Increasing atmospheric CO2 drives ocean acidification, with some of the fastest rates observed on Arctic shelves. Lowest pH levels occur near the coasts due to the additional effects of land-derived carbon sources. However, the impact of anthropogenic climate change on seawater acidity mediated by increasing permafrost thaw and coastal erosion remains unknown. Here, we find that the increase in organic matter release by coastal permafrost erosion over the 20th century enhances the interannual variability of seawater pH in Arctic near-shore areas. As a consequence, carbonate undersaturation events, stressing marine ecosystems, become more frequent, intense, and longer-lasting. We account for synergistic effects of anthropogenic climate change by combining increasing atmospheric CO2 levels and coastal erosion rates, utilizing a novel global model with high-resolution grid refinement towards coastal regions and improved process representation of shelf-specific carbon dynamics. By considering the role of coastal permafrost erosion, we conclude that critical Arctic Ocean acidification states will emerge earlier than simulated by the current generation of Earth system models. Our results emphasize the importance of understanding permafrost-ocean interactions and their increasing impact on marine ecosystems, especially in the rapidly changing Arctic region.

We present the first global ocean-biogeochemistry model that uses a telescoping high resolution for an improved representation of coastal carbon dynamics: ICON-Coast. Based on the unstructured triangular grid topology of the model, we globally apply a grid refinement in the land-ocean transition zone to better resolve the complex circulation of shallow shelves and marginal seas as well as ocean-shelf exchange. Moreover, we incorporate tidal currents including bottom drag effects, and extend the parameterizations of the model’s biogeochemistry component to account explicitly for key shelf-specific carbon transformation processes. These comprise sediment resuspension, temperature-dependent remineralization in the water column and sediment, riverine matter fluxes from land including terrestrial organic carbon, and variable sinking speed of aggregated particulate matter. The combination of regional grid refinement and enhanced process representation enables for the first time a seamless incorporation of the global coastal ocean in model-based Earth system research. In particular, ICON-Coast encompasses all coastal areas around the globe within a single, consistent ocean-biogeochemistry model, thus naturally accounting for two-way coupling of ocean-shelf feedback mechanisms at the global scale. The high quality of the model results as well as the efficiency in computational cost and storage requirements proves this strategy a pioneering approach for global high-resolution modeling. We conclude that ICON-Coast represents a new tool to deepen our mechanistic understanding of the role of the land-ocean transition zone in the global carbon cycle, and to narrow related uncertainties in global future projections.