Decadal and spatially complete global surface chlorophyll-a data record from satellite and BGC-Argo observations

Abstract. Decadal-scale satellite-based climate data records of chlorophyll-a (chl-a), an essential climate variable, are now readily available at high accuracy and precision. These data are being extensively used for research and, increasingly, for operational services. However, these observations rely on availability of sunlight and the satellite sensor being able to view the ocean, so there are gaps in data due to the presence of clouds and more widely during the polar winter. This is an issue when spatially complete data are needed for global climate studies, or as inputs to machine learning methods and for data assimilation. Whilst addressing cloud cover is well studied, methodologies to overcome missing data due to the polar winter has received little attention and simple approaches to overcome these gaps can lead to unrealistic values. Biogeochemical Argo (BGC-Argo) floats have widely been deployed, and they represent an opportunity to address these gaps. We present an approach that combines BGC-Argo data and a satellite chl-a climate data record to produce a spatially and temporally complete, global monthly chl-a record between 1997 and 2023 at 0.25° spatial resolution. Clouds gaps were filled using an established spatial kriging approach. Polar wintertime chl-a were reconstructed using relative changes between the wintertime BGC-Argo chl-a, and the previous autumntime or next springtime satellite observations, for individual hemispheres. Uncertainties were calculated on a per-pixel basis to retain the underlying uncertainty fields in the climate data record and were modified to account for the uncertainties related to the gap filling. The seasonal cycles in the resulting polar data are consistent with light availability. Clear interannual and inter-hemisphere variability in the wintertime chl-a were observed. Independent assessment of solely the gap filled wintertime chl-a estimates against in situ data (N = 204 total) indicates that the accuracy and precision of the underlying satellite data, a key component of a climate data record, are maintained. The 25 year global and spatially complete chl-a data, that are consistent with the underlying climate data record can be downloaded from Zenodo (Ford et al., 2025b).