Validation of the MetOp-A total ozone data from GOME-2 and IASI using reference ground-based measurements at the Iberian Peninsula

One of the most important atmospheric composition products derived from the first EUMETSAT Meteorological Operational satellite (MetOp-A) is the total ozone column (TOC). For this purpose, MetOp-A has two instruments on board: the Global Ozone Monitoring Experiment 2 (GOME-2) that retrieves the TOC data from the backscattered solar ultraviolet–visible (UV–Vis) radiance, and the Infrared Atmospheric Sounding Interferometer (IASI) that uses the thermal infrared radiance to derive TOC data. This paper focuses on the simultaneous validation of the TOC data provided by these two MetOp-A instruments using the measurements recorded by five well-calibrated Brewer UV spectrophotometers located at the Iberian Peninsula during the complete 2009. The results show an excellent correlation between the ground-based data and the GOME-2 and IASI satellite observations (R2 higher than 0.91). Differences between the ground-based and satellite TOC data show that the IASI instrument significantly overestimates the Brewer measurements (about 4.4% when all five ground-based stations are jointly used). In contrast, the GOME-2 instrument shows a slight underestimation (~ 1.6%). In addition, the absolute relative differences between the Brewer and GOME-2 data are quite smaller (about a factor higher than 2) than the Brewer–IASI absolute differences. The satellite viewing geometry (solar zenith angle and the view zenith angle) has no significant influence on the Brewer–satellite relative differences. Moreover, the analysis of these relative differences with respect to the ground-based TOC data indicates that GOME-2 instrument presents a slight underestimation for high TOC values. Finally, the IASI–GOME-2 correlation is high (R2 ~ 0.92), but with a mean relative difference of about ± 6% which could be associated with the bias between UV–Vis and infrared spectroscopy used in the retrieval processes.

Research Highlights
► IASI significantly overestimates the Brewer data (~ 4.4%).
► GOME-2 shows a slight underestimation (~ 1.6%) of Brewer data.
► The satellite viewing geometry has no influence on the Brewer–satellite differences.
► Brewer–sat. differ. may be due to the bias between UV–Vis and infrared spectroscopy.