SOCOMv2: On the strengths and limits of pCO2 interpolations products to estimate the ocean carbon sink

The ocean is an important sink for anthropogenic carbon dioxide (CO₂), but recent data from the Global Carbon Budget (GCB) highlight discrepancies in ocean carbon uptake estimates. Since the early 2000s, reconstructions of in-water CO₂ fugacity (fCO₂) using advanced interpolation techniques (data-products) have shown a growing divergence from estimates derived from global hindcast model simulations. This offsets in the mean flux amounts to approximately 0.49 GtC per year in the decade 2014-2023. The reasons for this discrepancy are not fully understood but may stem from a combination of factors including insufficient data coverage, uncertainties in scaling measurement-based estimates, and errors in model simulations. Previous studies suggests that biases in the fCO₂ data-products from the under-sampled Southern Hemisphere, may contribute significantly to this gap.

To address these concerns, the Surface Ocean CO₂ Mapping project has launched its second phase (SOCOMv2). This initiative aims to identify and quantify the accuracy and uncertainties related to data availability, changing observational networks, and input data. SOCOMv2 includes four key experiments: 1) a comprehensive geospatial uncertainty analysis, and three subsampling studies employing: 2) GCB hindcast simulations to capture true climate variability, 3) large ensemble simulations representing multiple climate states, and 4) idealized carbon uptake scenarios without climate variation. These efforts aim to provide a clearer understanding of the underlying factors contributing to the observed discrepancies in ocean carbon uptake estimates.

Results from the GCB subsampling hindcast simulation experiments reveal that individual fCO₂ data-product reconstructions can significantly overestimate or underestimate both the annual mean and the trend of the ocean carbon sink relative to the models ‘truth’. Nonetheless, the ensemble mean of the fCO₂ data-products tends to exhibit only a small overestimation of the model ‘truth’ ocean carbon sink. These discrepancies highlight the impact of limited data coverage and the inherent challenges of extrapolating from sparse measurements but cannot fully explain the observed divergence between models and fCO₂ reconstructions in the GCB.

SOCOMv2 aims to improve the accuracy and precision of ocean carbon flux estimates, contributing to improved observational approaches and guiding policy development for climate mitigation. SOCOMv2 efforts have been driven by the community, with supporting funding within a larger European Space Agency ocean carbon study (Ocean Carbon for Climate).