Table Of Contents

This Page

Notes on “Nutrient supply to the surface waters of the North Atlantic: A model study” by Oschlies (2002a)

“The simulated annual mean input of nitrate into the euphotic zone is separated into different supply routes, namely, turbulent vertical mixing, vertical advection, and horizontal transport. Vertical mixing is found to be the dominant supply mechanism in the subpolar North Atlantic, while horizontal advection provides most of the simulated nitrate input into the subtropical gyre. [...] The model results are used to investigate to what extent the advective input of organic matter could possibly explain a local imbalance between new and export production. It turns out that the simulated advective input of organic matter alone, which in the model provides nitrogen at a rate similar to that arising from nitrate supply, is not sufficient to explain observed oxygen consumption rates.”

“Of particular concern is the model’s failure to sustain observed levels of primary production in the subtropical gyre. [...] It is not yet clear whether the underestimation in simulated primary production is caused mainly by too weak levels of new production or by too low levels of regenerated production. The present study attempts to clarify this issue by investigating how realistically the model simulates the nitrate input into the upper layer of the North Atlantic.”

“Vertical turbulent mixing, which in the model diagnostics includes convective overturning, provides a net upward transport of nitrate almost everywhere in the model domain.”

“the simulated nitrate input by vertical advection is mainly a result of the large-scale atmospheric wind patterns. In particular, the entire subtropical gyre is a region of Ekman downwelling with an associated export of nitrate out of the euphotic zone.”

In the North Atlantic subtropical gyre, vertical advection is downward and horizontal transport is convergent with the net result being still downward –net advection removes nitrate from the euphotic zone. This is counter-balanced by the net input of nitrate by vertical mixing.

“At the northern flank of the subtropical gyre, southward Ekman transport, southward recirculation originating from the Gulf Stream and North Atlantic Current, and also mixing by eddies all act to transfer nutrients into the subtropical gyre. A more detailed inspection of the model results shows that this lateral nutrient flux has a strong seasonal cycle with almost all of the input taking place in winter when the southward Ekman transport is strongest and when mixing is deep and nitrate concentrations are high to the north (Figure 7).”

Fig. 8 shows the zonal-averaged meridional section of potential density and nitrate from climatology and from the model.

“The results presented support the hypothesis that the underestimation of simulated primary production is not caused by an underestimation in the nitrate supply. The model’s tendency to produce slightly too low SSTs over the subtropical gyre indicates an overestimation of diapycnal mixing in this region. Because of the close correspondence of nitracline and thermocline in the subtropics, it seems unlikely that the model should, by whatever possibly unresolved process, underestimate diapycnal mixing of nitrate into the subtropical gyre’s euphotic zone.”

Additional notes (07/06/2010)

  • See this remark on the difference between horizonal advection and diffusion:

    “Note that the analysis of the simulated horizontal nitrate transport does not distinguish between horizontal advection and horizontal diffusion. In a high-resolution model such a distinction would be of little physical relevance because mixing by geostrophic turbulence is assumed to be resolved partially and the amount of mixing required for numerical stability depends primarily on the grid resolution. Hence any lateral nitrate transport that may arise from horizontal mixing by mesoscale eddies is included in this picture.”

  • Simulated new primary production (via nitrate input into the surface layer) is consistent with the observations. This is true for all models so far, even the low-resolution one. it is true also in the 1/10-of-a-degree simulation of McGillicuddy et al. (2003), although the authors of this papers stress the point that the contribution of the various physical mechanisms responsible for the input of nitrate into the euphotic zone differ between the models.

  • What all models underestimate is the total primary production. Because new primary production is consistent with observations then the underestimation has to be due to an underestimation of regenerated production.

  • The present model does not resolve the discrepancy that observed export production (from oxygen utilization rate) is much higher than new primary production. The horizontal input of organic matter in the model, albeit positive, is too low to explain this discrepancy. Other processes, such as nitrogen fixation, lateral input of dissolved organic nutrients and subduction/oxidation of dissolved organic matter, not included in the model could explain the discrepancy. (Question: is this consistent with the problem that Johnson et al. (accepted, 2010) try to explain?)