Impact of waves on phytoplankton activity on the Northwest European Shelf: insights from observations and km-scale coupled models

The spring bloom is an annual event in temperate regions of the North Atlantic Ocean where the abundance of photosynthetic plankton increases dramatically. The timing and intensity of the spring bloom is dependent on underlying physical conditions that control ocean stratification and mixing. Although surface waves can be an important source of turbulent kinetic energy to the surface mixed layer, they have seldom been considered explicitly in studies of bloom formation. Here, we investigate the role of surface waves in bloom formation using a combination of satellite observations and numerical models. Satellite observations show a positive correlation between wave activity and chlorophyll concentration in the Northwest European shelf (May–September). In the deeper Northeast Atlantic, increased wave activity correlates with lower chlorophyll during periods of high phytoplankton activity (March-May) and higher chlorophyll when activity is low (below 54° N, July–September). We use a first-of-its-kind, km-scale, two-way coupled ocean-wave model system to investigate both the relationship between wave-driven mixing and bloom formation, and the sensitivity of model results to the method by which wave-driven mixing is parameterised. In deep regions, during the spring bloom, a wave-driven mixing event is likely to mix surface chlorophyll to deeper layers, away from light. In contrast, when phytoplankton activity is low in deep regions, wave-driven mixing can entrain nutrients, fueling the growth of nutrient starved phytoplankton near the surface. In June–October, in shallow but weakly stratified regions of the shelf, surface chlorophyll tends to be elevated following a wave-driven mixing event, which can bring to the surface both phytoplankton and nutrients from deeper layers. When contrasted with ocean-only runs, the two way-coupled ocean-wave model tends to produce greater vertical mixing and a delay in bloom onset. These results indicate bloom dynamics are sensitive to the way in which waves are modelled, and that the role of waves in bloom formation should be considered in future studies.