comparison of gto-ecv and adjusted merra-2 total ozone columns from the last 2 decades and assessment of interannual variability

In this study we compare the satellite-based Global Ozone Monitoring Experiment (GOME)-type Total Ozone Essential Climate Variable (GTO-ECV) record, generated as part of the European Space Agency's Climate Change Initiative (ESA-CCI) ozone project, with the adjusted total ozone product from the Modern Era Retrospective Analysis for Research and Applications version 2 (adjusted MERRA-2) reanalysis, produced at the National Aeronautics and Space Administration (NASA) Global Modeling and Assimilation Office (GMAO). Total ozone columns and associated standard deviations show a very good agreement in terms of both spatial and temporal patterns during their 23-year overlap period from July 1995 to December 2018. The mean difference between adjusted MERRA-2 and GTO-ECV $\mathrm{5}{}^{\circ }\times \mathrm{5}{}^{\circ }$ monthly mean total ozone columns is $-\mathrm{0.9}\pm \mathrm{1.5}$ %. A small discontinuity in the deviations is detected in October 2004, when data from the Ozone Monitoring Instrument (OMI) were ingested in the GTO-ECV and adjusted MERRA-2 data records. This induces a small overall negative drift in the differences for almost all latitude bands, which, however, does not exceed 1 % per decade. The mean difference for the period prior to October 2004 is $-\mathrm{0.5}\pm \mathrm{1.7}$ %, whereas the difference is $-\mathrm{1.0}\pm \mathrm{1.1}$ % for the period from October 2004 to December 2018. The variability in the differences is considerably reduced in the period after 2004 due to a significant increase in data coverage and sampling. In the tropical region, the differences indicate a slight zonal variability with negative deviations over the Atlantic, Africa, and the Indian Ocean and positive deviations over the Pacific. Ozone anomalies and the distribution of their statistical moments indicate a very high correlation among both data records as to the temporal and spatial structures. Furthermore, we evaluate the consistency of the data sets by means of an empirical orthogonal function (EOF) analysis. The interannual variability is assessed in the tropics, and both GTO-ECV and adjusted MERRA-2 exhibit a remarkable agreement with respect to the derived patterns. The first four EOFs can be attributed to different modes of interannual climate variability, and correlations with the Quasi-Biennial Oscillation (QBO), the El Niño–Southern Oscillation (ENSO) signal, and the solar cycle were found.