We report on updated trends using different merged datasets from satellite and ground-based observations for the period from 1979 to 2016. Trends were determined by applying a multiple linear regression (MLR) to annual mean zonal mean data. Merged datasets used here include NASA MOD v8.6 and National Oceanic and Atmospheric Administration (NOAA) merge v8.6, both based on data from the series of Solar Backscatter UltraViolet (SBUV) and SBUV-2 satellite instruments (1978–present) as well as the Global Ozone Monitoring Experiment (GOME)-type Total Ozone (GTO) and GOME-SCIAMACHY-GOME-2 (GSG) merged datasets (1995–present), mainly comprising satellite data from GOME, the Scanning Imaging Absorption Spectrometer for Atmospheric Chartography (SCIAMACHY), and GOME-2A. The fifth dataset consists of the monthly mean zonal mean data from ground-based measurements collected at World Ozone and UV Data Center (WOUDC). The addition of four more years of data since the last World Meteorological Organization (WMO) ozone assessment (2013–2016) shows that for most datasets and regions the trends since the stratospheric halogen reached its maximum (∼ 1996 globally and ∼ 2000 in polar regions) are mostly not significantly different from zero. However, for some latitudes, in particular the Southern Hemisphere extratropics and Northern Hemisphere subtropics, several datasets show small positive trends of slightly below +1 % decade⁻¹ that are barely statistically significant at the 2σ uncertainty level. In the tropics, only two datasets show significant trends of +0.5 to +0.8 % decade⁻¹, while the others show near-zero trends. Positive trends since 2000 have been observed over Antarctica in September, but near-zero trends are found in October as well as in March over the Arctic. Uncertainties due to possible drifts between the datasets, from the merging procedure used to combine satellite datasets and related to the low sampling of ground-based data, are not accounted for in the trend analysis. Consequently, the retrieved trends can be only considered to be at the brink of becoming significant, but there are indications that we are about to emerge into the expected recovery phase. However, the recent trends are still considerably masked by the observed large year-to-year dynamical variability in total ozone.

We use height‐resolved and total column satellite observations and 3‐D chemical transport model simulations to study stratospheric ozone variations during 1998–2017 as ozone‐depleting substances decline. In 2017 extrapolar lower stratospheric ozone displayed a strong positive anomaly following much lower values in 2016. This points to large interannual variability rather than an ongoing downward trend, as reported recently by Ball et al. (2018, https://doi.org/10.5194/acp‐18‐1379‐2018). The observed ozone variations are well captured by the chemical transport model throughout the stratosphere and are largely driven by meteorology. Model sensitivity experiments show that the contribution of past trends in short‐lived chlorine species to the ozone changes is small. Similarly, the potential impact of modest trends in natural brominated short‐lived species is small. These results confirm the important role that atmospheric dynamics plays in controlling ozone in the extrapolar lower stratosphere on multiannual time scales and the continued importance of monitoring ozone profiles as the stratosphere changes.

We report on updated trends using different merged zonal mean total ozone datasets from satellite and ground-based observations for the period from 1979 to 2020. This work is an update of the trends reported in Weber et al. (2018) using the same datasets up to 2016. Merged datasets used in this study include NASA MOD v8.7 and NOAA Cohesive Data (COH) v8.6, both based on data from the series of Solar Backscatter Ultraviolet (SBUV), SBUV-2, and Ozone Mapping and Profiler Suite (OMPS) satellite instruments (1978–present), as well as the Global Ozone Monitoring Experiment (GOME)-type Total Ozone – Essential Climate Variable (GTO-ECV) and GOME-SCIAMACHY-GOME-2 (GSG) merged datasets (both 1995–present), mainly comprising satellite data from GOME, SCIAMACHY, OMI, GOME-2A, GOME-2B, and TROPOMI. The fifth dataset consists of the annual mean zonal mean data from ground-based measurements collected at the World Ozone and Ultraviolet Radiation Data Centre (WOUDC). Trends were determined by applying a multiple linear regression (MLR) to annual mean zonal mean data. The addition of 4 more years consolidated the fact that total ozone is indeed slowly recovering in both hemispheres as a result of phasing out ozone-depleting substances (ODSs) as mandated by the Montreal Protocol. The near-global (60^∘ S–60^∘ N) ODS-related ozone trend of the median of all datasets after 1995 was 0.4 ± 0.2 (2σ) %/decade, which is roughly a third of the decreasing rate of 1.5 ± 0.6 %/decade from 1978 until 1995. The ratio of decline and increase is nearly identical to that of the EESC (equivalent effective stratospheric chlorine or stratospheric halogen) change rates before and after 1995, confirming the success of the Montreal Protocol. The observed total ozone time series are also in very good agreement with the median of 17 chemistry climate models from CCMI-1 (Chemistry-Climate Model Initiative Phase 1) with current ODS and GHG (greenhouse gas) scenarios (REF-C2 scenario). The positive ODS-related trends in the Northern Hemisphere (NH) after 1995 are only obtained with a sufficient number of terms in the MLR accounting properly for dynamical ozone changes (Brewer–Dobson circulation, Arctic Oscillation (AO), and Antarctic Oscillation (AAO)). A standard MLR (limited to solar, Quasi-Biennial Oscillation (QBO), volcanic, and El Niño–Southern Oscillation (ENSO)) leads to zero trends, showing that the small positive ODS-related trends have been balanced by negative trend contributions from atmospheric dynamics, resulting in nearly constant total ozone levels since 2000.

A new algorithm approach called Weighting Function Differential Optical Absorption Spectroscopy (WFDOAS) is presented which has been developed to retrieve total ozone columns from nadir observations of the Global Ozone Monitoring Experiment. By fitting the vertically integrated ozone weighting function rather than ozone cross-section to the sun-normalized radiances, a direct retrieval of vertical column amounts is possible. The new WFDOAS approach takes into account the slant path wavelength modulation that is usually neglected in the standard DOAS approach using single airmass factors. This paper focuses on the algorithm description and error analysis, while in a companion paper by Weber et al. (2004) a detailed validation with groundbased measurements is presented. For the first time several auxiliary quantities directly derived from the GOME spectral range such as cloud-top-height and cloud fraction (O₂-A band) and effective albedo using the Lambertian Equivalent Reflectivity (LER) near 377nm are used in combination as input to the ozone retrieval. In addition the varying ozone dependent contribution to the Raman correction in scattered light known as Ring effect has been included. The molecular ozone filling-in that is accounted for in the new algorithm has the largest contribution to the improved total ozone results from WFDOAS compared to the operational product. The precision of the total ozone retrieval is estimated to be better than 3% for solar zenith angles below 80°.

Within the European Space Agency's Climate Change Initiative, total ozone column records from GOME (Global Ozone Monitoring Experiment), SCIAMACHY (SCanning Imaging Absorption SpectroMeter for Atmospheric CartograpHY), and GOME‐2 have been reprocessed with GODFIT version 3 (GOME‐type Direct FITting). This algorithm is based on the direct fitting of reflectances simulated in the Huggins bands to the observations. We report on new developments in the algorithm from the version implemented in the operational GOME Data Processor v5. The a priori ozone profile database TOMSv8 is now combined with a recently compiled OMI/MLS tropospheric ozone climatology to improve the representativeness of a priori information. The Ring procedure that corrects simulated radiances for the rotational Raman inelastic scattering signature has been improved using a revised semi‐empirical expression. Correction factors are also applied to the simulated spectra to account for atmospheric polarization. In addition, the computational performance has been significantly enhanced through the implementation of new radiative transfer tools based on principal component analysis of the optical properties. Furthermore, a soft‐calibration scheme for measured reflectances and based on selected Brewer measurements has been developed in order to reduce the impact of level‐1 errors. This soft‐calibration corrects not only for possible biases in backscattered reflectances, but also for artificial spectral features interfering with the ozone signature. Intersensor comparisons and ground‐based validation indicate that these ozone data sets are of unprecedented quality, with stability better than 1% per decade, a precision of 1.7%, and systematic uncertainties less than 3.6% over a wide range of atmospheric states.

In preparation of the TROPOMI/S5P launch in early 2017, a tropospheric ozone retrieval based on the convective cloud differential method was developed. For intensive tests we applied the algorithm to the total ozone columns and cloud data of the satellite instruments GOME, SCIAMACHY, OMI, GOME-2A and GOME-2B. Thereby a time series of 20 years (1995–2015) of tropospheric column ozone was generated. To have a consistent total ozone data set for all sensors, one common retrieval algorithm, namely GODFITv3, was applied and the L1 reflectances were also soft calibrated. The total ozone columns and the cloud data were input into the tropospheric ozone retrieval. However, the tropical tropospheric column ozone (TCO) for the individual instruments still showed small differences and, therefore, we harmonised the data set. For this purpose, a multilinear function was fitted to the averaged difference between SCIAMACHY's TCO and those from the other sensors. The original TCO was corrected by the fitted offset. GOME-2B data were corrected relative to the harmonised data from OMI and GOME-2A. The harmonisation leads to a better agreement between the different instruments. Also, a direct comparison of the TCO in the overlapping periods proves that GOME-2A agrees much better with SCIAMACHY after the harmonisation. The improvements for OMI were small. Based on the harmonised observations, we created a merged data product, containing the TCO from July 1995 to December 2015. A first application of this 20-year record is a trend analysis. The tropical trend is 0.7 ± 0.12 DU decade⁻¹. Regionally the trends reach up to 1.8 DU decade⁻¹ like on the African Atlantic coast, while over the western Pacific the tropospheric ozone declined over the last 20 years with up to 0.8 DU decade⁻¹. The tropical tropospheric data record will be extended in the future with the TROPOMI/S5P data, where the TCO is part of the operational products.