Addressing degradation and geometry effects to develop consistent global solar induced fluorescence records from GOME-2A and GOME-2B (2007-2023)

Space-based solar-induced fluorescence (SIF) observations provide critical insights into vegetation activity over time. The GOME-2A and GOME-2B instruments offer extensive global SIF data spanning 2007 to 2023. However, calibration issues and instrument degradation have complicated the generation of consistent long-term records. Combining SIF products with differing viewing geometries and spatio-temporal coverage remains challenging, even for similar satellite instruments GOME-2A and GOME-2B.

We present the SIFTER v3 algorithm, developed to deliver a more accurate and reliable SIF record for the 2007–2023 period. Using newly reprocessed level-1b R3 data from EUMETSAT ensures SIFTER v3 processes the GOME-2A and GOME-2B retrievals with consistent calibration settings. Despite the improvements in R3 reflectances, additional corrections for long-term degradation in the 734–758 nm retrieval window are necessary for both GOME-2A and GOME-2B. This concerns in-flight corrections that address time, wavelength, and scan-angle dependent reflectance degradation. 

After applying these corrections, the SIFTER v3 dataset exhibits enhanced consistency, aligning closely with NASA GOME-2A data and GPP estimates from FluxSat and FLUXCOM-X. To produce a coherent SIF record suitable for climate analysis, the algorithm addresses (1) spatio-temporal sampling differences and (2) viewing geometry dependencies. Co-sampling GOME-2A and GOME-2B significantly improves consistency and reduces inter-sensor SIF offsets by up to 15%. Notably, we demonstrate that GOME-2 measures up to 30% higher SIF in the orbit’s western regions, where vegetation is sunlit, compared to shaded vegetation in the east. By quantifying these geometry effects across regions and seasons, we propose corrections to make level-2 SIF data suitable for daily applications.