Table Of Contents

This Page

Notes on “Can eddies make ocean deserts bloom?” by Oschlies (2002b)

The abstract is good summary.

“Note, that explaining the sparseness of direct observations of the eddy-pumping process by under-representation of eddies in the in situ observations is problematic: Statistically, there is no reason why in situ measurements should have undersampled (or oversampled) these events. The likelihood of systematic sampling errors should be particularly small for the long-term time-series sites in the subtropical oceans, the Bermuda Atlantic Time-series Study (BATS) and the Hawaii Ocean Time-series (HOT) with typical sampling intervals (monthly to fortnightly) similar to the lifetime of eddy signals [e.g., McNeil et al., 1999].”

“Because sinking is relative to the ambient water and thus constitutes a diapycnal transport, remineralization of organic matter occurs on denser (and locally deeper) isopycnals than its formation (Figure 1). Without recharging nutrient levels of the lighter isopycnals, the long-term effect of eddies would therefore be a net deepening of the nutricline to a level at which eddies cannot anymore raise nutrient-replete isopycnals into the euphotic zone, and eddy-pumping would cease. Irrespective of the level of eddy activity, the bottleneck of upper ocean nutrient supply still is the diapycnal flux that may either occur locally (i.e., vertically) or be associated with remote water mass transformation and subsequent isopycnal (i.e., lateral) transport.” [bold characters are mine]

The model has 2/15th x 1/9th of a deg. in the horizontal with 37 levels in the vertical.

In their Fig. 1, they imply that the time scale of remineralization is fast compared to the time scale of eddy formation and decay. I am not sue this is quite right.

They use BATS to validate the model: “A comparison with BATS data demonstrates relatively good agreement between simulated and observed mean nitrate profiles as well as of the variance of vertical displacements of nitrate surfaces, their correlation with surface height anomalies, and the correlation between nitrate and density (Figure 4).”

“Interestingly, regressing vertical displacements of nitrate surfaces to SSH anomalies can explain about 40% of the variance below about 300 m whereas the explained variance approaches zero near the euphotic zone both in the observations and in the model (Figure 4c). At least near Bermuda, attempts to estimate the eddy-induced nutrient supply from satellite altimetry may therefore be problematic.”

In Fig. 5, the annual cycle of vertical profile of nitrate is plotted with that of isopycnal: there are significant differences between contours of nitrate and potential density. Also shown is the cumulative input of nitrate: the largest contribution is from vertical mixing.

“For the first event, the s0 = 26.1 kg m^3 isopycnal is displaced from a depth of about 190 m to some 60 m at the maximum of the negative SSH anomaly. However, this is not accompanied by a similar vertical excursion of nitrate isolines in the upper 200 m. This loosening of the otherwise tight correlation of nitrate and density [McGillicuddy and Robinson, 1997] just below the euphotic zone indicates that past excursions of this isopycnal into the light-lit surface layer have already reduced its nitrate load. Had we assumed rapid restoring to climatological mean nitrate concentrations (as was done in some previous studies), a total nitrate input of more than 0.05 mol m^2 would be diagnosed from the vertical displacement of isopycnals in the density range 26.0–26.18 kg m^3 into the euphotic zone during eddy event 1.”

About observations of the North Pacific:

Abell, J., S. Emerson, and P. Renaud, Distributions of TOP, TON and TOC in the North Pacific subtropical gyre: Implications for nutrient supply in the surface ocean and remineralization in the upper thermocline, J. Mar. Res., 58, 203–222, 2000.

With higher resolution, nitrate supply decreases at mid-latitudes mainly because of a more realistic and shallower mixed layer. It decreases also in the subtropics due to “a reduction of spurious upwelling along the American coast”; in the interior, there is no change with the lower resolution model.

“Near Bermuda, a comparison with previous estimates of the nutrient budget also bears some surprises: In this region the model predicts an average eddy-induced nitrate supply of only 0.029 mol m^2/yr (of which 0.027 mol m^2/yr is due to vertical advection). This is only about 15% of previous estimates based on methods that restore to climatological nutrient concentrations below the euphotic zone [McGillicuddy and Robinson, 1997; Siegel et al., 1999]. Applying the method of Siegel et al. [1999] (also used by McGillicuddy et al. [1998]) to infer nutrient supply from satellite altimetry and statistics derived from in situ data to the (1/9) deg. model output turns out to overestimate the ‘‘true’’ simulated eddy-induced nitrate supply by more than 600%. In fact, this overestimate (0.18 mol m^2/yr) is very close to the original estimates derived from satellite and in situ data.”