Turbulence measurements suggest high rates of new production over the shelf edge in the northeastern North Sea during summer

Bendtsen, Jørgen; Richardson, Katherine

doi:https://doi.org/10.5194/bg-15-7315-2018

Articles | Volume 15, issue 23

https://doi.org/10.5194/bg-15-7315-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/bg-15-7315-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 23

Research article

|

10 Dec 2018

Research article |

| 10 Dec 2018

Turbulence measurements suggest high rates of new production over the shelf edge in the northeastern North Sea during summer

Jørgen Bendtsen and Katherine Richardson

Download

Final revised paper (published on 10 Dec 2018)
Supplement to the final revised paper
Preprint (discussion started on 07 Sep 2018)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review', Anonymous Referee #1, 21 Sep 2018
- AC1: 'AC: Response to the referee 1 comments', Jørgen Bendtsen, 29 Oct 2018
RC2: 'REVIEW of “Turbulence measurements suggest high rates of new production over the shelf edge in the north-eastern North Sea during summer” by Bendtsen and Richardson.', Anonymous Referee #2, 24 Sep 2018
- AC2: 'AC: Response to the referee 2 comments', Jørgen Bendtsen, 29 Oct 2018

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Reconsider after major revisions (30 Oct 2018) by Emilio Marañón

AR by Jørgen Bendtsen on behalf of the Authors (30 Oct 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (01 Nov 2018) by Emilio Marañón

RR by Anonymous Referee #1 (02 Nov 2018)

Suggestions for revision or reasons for rejection

Thank you to the authors for a well-structured response to the earlier review.
Most of the queries I had have been answered/clarified. I have just a few final points:

1. I still do not fully understand how the individual, single-cast turbulence measurements (or averages based on a handful of closely-spaced single profiles) compare to the tidal variability seen at the time series station. Tides are clearly important in determining the variability in mixing, but this has not been fully explained. In Fig. S1 there appears to be sufficient data to calculate an average turbulence profile with a measure of variability (e.g. 95% confidence intervals using a bootstrap technique on the profiles). Then comparing average profiles and confidence intervals based on a handful of closely-spaced stations elsewhere in the survey would be useful in indicating if there were significant contrasts in mixing outside of the tidal variability. This is different to the analysis done on the paired shear probes (which is a nice, robust QC of the data, but doesn’t tell us about tidal variability).

In the revised manuscript (section 3.7, page 13) it is stated that turbulence measurements were not made at night. Fig. S1 suggests they were made at night – assuming the DoY 202.0 is midnight. In Fig. 8 it looks like data were indeed not collected for a short time at night.

On page 14, line 10 – I do not understand why the observation that turbulent dissipation was similar to tidal energy dissipation (1st sentence) then suggests that there must be another source of turbulence (2nd sentence). To me the 2nd sentence simply does not follow on from the claim in the 1st sentence.

Looking at Fig. 5 it does look like additional turbulence in the bottom boundary layer over the shelf edge. Is the claim that this turbulence is not tidal? If not, are there any published reports of slope currents or eddies that could be cited as suggestive of the source of the mixing?

In the supplementary material the vertical turbulence profiles are described as being “3 minutes” apart. But this is not the temporal resolution of the profiles. What is the typical time interval between the start of each profile (i.e. the period of the turbulence profiler yo-yos)?

2. On page 17 it is now more clearly articulated than the relatively strong turbulent flux of nitrate is a consequence of (1) deeper nutricline and (2) “elevated” mixing. But it is not clear that the increased nutrient flux is at the shelf edge. It might be shown in Fig. 6, but I cannot work out where the profiles in Fig. 6 came from. The caption refers to “summarised in the tables (e-h)”. I assume e-h refers to the profiles in Fig. 6, but which “tables”? Neither of the tables in the manuscript have location information. The nitrate flux has a clear peak in profile g, associated with a peak in turbulent dissipation, but it is not clear where on a transect this was. Can the positions of the profiles be marked on one of the earlier sections? Alternatively does this “elevated” shelf edge mixing rest on the evidence of the sections in Fig. 5 (though the relationship to the nutricline is not obvious)?

In general avoid the word “elevated”. It is ambiguous – do you mean “increased” or “raised upward in the water column”? I’m assuming “increased”….better to make this clearer throughout.

3. Line 15, page 17. The largest nitrate fluxes of 1 – 1.5 mmol m-2 day-1 are stated (with reference to Fig. 7), but I cannot see the parameters behind these in any of the data presented. Table 2 shows mean fluxes (it does not clearly say where from in the caption), which are very low, and the maximum value I can see in Fig. 6 (g) is about 0.3 mmol m-2 day-1. There only are a handful of points in Fig 7 with these high values of nitrate flux, so without any other information (e.g. profiles of nitrate and turbulence) there is nothing to support the underlying reasons for these fluxes - they could easily be interpreted as outliers. In other words, what is the link between these high flux points, the patterns of turbulence in Fig. 5 and the flux profiles in Fig. 6? Are you attributing these high fluxes to the increased boundary mixing and deeper nutricline idea (see also point 5 below….)?

The cited values of nitrate fluxes to the SCM from other shelf studies are all daily-means, so make it clear in the text if you are comparing maximum values with daily means, or daily means with daily means.

4. On page 18 it is stated that the maximum nitrate flux was on average found below the nutricline depth (by 6.4 m). Given the vertical resolution of the nitrate samples, this is not significant. Also, the maximum flux is stated as being below the SCM – but if the SCM depth is defined as the maximum chl concentration, this is not surprising. The peak of typical shelf sea SCM is often coincident with the point where nitrate hits zero (within the constraints of the resolution in this study, that looks like a reasonable claim here as well), so the relevant region over which to look at the nitrate flux is generally in the low part of the SCM rather than at the peak. Nitrate reaching zero at the SCM peak then lends support to the statement that recycling must be important further up in the SCM (or, indeed, motility given the very low turbulence in the SCM).

5. The addition of the satellite imagery is really interesting. It looks to me (comparing Figs 1 and 7) that the high values of nitrate flux are coincident with fronts in SST associated with what might be an outflow from the Baltic. I have no idea if this is coincidence or causal, but could this be tentative evidence of another source of flow and shear in the system? In relation to my point 3 above, could the mechanism underpinning these highest fluxes that you see be different from the boundary-driven turbulence plus deeper nutricline idea? The more I look at the data, the more I think there may be two mechanisms behind the pattern of nitrate fluxes: (1) shelf edge boundary mixing combined with deeper nutricline for the “typical” fluxes of 0.1- 0.3 mmol m-2 day-1, and frontal or mesoscale shear associated with the warm surface water from the Baltic driving the larger fluxes (which are patchy in the survey data, but could be quite significant given the scale of the SST signal). So the highest f-ratios might be frontal signatures. Obviously (!) don’t try to open up a new area of analysis – but there may be published work on Baltic outflows to the N Sea that might help justifying this as a reasonable hypothesis in the discussion.

Hide

RR by Anonymous Referee #2 (07 Nov 2018)

Suggestions for revision or reasons for rejection

The authors correctly addressed my main concerns and the revised manuscript was significantly improved. The revised manuscript includes, among others, a more thorough evaluation of the uncertainties associated with the TKE dissipation and nitrate fluxes (extended Figure 6) calculations; new panels in Figures 2 and 7 displaying the distribution of nitrate in transect Tr4, and the distribution of nitrate fluxes across the shelf edge; microstructure and CTD data from a 36-hour time-series, and a Table presenting the median of relevant parameters for the different regions. Those modifications strengthen the communication of the manuscript’s main conclusions. I have only a set of relatively minor points that I would like the authors to address before the manuscript is published.

GENERAL COMMENTS:
I contend that some weaknesses remain in the communication of the main conclusions. See for example the following sentences:

“Estimated nitrate fluxes due to turbulent vertical mixing into the euphotic zone were up to 0.5 - 1 mmol N m-2 d-1 over the shelf-edge (f-ratios > 0.1) while values of < 0.1 mmol N m-2 d-1 were found in the deeper open area north of the shelf-edge.” (Abstract, P.1 L. 13-14)

“Nitrate fluxes were generally very low (< 0.1 mmol N m-2 d-1) in the deeper area north of the shelf edge due to low vertical mixing in the upper 50 m. However, elevated nutrient fluxes of ~1 mmol N m-2 d -1 were seen in the shelf edge zone and near the Norwegian coast. This resulted in f-ratios above 0.10 in these regions compared to < 0.02 for the remainder of the study area (Fig. 7, S28d, e).” (P. 14, L. 10-14)

“Significant NP was found above the shelf-edge where vertical nitrate fluxes of 0.5 - 1.5 mmol N m-2 d-1 implied f-ratios above 0.10, whereas very low nutrient fluxes characterised the open area above the Norwegian Trench (f-ratios < 0.02).” (Conclusions, P. 21, L. 15-17)

Although the nitrate diffusive flux was as high as ~1 mmol m-2 d-1 and f>0.10 in some stations over the shelf edge area, those values seem not to be representative of the “mean state”, according to the median values presented in Table 2. The median flux in the shelf edge area and the corresponding f-ratio (0.11 mmol m-2 d-1, f-ratio = 0.021) are well below the values reported in these sentences. I would suggest to rephrase the sentences to acknowledge that, although values of the f-ratio/nitrate fluxes at individual stations could be as high as 0.1 and 1 mmol m-2 d-1, those values do not correspond to the regional means.

SPECIFIC COMMENTS:
- This sentence “Every cast provided two shear-measurements and, in general, there was relatively small deviation between the two shear probes. Therefore, all measurements were included in the analyses and values from the two shear-probes were averaged.” (P.4, L 20) seems contradictory with the following paragraph (“This led only to the removal of a few pairs of epsilon from the dataset”, P.4, Line 32)

- Although I think that I understand what the authors meant with these two sentences: “ Thus, PBmax was significantly lower at the SCM, in general accordance with previous studies (see review in Richardson et al., 2016) whereas αB showed a weak decrease with depth (with overlapping uncertainty intervals between the two depth levels). In general, αB has been found to increase with depth (resulting in a more efficient photosynthetic response at low light levels) and inspection of the vertical distribution showed at tendency to higher values between 15 -25 m depth (i.e. 2 - 3 e-folding depth of PAR) and lower values below 30 m resulting in a lower median value from SCM level.” (P. 12, L. 15-19); they might also sound a bit contradictory and probably need some clarification.

- In P 14, Line 5: “cf. Fig. 2c,d” I think the authors wanted to refer to Supplementary Figure 2 here. Also, the description of the figure remains extensive in the text. Although it contains some repeated information, I would also agree if the authors decided to keep this figure (or a reduced version of it) in the main manuscript file. My main concern with this figure was that the cross-shelf variability of the nitrate fluxes was nuanced here, but this is now clearly illustrated in Figure 7c.

- I appreciate that the authors extended the discussion about the calculation of photosynthetic parameters and its uncertainties. However, I still would like to see how the fitted curves and the data actually look like, as they are a central point of the manuscript. Could you show some figure as example to me and, maybe, include them in the manuscript or as supplementary information?

-Figure 7 (caption): Could you add the integration depth in the Figure caption as well?

- Table 2 (caption). “...depth intervals of depth,”, please remove one “depth”

Referee Report: PDF

Hide

ED: Publish subject to minor revisions (review by editor) (12 Nov 2018) by Emilio Marañón

AR by Jørgen Bendtsen on behalf of the Authors (24 Nov 2018) Author's response Manuscript

ED: Publish as is (27 Nov 2018) by Emilio Marañón

Download

Article (8137 KB)
Full-text XML

Short summary

New production based on nutrients entering the well-lit surface layer is important for understanding marine ecosystems. Measurements of primary production and turbulence across the shelf edge in the northeastern portion of the North Sea show that new production is concentrated around the shelf-edge zone. The shelf-edge zone is, therefore, a major nutrient supplier to the productive surface layer and makes this area important for higher trophic levels such as zooplankton and fish.