Description: This paper discusses an observing system simulation experiment which reveals the difference in primary production of (i) phytoplankton moving freely in the turbulent mixed layer of the upper ocean and (ii) a sample of the same population held in a bottle at fixed depths. The results indicate the tendency of incubation measurements to overestimate phytoplankton production rates by up to 40%. Differences in primary production depend to a first approximation on the vertical extent of mixing and on water turbidity. A simple model was constructed leading to a non-linear calibration function which relates the difference in primary production to surface irradiance, mixing depth and to the depth of the euphotic zone. This function has been applied to calibrate the production rates simulated at fixed depths, and the corrected values were verified by comparisons with productivities in the turbulent environment. The calibration function was found to be capable of reducing the differences significantly.

Description: One of the critical issues in large-scale physical/biological coupled models is the survival of zooplankton in a water column circulating an anticyclonic gyre. Survival is most at risk in regions where the phytoplankton food supply is low due to environmental stress by light-limitation (deep mixing in winter) or nutrient limitation (oligotrophy). To investigate this problem we simulated the ecosystem in a 1 m2 cross-section water column, using the Lagrangian Ensemble method in which plankton are treated as particles following independent trajectories through the changing environment. In this first part of a two-part article we report the results of simulating the ecosystem in a water column located off the Azores, where winter mixing reaches 200 m and there is seasonal, but not permanent oligotrophy. The model features diatoms and herbivorous copepods subject to carnivorous predation, with remineralization of carbon and nitrogen by bacteria attached to detritus and faecal pellets. The copepods become extinct after failing to reproduce in years of low food supply. We show that the risk of extinction can be reduced by allowing cannibalism or by reducing carnivorous predation; we discuss other possibilities: enhancing the food supply by adding new guilds of phytoplankton, and relaxing oligotrophy by allowing other sources of nitrogen injection into the euphotic zone.

Description: One of the most striking features of the upper North Atlantic Ocean is an extensive layer of water with temperature close to 18°C and salinity close to 36.5‰, (ref. 1). This 18°C water is formed by winter convection in the Sargasso sea2,3, but aspects of the annual rate of 18°C water formation remain obscure4. We have simulated this water mass formation by integrating a one-dimensional model along a 4-yr trajectory of a water column circulating around the Sargasso Sea. Winter convection is deep (≥200 m) in regions where the ocean suffers a net annual heat loss to the atmosphere, and shallow (≤lOOm) where the ocean gains heat each year. The origin of the thermostad (nearly isothermal layer) is a thick layer of nearly homogeneous water subducted beneath the seasonal boundary layer in the year that the water column passes through the line dividing annual cooling from annual heating. We estimate the annual production of 18°C water to be 446,000 km3 yr−1. Downstream, more stratified central water is formed each year at a rate that depends more on Ekman pumping (wind-forced convergence) than on the decreasing depth of winter convection

Description: The current status of the Sverdrup theory for the initiation of plankton blooms is examined. A prescription is given for the computation of the Sverdrup critical depth, using recently-published algorithms for mixed-layer primary production and a generalised loss term. Using no further information, the intrinsic rate of increase of phytoplankton biomass in the mixed layer can also be found. This rate, compared against the local frequency of storm occurrence, provides an alternative criterion for the initiation of blooms. The Eulerian (bulk property) methods used to derive these results are contrasted with the Lagrangian Ensemble method. The Lagrangian approach provides one avenue to the elaboration of the Sverdrup criterion to include the effect of processes with characteristic timescales small compared to one day. The incidence of blooms in the apparent absence of vertical stratification is reviewed: it is concluded that these observations do not undermine the basic logic of the Sverdrup theory. However, they do provoke interest in a re-examination of the feedbacks between the physical and biological dynamics in the mixed layer: an example is given. Finally, suggestions are made for further work in this subject area.

Description: Ocean biogeochemical models are routinely applied to assess the net global impact of ocean acidification and warming on pelagic CaCO3 cycling. As with respect to the net change of global air-sea carbon fluxes affected by the reduced calcification under future CO2 conditions, these models diverge by a factor of four. The standard method to evaluate modeled CaCO3 cycles is to compare alkalinity and CaCO3 saturation states with observations. In general, state-of-the-art models do feature strong deviations and it is unclear if, or to what extent, these are driven by a deficient representation of physics (ocean circulation) or a deficient representation of biogeochemistry. This points to a strong need for improvement of the data-based evaluation of the base state of global biogeochemical models used for ocean acidification research. Here we apply the TA* method to output from a variety of model experiments and observations (GLODAP). This method was originally developed to separate the signals of CaCO3 production and dissolution from the large, conservative alkalinity background in observations and is also a critical part of approaches used to quantify the inventory of anthropogenic CO2 in the ocean. The aim is our study is twofold. First, to assess the TA* method using additional explicit representations of preformed alkalinity, accumulated CaCO3 dissolution, and organic matter remineralisation in our models. And second, we aim to disentangle deficiencies in the physical and biogeochemical CaCO3-cyle module in a series of ocean biogeochemical models of increasing complexity. Finally, our modeling study provides a critical assessment of the ‘mystery of shallow CaCO3 dissolution’, i.e. apparent dissolution of major CaCO3 minerals well above their saturation horizon.