8-day SIF at 0.05 degree

Photosynthesis (or Gross Primary Production, GPP) is a key process in hydrology and ecology linking carbon and water cycles. The newly available satellite-retrieved solar-induced chlorophyll fluorescence (SIF) is a direct measure of photosynthesis globally. However, the existing satellite SIF records are limited by low spatial resolution, sparse data acquisition, or both. The situation changed in 2018 with the launched TROPOspheric Monitoring Instrument (TROPOMI), from which SIF can be retrieved with unprecedented resolution, but the short temporal coverage of this data record precludes its use for many applications. This study reconstructs a long-term contiguous and high resolution TROPOMI SIF product over 2001-2019 at a 0.05° and 8-day resolution using machine learning combined with the physiological constraints. The model provides a significantly more accurate prediction of SIF (R²= 0.919, regression slope=1.002), which is consistent with GPP estimations and other independent SIF measurements. The long-term, high-resolution, contiguous global coverage of this dataset will be valuable for estimating GPP to constrain the global carbon budget and related water fluxes.