the high latitudes. The stability of various GB and satellite ozone profile time series is then evaluated by comparing with the ozone lidar data for each station. All ozone profile measurement techniques show their best agreement (±3%) with lidars in the 20-40km altitude range and the estimated drifts are less than ±0.3%yr−1 at all stations. Comparatively large biases and drifts are computed below 20 and above 40 km. A combined time series of the relative differences of SAGE II, HALOE and Aura MLS with respect to the lidar measurements at the six lidar sites is constructed to obtain long-term data sets from 1985 to 2010. The relative drifts derived from these combined data of 27 years are very small, within ±0.2%yr−1. Then, stratospheric ozone trends are estimated at Meteorological Observatory Hohenpeissenberg (MOHp) using Dobson, and at Haute-Provence Observatory (OHP) using Dobson and SAOZ total column measurements and various GB and satellite ozone profiles. For that a multiple regression model is developed using different explanatory variables such as Quasi Biennial Oscillation (QBO), North Atlantic Oscillation (NAO), solar flux, eddy heat flux, aerosols and trend. The PWLTs computed from the ozone column at OHP and MOHp show significant negative (−1.4 ± 0.29DUyr−1) and positive (0.55 ± 0.29DUyr−1) values before and after 1997, respectively, indicating a clear signal of ozone recovery at these latitudes after 1996. Vertical distribution of ozone trends based on PWLT model, estimated using the all instrument average at OHP exhibit about −0.5± 0.1%yr−1 in the 16-22km range and about −0.8 ± 0.2%yr−1 in the 38-45km region before 1997. Significant positive trends (0.2 ± 0.05-0.3 ± 0.1%yr−1) are estimated in the 15-45km altitude region after 1996. These significant ozone profile trends in the respective periods corroborate those derived from the ozone total column and hence, provide signs of ozone recovery in the northern mid-latitudes. The trends based on both PW and EESC regressions are similar and significant before 1997 while they differ slightly after 1996, with the largest value in the PW regression. In addition, the most recent increase in ozone after 1996 is due to the increase in QBO and planetary wave drive. For instance, QBO, NAO and heat flux contribute about 20-26DU to the large total ozone anomaly of 25-30DU in the winter/spring months in 2010. Therefore, this thesis presents some new and interesting results on the mid-latitude stratospheric ozone recovery.