LAUSR

Abstract Using the Solar Back-scatter Ultraviolet merged ozone data during 1980–2016, the recovery of total ozone column (TOC) is examined in 5° wide latitude bands from 65°S to 65°N across the globe. Since the variability of TOC is influenced by various natural and anthropogenic proxies, a multiple linear regression (MLR) is employed to remove the dynamic variability. Among the proxies, lower stratospheric temperature at ~100hPa plays a significant role in the dynamics of the TOC variability. The MLR model has significantly improved after including the lower stratospheric temperature along with other proxies and the standard deviation of the estimated TOC trend has significantly reduced about 35 to 43% in 50°N-65°N and 16 to 20% in 45°S-65°S after using the lower stratospheric temperature during both the depletion (1980–1996) and recovery (1997–2016) phases. Cooling of lower-middle stratospheric temperature are observed during the depletion phase with −0.09 °C to −0.02 °C year−1. However, during the recovery phase, the temperature at the lower stratosphere is warming. Due to weak signals, the estimated warming trends are not statistically significant. The TOC trends are examined by the increasing or decreasing emission of chlorofluorocarbon, indicated by Equivalent Effective Stratospheric Chlorine (EESC) index. During the recovery phase, the EESC varies from 0.82±0.01 to 0.29±0.005 DU year−1 at 45–65°S using the lower stratospheric temperature. In the northern hemisphere, the recovery trends are weak and varied from 0.21±0.004 to 0.25±0.005 DU year−1 and these trends are statistically significant. This study suggests the importance of the lower stratospheric temperature to derive proper trends using the MLR model.