Abstract. Given the importance of the long-term monitoring of the evolution of ozone and following the frequent disruptions of satellites operations, the understanding of quality and stability of ground-based instruments measurements (Dobson, Brewer, SAOZ) performing over several decades, is essential. This is not only for their own records, but also for evaluating the performance of satellites systems and the reprocessing of their data. Many data sets inter-comparisons between Dobson/Brewer networks and satellites are already available (e.g. Redondas et al. 2014 and references therein), for the 40 years of SBUV (Labow et al. 2013) or for the OMI AURA (Balis et al., 2007; McPeters et al. 2008). Here, we evaluate the performance of SAOZ Total Ozone Column (TOC) measurements carried out in Arosa between October 2015 and March 2017 by comparison to simultaneous Dobson, Brewer and satellites observations available there. On average, when using Serdyuchenko et al. (2014) ozone cross-sections for all measurements and correcting the Dobson for its temperature dependence, Dobson, Brewer and SAOZ agree within 1 %, providing confidence to the long-term SAOZ measurements carried out all over the world within the international Network for the Detection of Atmospheric Composition Change (NDACC). However, the differences between SAOZ and Dobson/Brewer do show a small seasonality of the order of 1–2 % generated by the use of a zonal mean TOMS v8 profile climatology in the SAOZ air mass factor (AMF) calculation, not taking into account seasonal variations.

On the satellites side, the differences with ground based TOCs show larger and highly variable biases (between −1.2 % and +2.4 %) as well as larger seasonality (around 2–3 %) on all satellites, except remarkably with the Solar Backscatter Ultra Violet instrument (SBUV) for which the bias is smaller than 0.4% and the seasonality less than 1 %. Asides from ozone absorption cross-sections, most important for satellites TOCs differences are the satellites measurements techniques: ozone profiles for SBUV, in contrast to all other nadir viewing satellites, sensitive like SAOZ, to ozone profile shape assumption.