Examination on total ozone column retrievals by Brewer spectrophotometry using different processing software

The availability of long-term records of the total ozone content (TOC) represents a valuable source of information in studies on the assessment of short and long-term changes and their impact on the terrestrial ecosystem. In addition, ground-based observations represent a valuable tool to validate satellite-derived products. To our knowledge, details about processing software packages to retrieve the TOC from Brewer spectrophotometer measurements are seldom specified in studies concerning such datasets, although some discrepancies can arise from the use of different algorithms and implementations. The deviations among retrieved TOCs from the Brewer instruments located at Rome and Aosta (Italy), using different processing software (Brewer Processing Software, O3Brewer software and EUBREWNET products (Level 1.5) are investigated. Ground-based TOCs are also compared with the Ozone Monitoring Instrument (OMI) TOC retrievals used as an independent dataset since no other instruments near the Brewer sites, are available. Although the overall agreement of the BPS and O3Brewer TOC data with EUBREWNET data is clearly very good (as expected) and in most cases within the Brewer declared uncertainty less than 2 %, it is worth noticing that slight differences have been seen depending on the software in use. Such differences become larger when the instrumental sensitivity exhibits a long-term drift and even in short-term episodes due to the different algorithm for the standard lamp correction. This work aims to provide useful information both for scientists engaged in ozone measurements with Brewer spectrophotometry and for stakeholders of the Brewer data products available at web-based platforms.