| Summary: | Model-derived estimates of marine new production are found to display systematic covariations with the underlying model architecture. Almost regardless of the formulation of biogeochemical processes, model-derived estimates of new production have more than doubled from about 0.15 molNm?2 yr?1 to values around 0.4 molNm?2 yr?1 when turning from early box models to more recent investigations using coarse-resolution general circulation models. Because none of these models resolves eddies, which have been shown to enhance biological production, a further increase in simulated new production with increasing model resolution might be expected. This study presents results from an eddy-permitting coupled biological–physical model that suggest a basin-scale new production of less than 0.3 molNm?2 yr?1 for the North Atlantic, i.e. substantially less than values typical for coarse-resolution models. Sensitivity experiments reveal that the amount of diapycnal mixing, described either explicitly or implicitly in the numerical discretization schemes, has a considerable effect on the simulated input of nutrients into the euphotic zone. Implications for coarse-resolution models used until now are that unrealistically high levels of explicit and implicit diapycnal diffusion may have been responsible for unrealistically high estimates of new production.
|
|---|