the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 06 Jan 2023
Satellite data reveal earlier and stronger phytoplankton blooms over fronts in the Gulf Stream region
Abstract. Fronts affect phytoplankton growth and phenology by locally reducing stratification and increasing vertical nutrient supply. Biomass peaks at fronts have been observed in-situ and linked to local nutrient upwelling, and reduced stratification over fronts has been shown to induce earlier blooms in numerical models. However observation of these biophysical interactions through satellite imagery have been scarce, despite the opportunity to quantify them at synoptic scales. Here we used twenty years of Sea Surface Temperature (SST) and Chlorophyll-a satellite data in a large region surrounding the Gulf Stream to quantify the impact of fronts on phytoplankton in contrasting regimes, from oligotrophy to bloom, and throughout the year. We computed an Heterogeneity Index (HI) from SST, and used it to sort fronts into weak and strong fronts based on HI thresholds. We observed that the localization of strong fronts corresponded to western boundary current fronts, and weak fronts to more ephemeral submesoscale fronts. We compared Chlorophyll-a distributions over strong fronts, weak fronts and outside of fronts. We assessed three metrics, the local enhancement of Chlorophyll-a over fronts, the global enhancement of Chlorophyll-a due to fronts at the scale of the region, and the lag in spring bloom onset due to fronts. We found that weak fronts lead to a local enhancement of Chlorophyll-a weaker than strong fronts, but because they are also more frequent they contribute equally to the regional Chlorophyll-a budget. We also find the the local enhancement of Chlorophyll-a was two to three times larger for the spring bloom than in the oligotrophic subtropical gyre. We also provide observational evidence that blooms start earlier over fronts, by one to two weeks. Nevertheless our results suggest that the spectacular impact of fronts at the local scale may be misleading, considering their impact on a regional scale budget remains limited.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(1271 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1271 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-1489', Anonymous Referee #1, 18 Feb 2023
The manuscript by Haeck et al., presents an analysis of the amplification of surface chlorophyll along weak and strong fronts (detected from satellite SST) in the Western North Atlantic region. The results show significant amplification of chlorophyll at the location of SST fronts, which, when averaged over the domain where fronts are found, increases the total chlorophyll by up to 5% on average over broad biomes. Furthermore, spring blooms are observed to occur earlier — by up to two weeks — along fronts in the subpolar biome.
Heterogeneity in the marine environment is a topic of great interest from physical to biological oceanography. Fronts and eddies usually draw a lot of interest, but the actual impacts on marine ecosystems remain poorly quantified. The manuscript by Haeck et al. thus sheds light on a topic of current relevance.
I found the manuscript very well written and the analysis thorough and sound. This is clearly the result of a lot of work, and the results are not only useful (because they address long standing questions), but also very stimulating. I can imagine a global extension of the approach, which would further advance our understanding of physical-biogeochemical coupling in the ocean, and inform ecological thinking. I recommend publication of the manuscript with only minor suggestions.
As a general criticism, I somewhat disagree with the negative spin on the results by the Authors. While 5 % amplification over large scales may not seem that large, it may still have important implications for ecology; after all, chlorophyll amplification can be significant along fronts (E.g. Fig. 8f,h); early onset of blooms along front may also be important for organisms phenology. Furthermore, the study quantifies only one aspect of the effects of front on phytoplankton — i.e., is the direct chlorophyll amplification; as discussed in the text, also because of the ephemeral nature of fronts, nutrients upwelled along fronts may be dispersed more broadly and contribute to an average amplification of chlorophyll that may not be directly co-located with fronts. Of course, this effect is hard, if not impossible to detect purely from remote-sensing. But in few instances in the abstract and conclusions the Authors could be more specific in stating that they quantify the specific effect of biomass amplification at fronts, not other possibly more widespread effects. Additionally, SST is an imperfect proxy for the expected phytoplankton response along fronts, which may lag more than a week relative to the excess of recently upwelled nutrients that may co-occur with cooler SSTs. While these caveats are discussed in the manuscript, they could be somewhat better highlighted.
Specific comments:
Line 11, “the global enhancement of Chlorophyll-a due to fronts” maybe clarify as “the global enhancement of Chlorophyll-a along fronts”
Line 16, “misleading”: I would use more nuance here, and avoid this term.
Line 17, “budget”: maybe “biomass amplification” or “chlorophyll amplification” would be more appropriate, since technically a budget (which implies some balance of different sources and sinks) has not been evaluated. See also “budget” in line 55 — at least clarify what the term means in this context.
Line 71: this sentence seems a bit obvious; it could be removed.
Line 122, “All pixels where water depth is less than 1500m are masked to exclude the continental shelf.”. This needs at least a sentence to justify the removal of shelf waters.
“Front detection” section. This is well described and builds nicely on previous work. I would only advise to clarify better why some choices were made and how different choices may affect the results; showing or stating that results are not very sensitive to specific thresholds or changes from previous methods would suffice.
Lines 204-206: I was somewhat confused by how the lag “L” was defined; maybe add a sentence to clarify its definition.
Lines 246-249: This seems an important point; lack of co-location (in space or time) of SST fronts and Chl maxima may be the consequence of interesting dynamical reasons, e.g. related to the timescales of phytoplankton response vs. the physical lifetime of a front. This could be discussed.
Lines 262-264: this entire paragraph doesn’t seem necessary; it could be removed to streamline the paper. In general the paper is on the long side, so some streamlining may help.Lines 306-307: Very interesting to observe the negative effect of fronts in summer, likely related to subduction as discussed later. This is a nice result.
Lines 315-319, “To quantify …”: this part may belong to Methods; it could also be clarified.
Lines 351-353: the other effect not quantified here is the broader supply and re-distribution of nutrients that may be caused by fronts; i.e., additional upwelled nutrients may not remain confined to the front that upwelled them, and be able to fertilize phytoplankton more broadly.
Line 385, “what is generally thought”: maybe add a reference, or it risks to be a straw man argument.
Also, related to the greater impact of fronts in bloom regions relative to oligotrophic regimes, the study by Yamamoto et al. (2018, Journal of Geophysical Research) provided (based on a mesoscale-eddy permitting model) evidence that the majority of nutrient supply to the euphotic zone of the oligotrophic gyres occurs by lateral eddy fluxes rather than vertical eddy fluxes. This certainly relates to the argument of deep nutriclines in these regions.
Lines 420-424: a clear sense of why blooms occur early along fronts is a bit hidden in this explanation, maybe clarify a bit.
Line 429: I think the results of early blooms on fronts may be important for phenology of zooplankton, and potential their predators, which could be more explicitly discussed.
Line 444, “due to fronts”: again here, it may be better to add more nuance, since only a direct effect of fronts on co-located chlorophyll was quantified here; indirect effects (e.g. nutrient redistribution) may be also important.
Technical comments:Line 14: “the the” —> “the”
Line 64, “spacial” —> “spatial”
Line 70, “built” —> “build”
Line 82, “contrasted” —> “contrasting”
Line 115: “tends to provide an underestimate of” —> “tends to underestimate”
Line 139: “Gulf stream” —> “Gulf Stream”
Line 276: “and this throughout” —> “and this holds throughout”
Line 326, “one months” —> “one month”
Caption of fig. 10, and other parts of the text: the use of “rest.” May not be completely clear, maybe use the full word (respectively?)
Line 429, “budget” —> “budgets”Citation: https://doi.org/10.5194/egusphere-2022-1489-RC1 -
AC1: 'Reply on RC1', Clément Haëck, 16 Mar 2023
RC1: The manuscript by Haeck et al., presents an analysis of the amplification of surface chlorophyll along weak and strong fronts (detected from satellite SST) in the Western North Atlantic region. The results show significant amplification of chlorophyll at the location of SST fronts, which, when averaged over the domain where fronts are found, increases the total chlorophyll by up to 5% on average over broad biomes. Furthermore, spring blooms are observed to occur earlier — by up to two weeks — along fronts in the subpolar biome.
Heterogeneity in the marine environment is a topic of great interest from physical to biological oceanography. Fronts and eddies usually draw a lot of interest, but the actual impacts on marine ecosystems remain poorly quantified. The manuscript by Haeck et al. thus sheds light on a topic of current relevance.
I found the manuscript very well written and the analysis thorough and sound. This is clearly the result of a lot of work, and the results are not only useful (because they address long standing questions), but also very stimulating. I can imagine a global extension of the approach, which would further advance our understanding of physical-biogeochemical coupling in the ocean, and inform ecological thinking. I recommend publication of the manuscript with only minor suggestions.AC1: We thank the reviewer for their positive appreciation of our work
RC: As a general criticism, I somewhat disagree with the negative spin on the results by the Authors. While 5 % amplification over large scales may not seem that large, it may still have important implications for ecology; after all, chlorophyll amplification can be significant along fronts (E.g. Fig. 8f,h); early onset of blooms along front may also be important for organisms phenology. Furthermore, the study quantifies only one aspect of the effects of front on phytoplankton — i.e., is the direct chlorophyll amplification; as discussed in the text, also because of the ephemeral nature of fronts, nutrients upwelled along fronts may be dispersed more broadly and contribute to an average amplification of chlorophyll that may not be directly co-located with fronts. Of course, this effect is hard, if not impossible to detect purely from remote-sensing. But in few instances in the abstract and conclusions the Authors could be more specific in stating that they quantify the specific effect of biomass amplification at fronts, not other possibly more widespread effects. Additionally, SST is an imperfect proxy for the expected phytoplankton response along fronts, which may lag more than a week relative to the excess of recently upwelled nutrients that may co-occur with cooler SSTs. While these caveats are discussed in the manuscript, they could be somewhat better highlighted.
AC: We thank the reviewer for their constructive criticism. We agree on all points raised and will discuss them in the revised manuscript:
- Regarding the negative spin on the results: in the context of previous observational and modeling studies that have shown very large amplifications at ocean fronts, our initial intention was to put them into a broader context, and to show that these large local amplifications contribute modestly to regional scale amplification. We agree we might have fallen too much on the negative side, and we propose a few modifications in the text to give some more insights on the widespread effects. For instance, we propose to change the last sentence of the abstract into something more nuanced (see answer to specific comment line 16).
- that 5% amplification of surface Chl might lead to greater amplification in terms of ecological implications (Stock et al. 2014, Lotze et al. 2019). We will mention these two papers to illustrate that point in our discussion (section 4.4):
- Stock CA, Dunne JP, John JG. 2014. Drivers of trophic amplification of ocean productivity trends in a changing climate. Biogeosciences. 11(24):7125–35
- Lotze HK, Tittensor DP, Bryndum-Buchholz A, Eddy TD, Cheung WWL, et al. 2019. Global ensemble projections reveal trophic amplification of ocean biomass declines with climate change. Proc. Nat. Ac. Sc. 1:201900194
We will also add in the plain language summary: “Nevertheless these relatively modest effects in term of surface chlorophyll might have more significant impacts for the rest of the ecosystem.” - that nutrient enrichments due to fronts might lead to phytoplankton biomass enhancement outside of fronts.
We agree and this result is actually seen for example in our Figure 2, as the Chl-a filament expands outside the area of elevated HI in the north west part of the domain. We will use this example to discuss that our estimate is a lower limit because the enrichments due to fronts may be seen outside of fronts. We will expand the discussion on how these enrichments outside of fronts are integrated in our analysis and how much they represent by showing an additional supplementary figure with the Chl-a distributions. We will expand the caveat section of the discussion. - that the mismatch between Chla and HI (seen in figure 2) may be due to the time evolution of the flow, to diffusion, and to the time lag between nutrient supplies and Chl increase.
These aspects will be discussed when Figure 2 is described, and we will come back to them in the final discussion.
RC: Specific comments:
Line 11, “the global enhancement of Chlorophyll-a due to fronts” maybe clarify as “the global enhancement of Chlorophyll-a along fronts”AC: We will rephrase the sentence as it is not exactly the enhancement along fronts, but really, the effect of fronts when integrated over a large region:
“… the local enhancement over fronts, the global enhancement that it implies at the regional scale, and...”RC: Line 16, “misleading”: I would use more nuance here, and avoid this term.
AC: We agree and propose to rephrase as follows: “Nevertheless, our results suggest that the spectacular impact of fronts at the local level is more limited when considered at the regional level”.
RC: Line 17, “budget”: maybe “biomass amplification” or “chlorophyll amplification” would be more appropriate, since technically a budget (which implies some balance of different sources and sinks) has not been evaluated. See also “budget” in line 55 — at least clarify what the term means in this context.
AC: We agree and will remove the term budget everywhere in the paper, and replace it with amplification.
RC: Line 71: this sentence seems a bit obvious; it could be removed.
AC: We agree and will shorten it
RC: Line 122, “All pixels where water depth is less than 1500m are masked to exclude the continental shelf.”. This needs at least a sentence to justify the removal of shelf waters.
AC: We agree and propose to add the following sentence: “In addition to these three biomes, other biomes are found on the continental shelf but were excluded from this study that focuses on open ocean conditions”.
RC: “Front detection” section. This is well described and builds nicely on previous work. I would only advise to clarify better why some choices were made and how different choices may affect the results; showing or stating that results are not very sensitive to specific thresholds or changes from previous methods would suffice.
AC: Indeed, on line 167 we say that we performed sensitivity tests on the parameter choices with no further comments. And it is only from line 280 that we say that our results are very weakly sensitive to the choice of parameters (with a reference to supplementary Fig. A1 which actually shows it). We will bring these two paragraphs together.
RC: Lines 204-206: I was somewhat confused by how the lag “L” was defined; maybe add a sentence to clarify its definition.
AC: We agree and we propose to change this entire paragraph in the method section, by incorporating also your other comment regarding lines 315-319 that are moved to the method section:
“Finally, the time series of the Chl-a median in the subpolar biome is characterized by a spring bloom, of which we measured the timing (onset date) both in the fronts and in the background, for each year. Because the spring bloom onsets propagates from south to north, these timings were inferred over latitudinal band of width 5°. Our estimates are based on averaged statistics over Eulerian time series, assuming that the bloom evolves coherently in the background (resp. over fronts) within each latitudinal bands in the subpolar biome. Thus we are assuming that front and non front pixels can be pooled apart to follow the bloom evolution over the two contrasting environments. This assumption is suggested by high-resolution models of the bloom (e.g. Lévy et al., 2005; Karleskind et al., 2011), but it is imperfect as the bloom evolves along Lagrangian trajectories. Nevertheless, we found very consistent results with this approach. We define the lag in bloom onset L as the bloom onset day in the fronts minus the bloom onset day in the background.
To extract the onset date, we filtered the Chl-a median time series with a low-pass Butterworth filter of order 2 and cutting frequency 1/20days−1. The filtered time-series displayed strong variations in their phenology from year to year, but a bloom was always discernible. We considered data from February to July, which allowed us to isolate the spring bloom and exclude the autumn bloom. First, we detected the maximum value of Chl-a in this time window, then defined the bloom onset as the time of maximum Chl-a derivative prior to the time of maximum Chl-a.
We defined the uncertainty in bloom onset date as the standard deviation of all days for which the Chl-a derivative is above 90% of its maximum value. We defined the uncertainty in lag L as the square root of the sum of squared uncertainties in fronts and background. Finally, we defined the mean lag (over the 20 years of data and for each latitudinal band) as the weighted averaged lag, with the weights equal to the inverse of the uncertainties; and we defined the mean lag uncertainty as the weighted standard deviation of all lag values around the mean lag, with the weights equal to the inverse of the uncertainties.“RC: Lines 246-249: This seems an important point; lack of co-location (in space or time) of SST fronts and Chl maxima may be the consequence of interesting dynamical reasons, e.g. related to the timescales of phytoplankton response vs. the physical lifetime of a front. This could be discussed.
AC: Yes we agree, this relates to one of your earlier comments, and we will discuss it here. We propose to add the following paragraph, and to provide an additional supplementary figure to show how the distributions of Chl differ between fronts and the background (attached here):
“the mismatch between the exact location of fronts at a given time and the exact location of the Chl-a response at the same time may be due to different factors, such as the very dynamic nature of fronts which can lead to the chaotic advection of phytoplankton (and/or nutrients) outside of fronts, the time-scale needed for phytoplankton to respond to nutrient supplies at fronts, or trophic interactions at the front (as shown, for instance, in the modelling study of Mangolte et al. 2022, their Figure 1d). This implies that there might be remnant of large Chl-a values due to frontal activity outside of fronts. Nevertheless, our results show that on average, the distribution of Chl-a at a given time differs over fronts and outside of fronts (supplementary figure), and also depends on the strength of fronts.“
RC: Lines 262-264: this entire paragraph doesn’t seem necessary; it could be removed to streamline the paper. In general the paper is on the long side, so some streamlining may help.
AC: We agree – the paragraph will be removed.
RC: Lines 306-307: Very interesting to observe the negative effect of fronts in summer, likely related to subduction as discussed later. This is a nice result.
AC: Thanks. We feel that this result is maybe not highlighted enough and will make it appear more clearly in the abstract and in the discussion of the revised version.
RC: Lines 315-319, “To quantify …”: this part may belong to Methods; it could also be clarified.
AC: Indeed, lines 315-321 will be moved to method and clarified. This relates to your comment regarding the computation of the lag L line 204-206. We will rewrite entirely the paragraph that goes from line 196-206 that explains the computation of the bloom timing (see above)
RC: Lines 351-353: the other effect not quantified here is the broader supply and re-distribution of nutrients that may be caused by fronts; i.e., additional upwelled nutrients may not remain confined to the front that upwelled them, and be able to fertilize phytoplankton more broadly.
AC: Indeed, in addition, we will complete this paragraph and add one reference:
“Moreover, our assessment of the effect of fronts on phytoplankton, based on surface Chl-a, is probably a lower estimate. First, it only measures the excess Chl-a at the surface but the biological signal due to the upwelling of nutrients at fronts often does not reach the surface and is more intense at sub-surface (Mourino et al., 2004; Ruiz et al., 2019). Secondly, the excess Chl-a due to fronts is not always located exactly over fronts because of chaotic advection during the time needed for phytoplankton to respond to nutrient enrichments at fronts (Abraham, 1998).“- Abraham E. 1998. The generation of plankton patchiness by turbulent stirring. Nature. 391(6667):577–80
RC: Line 385, “what is generally thought”: maybe add a reference, or it risks to be a straw man argument.
AC: Indeed, we will remove this statement and add:
“Finally, an important outcome of this study is that the amplification in phytoplankton biomass at fronts can be stronger in bloom regimes than in oligotrophic regimes. Indeed, in eutrophic regions, where nutrients are not limiting productivity, fronts have generally been shown to subduct excess nutrients (Oschlies, 2002; Gruber at al., 2011) and excess biomass (Lathuiliere et al., 2010) rather than to increase biomass as in oligotrophic regions. This effect of a decreased biomass, suggestive of subduction, is observed here in the northern parts of the subpolar biome between May and September (Fig. 8b-d). But we also find that in the subpolar biome, the Chl-a excess over fronts can reach 150% during the bloom (Fig. 8f), and 50% during summer (Fig. 8h), while in the permanent subtropical biome it never exceeds 10% (Fig. 6b).”
- Oschlies A. 2002. Can eddies make ocean deserts bloom. Glob. Biogeochem. Cyc. 16(4):1106
- Gruber N, Lachkar Z, Frenzel H, Marchesiello P, Münnich M, et al. 2011. Eddy-induced reduction of biological production in eastern boundary upwelling systems. Nature Geoscience. 4(11):787–92
- Lathuiliere C, Lévy M, Echevin V. 2010. Impact of eddy-driven vertical fluxes on phytoplankton abundance in the euphotic layer. J. Plankton Res. 33:827–31RC: Also, related to the greater impact of fronts in bloom regions relative to oligotrophic regimes, the study by Yamamoto et al. (2018, Journal of Geophysical Research) provided (based on a mesoscale-eddy permitting model) evidence that the majority of nutrient supply to the euphotic zone of the oligotrophic gyres occurs by lateral eddy fluxes rather than vertical eddy fluxes. This certainly relates to the argument of deep nutriclines in these regions.
AC: Yes, we agree and this is a very important point. We thank you and the second reviewer for raising it. We will change our introduction, discussion and conclusion to highlight that increased Chl-a over fronts is not necessarily the result of vertical advection but may also be related to lateral transport and particularly to the nutrient stream (as our results actually show evidence for it, and we saw it thanks to your feedback).
• We will replace “vertical nutrient supply” by “nutrient supply” in the first sentence of the abstract
• In the 2nd line of the abstract, we will add “linked to local nutrient upwelling and/or to horizontal transport”
• Abstract, line 15, we will add “the local enhancement of Chl-a was particularly strong in the latitude band of the Gulf Stream, likely reflecting the lateral supply of nutrients through the nutrient stream”
• We will revise the second paragraph of the introduction. line 42 we will add: “In addition, western boundary currents like the Gulf Stream (Peligry et al., 1996) and the Kuroshio (Long et al., 2022) are also the locus of intense nutrient transport over long distances, or nutrient streams. We will also remove lines 43 to 46
• Introduction line 78, we will also refer to the nutrient stream as a potential process
• Method section line 132: we will add that the Gulf Steam conveys nutrients and that the nutrient stream falls within the seasonal subtropical biome
• Discussion section 4.2 will be strongly modified to discuss the idea that the stronger local increase over strong fronts in the intermediate region likely reflects the lateral transport of nutrients by the nutrient stream. We will also discuss lateral eddy fluxes of nutrients which are more important at the gyre boundaries (Yamamoto et al. 2018, Spingys et al., 2021). We will make it clear that the local increase of Chl-a over fronts is not only the consequence of local upwelling of nutrients but also of lateral eddy fluxes of nutrients.- Long Y, Guo X, Zhu X-H, Li Z. 2022. Nutrient streams in the North Pacific. Prog. Oceanogr. 202:102756
- Pelegrí JL, Csanady GT, Martins A. 1996. The North Atlantic nutrient stream. Journal of oceanography. 52(3):275–99
- Yamamoto A, Palter JB, Dufour CO, Griffies SM, Bianchi D, et al. 2018. Roles of the Ocean Mesoscale in the Horizontal Supply of Mass, Heat, Carbon, and Nutrients to the Northern Hemisphere Subtropical Gyres. J. Geophys. Res. Ocean. 123(10):7016–36
- Spingys CP, Williams RG, Tuerena RE, Garabato AN, Vic C, et al. 2021. Observations of Nutrient Supply by Mesoscale Eddy Stirring and Small-Scale Turbulence in the Oligotrophic North Atlantic. Global Biogeochemical Cycles. 35(12):e2021GB007200RC: Lines 420-424: a clear sense of why blooms occur early along fronts is a bit hidden in this explanation, maybe clarify a bit.
AC: We propose this new formulation:
“Several field and modeling studies have shown that frontal dynamics, by tilting existing horizontal density gradients, increase the vertical stratification of surface mixed layers (Taylor and Ferrari, 2011), which can lead to the stratification of the mixed layer prior to seasonal stratification. Given that the surface spring bloom is triggered by increased stratification, this effect can cause earlier local phytoplankton blooms over fronts compared to surrounding areas”RC: Line 429: I think the results of early blooms on fronts may be important for phenology of zooplankton, and potential their predators, which could be more explicitly discussed.
AC: Indeed, we propose to add this comment:
“This effect alone is unlikely to affect primary productivity, but may impact phytoplankton competition at the onset of the bloom season, affect their predators, and the phenology of zooplankton.“RC: Line 444, “due to fronts”: again here, it may be better to add more nuance, since only a direct effect of fronts on co-located chlorophyll was quantified here; indirect effects (e.g. nutrient redistribution) may be also important.
AC: Yes, we propose to use “associated with fronts” instead of “due to fronts”
RC: Technical comments:
Line 14: “the the” —> “the” Line 64, “spacial” —> “spatial” Line 70, “built” —> “build” Line 82, “contrasted” —> “contrasting” Line 115: “tends to provide an underestimate of” —> “tends to underestimate” Line 139: “Gulf stream” —> “Gulf Stream” Line 276: “and this throughout” —> “and this holds throughout” Line 326, “one months” —> “one month” Caption of fig. 10, and other parts of the text: the use of “rest.” May not be completely clear, maybe use the full word (respectively?) Line 429, “budget” —> “budgets”
Citation: https://doi.org/10.5194/egusphere-2022-1489-RC1AC: Thanks again for your time in providing these very useful and detailed comments
Attached: new supplementary figure showing the distributions of Chl-a by season, biome, and front type.
-
AC1: 'Reply on RC1', Clément Haëck, 16 Mar 2023
-
RC2: 'Comment on egusphere-2022-1489', Anonymous Referee #2, 01 Mar 2023
This is the review of “Satellite data reveal earlier and stronger phytoplankton blooms over fronts in the Gulf Stream region”, by Haëck et al.
The authors build on the methods from Levine et al. 2016 to investigate the impact of fronts—as diagnosed by a heterogeneity index (HI) based on sea surface temperature—on surface chlorophyll-a—as estimated by remote sensing—in different biomes in the Western North Atlantic. Namely, a permanently subtropical, a seasonally subtropical and a subpolar biome. Based on the HI, they estimate the occurrence of weak (5<HI<10) and strong (HI>10) fronts in each biome. Regions with HI less than 5 were considered to be background. The authors noted that high HI values were associated with quasi-permanent fronts while the weak fronts were more ephemeral. Data resolution (4 km) allows the authors to estimate HI in a square of around 24 km (49 grid points).
The authors find that fronts increase surface phytoplankton by about 5% over the region, which they claim is a “much weaker than expected” impact ( although the authors are not explicit about the sources of the high expectations). Secondly, the authors claim that the nutrient supply at fronts more than doubled chlorophyll concentrations during the spring bloom Thirdly, the authors claim that the spring bloom over fronts occur one to two weeks earlier than in background regions.
I find the study interesting. It is well written, well organized and the methodology is sound. I do, however, have a few comments which I believe should be clarified before publication.
Although it is not explicitly stated, the hypothesis the authors are testing is : SST fronts impact surface ocean chlorophyll via submesoscale vertical advection due to unbalanced motions. By stating that the the impact of SST fronts is “much weaker than expected”, they seem to reject that hypothesis.
The premise is that the impact of SST fronts on surface chlorophyll occurs via submesoscale motions (e.g. ageostrophic secondary circulations due to frontogenesis.). By supplying nutrients into a nutrient-depleted surface layer via submesoscale vertical advection, biomass would increase. The problem with this premise is that it is only valid in the permanently oligotrophic region. In the seasonally subtropical and the subpolar biomes, where fronts are more frequent and stronger, lateral advection plays a much more important role due to large-scale climatological horizontal gradients in surface properties. In the Gulf Stream region, not only is the current know to be a “nutrient stream” (Pelegri and Csanady, 1991), with robust horizontal transport of nutrients, but also the front is maintained by the thermal wind shear. There certainly is imbalance and ageostrophic motion, but I am not sure if that is truly discernible from the submesoscale instabilities the authors seem to be mostly concerned about.
In addition, it is unclear whether or not the impact of mesoscale eddies on chlorophyll is removed from the analysis by considering a 24 km square. It is known that the curvature induced by mesoscale eddies impact chlorophyll-a distribution (Siegel et al. 2011). While the analysis was performed on a larger scale in the case of Siegel et al. 2011, it is unclear if the mesoscale signal is removed from the current analysis. Could, for example, some of the detected frontal regions be associated with the edges of mesoscale eddies?
With regards to the permanently stratified, there’s observational evidence (Johnson et al. 2010) that episodic injections, due to meso- or submesoscale processes induce vertical advection of nutrients but that these nutrients get consumed before reaching the surface, so the signal may not be captured by SST in these regions. In addition, there may be photo-inhibition which prevents phytoplankton to be near the surface in oligotrophic regions, even though they may be impacted by frontal motions. The authors partially mention that in lines 351-353, but the studies they cite make use of the QG-omega equation to estimate the vertical velocities, which may be on the mesoscale end of the frontal structures.
Lastly, it is not clear how the authors reconcile the “considerable evidence that submesoscale motions influence nutrient and light environments” with the “smaller than expected” impact of fronts on chlorophyll-a. In addition, conclusions from this work seem to differ from those of Lie and Levine 2016, which showed a 40% increase in the winter. Is it only due to the different resolution in the SST and chlorophyll-a products?
Specific comments:
Title: “earlier and stronger” than what?
Lines 29-31: “largely explained by consistent physical forcing and environmental conditions” - not clear what the authors are referring to. Aren’t the environmental conditions a consequence of the physical forcing?
Line 31 : “wind-driven circulation” : the authors could be more explicit in saying that negative wind stress curl induces downwelling in the subtropical gyres while the opposite occurs on subplolar gyres.
LIne 149: the authors seem to differentiate the “atmospherically and topographically”-controlled fronts from “submesoscale” fronts. This distinction is not so straightforward as submesoscale
Figure 2c: I suggest a qualitative or discreet colorbar for the different categories of HI. In other words, what matters to the reader are the intervals, not the continuous values.
Line 197: typo. “Onsets propagates”. ~Remove “s”.
LIne 221: It is also consistent with baroclinic instability, which is not a submesoscale process.
Line 237: typo : add “s” to “front”.
Lines 243-244: not clear. Are you saying that strong fronts are counted as weak fronts during summer? Isn’t the methodology robust enough to have a consistent classification?
Line 249: I assume this does not affect the interpretation of regions affected by fronts right? Otherwise, it would required a more detailed analysis of points with high chlorophyll-a in regions of low HI.
References
Johnson, K., Riser, S. & Karl, D. Nitrate supply from deep to near-surface waters of the North Pacific subtropical gyre. Nature 465, 1062–1065 (2010). https://doi.org/10.1038/nature09170
Pelegrí, J. L., and Csanady, G. T. (1991), Nutrient transport and mixing in the Gulf Stream, J. Geophys. Res., 96( C2), 2577– 2583, doi:10.1029/90JC02535.
Siegel, D. A., Peterson, P., McGillicuddy, D. J., Maritorena, S., and Nelson, N. B. (2011), Bio-optical footprints created by mesoscale eddies in the Sargasso Sea, Geophys. Res. Lett., 38, L13608, doi:10.1029/2011GL047660.
Citation: https://doi.org/10.5194/egusphere-2022-1489-RC2 -
AC2: 'Reply on RC2', Clément Haëck, 16 Mar 2023
RC2: This is the review of “Satellite data reveal earlier and stronger phytoplankton blooms over fronts in the Gulf Stream region”, by Haëck et al.
The authors build on the methods from Levine et al. 2016 to investigate the impact of fronts—as diagnosed by a heterogeneity index (HI) based on sea surface temperature—on surface chlorophyll-a—as estimated by remote sensing—in different biomes in the Western North Atlantic. Namely, a permanently subtropical, a seasonally subtropical and a subpolar biome. Based on the HI, they estimate the occurrence of weak (5<HI<10) and strong (HI>10) fronts in each biome. Regions with HI less than 5 were considered to be background. The authors noted that high HI values were associated with quasi-permanent fronts while the weak fronts were more ephemeral. Data resolution (4 km) allows the authors to estimate HI in a square of around 24 km (49 grid points).
The authors find that fronts increase surface phytoplankton by about 5% over the region, which they claim is a “much weaker than expected” impact (although the authors are not explicit about the sources of the high expectations). Secondly, the authors claim that the nutrient supply at fronts more than doubled chlorophyll concentrations during the spring bloom. Thirdly, the authors claim that the spring bloom over fronts occur one to two weeks earlier than in background regions.
I find the study interesting. It is well written, well organized and the methodology is sound. I do, however, have a few comments which I believe should be clarified before publication.
Although it is not explicitly stated, the hypothesis the authors are testing is: SST fronts impact surface ocean chlorophyll via submesoscale vertical advection due to unbalanced motions. By stating that the the impact of SST fronts is “much weaker than expected”, they seem to reject that hypothesis.
The premise is that the impact of SST fronts on surface chlorophyll occurs via submesoscale motions (e.g. ageostrophic secondary circulations due to frontogenesis). By supplying nutrients into a nutrient-depleted surface layer via submesoscale vertical advection, biomass would increase. The problem with this premise is that it is only valid in the permanently oligotrophic region. In the seasonally subtropical and the subpolar biomes, where fronts are more frequent and stronger, lateral advection plays a much more important role due to large-scale climatological horizontal gradients in surface properties. In the Gulf Stream region, not only is the current know to be a “nutrient stream” (Pelegri and Csanady, 1991), with robust horizontal transport of nutrients, but also the front is maintained by the thermal wind shear. There certainly is imbalance and ageostrophic motion, but I am not sure if that is truly discernible from the submesoscale instabilities the authors seem to be mostly concerned about.AC2: We are very grateful to the reviewer for their positive assessment of our work, and especially for their sound criticism that we should have examined and discussed the role of lateral advection more thoroughly. This was indeed missing and is an important part of the story. We agree that particularly in the two biomes that are more affected by the Gulf Stream, it is very likely that the signal that we detect is due to the lateral supply of nutrients by the Gulf Stream. This view can explain why the signal is maximum and so much stronger than elsewhere at around 40°N in our figure 9a. We propose to add this in our discussion and highlight it in our results and abstract (see also our answer AC1 to referee 1 on this same point).
RC: In addition, it is unclear whether or not the impact of mesoscale eddies on chlorophyll is removed from the analysis by considering a 24 km square. It is known that the curvature induced by mesoscale eddies impact chlorophyll-a distribution (Siegel et al. 2011). While the analysis was performed on a larger scale in the case of Siegel et al. 2011, it is unclear if the mesoscale signal is removed from the current analysis. Could, for example, some of the detected frontal regions be associated with the edges of mesoscale eddies?
AC: We believe that considering a 24 km square removes a large part of the mesoscale signal from the analysis since mesoscale eddies in this region have diameters roughly between 50 and 200 km. But indeed, mesoscale and submesoscale dynamics are often strongly intertwined, and eddy edges can be considered as small frontal areas and can behave as such. We will add a paragraph to discuss what is captured by our analysis in the method section.
RC: With regards to the permanently stratified, there’s observational evidence (Johnson et al. 2010) that episodic injections, due to meso- or submesoscale processes induce vertical advection of nutrients but that these nutrients get consumed before reaching the surface, so the signal may not be captured by SST in these regions. In addition, there may be photo-inhibition which prevents phytoplankton to be near the surface in oligotrophic regions, even though they may be impacted by frontal motions. The authors partially mention that in lines 351-353, but the studies they cite make use of the QG-omega equation to estimate the vertical velocities, which may be on the mesoscale end of the frontal structures.
AC: Indeed, in Mourino et al., they lacked the full resolution to estimate the submesoscale W, but we wanted to mention this reference because of the vertical profiles of Chl (their Fig4) which showed Chl-a max at 100m, with finescale biomass enhancement at 75m. In Ruiz et al. (2019) their observations are complemented by a high-resolution model with strong submesoscale W where the Chl-a enhancement is also at subsurface (close to 40 m). In these two cases, the enhancement cannot be seen with surface data. We will extend the discussion in lines 351-353 to include these points. In addition, going back to an “old” submesoscale model that looked at nutrient injection in the permanently stratified oligotrophic gyre, the enhancement was much more pronounced at subsurface (Levy et al., 2001, Fig. 9f) but was nevertheless also seen at the surface.
- Lévy M, Klein P, Treguier A. 2001. Impact of sub-mesoscale physics on production and subduction of phytoplankton in an oligotrophic regime. J. Mar. Res. 59(4):535–65RC: Lastly, it is not clear how the authors reconcile the “considerable evidence that submesoscale motions influence nutrient and light environments” with the “smaller than expected” impact of fronts on chlorophyll-a. In addition, conclusions from this work seem to differ from those of Lie and Levine 2016, which showed a 40% increase in the winter. Is it only due to the different resolution in the SST and chlorophyll-a products?
AC: We will remove “smaller than expected”. What we intended to say here, is that the impact of the regional scale are smaller than the impact at the scale of the front. We propose the following rephrasing for the last sentence of our abstract, to explain how we reconcile these two apparently conflicting views:
“Nevertheless, our results suggest that the spectacular impact of fronts at the local level is more limited when considered at the regional level”.
We will also be more cautious in our revision, and this answers some of the concerns of the first reviewer as well, on the fact that the “small impact” may in fact 1) be underestimated and 2) have stronger impacts on the ecosystem as a whole.
Also, indeed, the impact of 40% find by Liu and Levine is larger than our findings in the permanent subtropical biome. And it is a very good question to ask why. It could be indeed be due to the resolution of the products, to the size of the window to compute the HI index (10 km for Liu and Levine versus 30 km here), or to the difference between the two regions. Thus we would need to compare the two products in the two regions (Atlantic versus Pacific) for a range of window sizes to understand what drives the difference. We did some tests with their product, but on a few images, and that did not seem to play a role. We also compared the results for different window sizes (Supplementary Fig): in the PSB in winter, there is indeed some degree of sensitivity, larger than elsewhere, mainly because the Chl-a values are very small, and the median value is very sensitive there to the area covered by fronts. But a more in depth study comparison, based on a statistically significant number of cases would be required. This would go beyond our objectives here but we intend to test it in the future, as we intend to extend our approach to larger zones of the global ocean. We will make sure that this comparison is properly highlighted in the paper.RC: Specific comments:
Title: “earlier and stronger” than what?AC: The title says “earlier and stronger phytoplankton blooms over fronts” so implicitly, “than outside of fronts”. We felt it would be a bit heavy to add this precision to the title, and we have hence explained it more clearly the abstract.
RC: Lines 29-31: “largely explained by consistent physical forcing and environmental conditions” - not clear what the authors are referring to. Aren’t the environmental conditions a consequence of the physical forcing?
AC: Not necessarily, if one thinks in terms of the nutrient distributions for instance, they are related to the physical forcings but also to their sources and sinks. We will make it clear that by environmental conditions, we are thinking of nutrients.
RC: Line 31 : “wind-driven circulation”: the authors could be more explicit in saying that negative wind stress curl induces downwelling in the subtropical gyres while the opposite occurs on subplolar gyres.
AC: Indeed, we will.
RC: LIne 149: the authors seem to differentiate the “atmospherically and topographically”-controlled fronts from “submesoscale” fronts. This distinction is not so straightforward as submesoscale
AC: This differentiation comes from the fact that there are indeed various ways to define submesoscale, either in terms of a specific scale range or in terms of Rossby number. Here we think of submesoscale as flows characterized by Rossby number of order one, and these flows encompass what we call “ephemeral submesoscale fronts”, and “persistent fronts” (such as the Gulf Stream here), which also have Rossby number of order one but have larger scale. We will make this clearer in the introduction.
RC: Figure 2c: I suggest a qualitative or discreet colorbar for the different categories of HI. In other words, what matters to the reader are the intervals, not the continuous values.
AC: In Fig 2c, we have overlaid the contours for HI=5 and HI=10 to the color bar. The same contours are overlaid in Fig 2a and 2b. This way, we both show the intervals, and we also show the continuous values. Indeed, part of our discussion is to say that there is a continuity in the relationship between HI and Chl, as shown in Figure 3.
RC: Line 197: typo. “Onsets propagates”. ~Remove “s”.
AC: Thanks
RC: Line 221: It is also consistent with baroclinic instability, which is not a submesoscale process.
AC: We will remove this sentence, and add that “Submesoscale flows are typically created by mesoscale eddy stirring or emerge spontaneously from baroclinic instability in surface or bottom boundary layers”
RC: Line 237: typo : add “s” to “front”.
AC: Thanks
RC: Lines 243-244: not clear. Are you saying that strong fronts are counted as weak fronts during summer? Isn’t the methodology robust enough to have a consistent classification?
AC: Indeed this was not clear, all the more that there is in fact also a weak reduction in the number of fronts in the subpolar biome in summer. We propose to remove this comment which was unclear and not supported by further analysis.
RC: Line 249: I assume this does not affect the interpretation of regions affected by fronts right? Otherwise, it would required a more detailed analysis of points with high chlorophyll-a in regions of low HI.
AC: See our answer to the first reviewer on that specific point. As a more detailed analysis, we propose to add one figure in supplementary material which shows the distributions of Chl-a (in addition to their median value) within each region and for each season. This figure shows that the quantity of high Chl-a points in regions of low HI (i.e. in the background) is much less than in the regions of large HI (the fronts), and that the median value captures that.
RC: References
Johnson, K., Riser, S. & Karl, D. Nitrate supply from deep to near-surface waters of the North Pacific subtropical gyre. Nature 465, 1062–1065 (2010). https://doi.org/10.1038/nature09170
Pelegrí, J. L., and Csanady, G. T. (1991), Nutrient transport and mixing in the Gulf Stream, J. Geophys. Res., 96( C2), 2577– 2583, doi:10.1029/90JC02535.
Siegel, D. A., Peterson, P., McGillicuddy, D. J., Maritorena, S., and Nelson, N. B. (2011), Bio-optical footprints created by mesoscale eddies in the Sargasso Sea, Geophys. Res. Lett., 38, L13608, doi:10.1029/2011GL047660.
Citation: https://doi.org/10.5194/egusphere-2022-1489-RC2AC: Thanks again for your time in providing these very useful and detailed comments.
Joined in AC1: new supplementary figure showing the distributions of Chl-a by season, biome, and front type.
Citation: https://doi.org/10.5194/egusphere-2022-1489-AC2
-
AC2: 'Reply on RC2', Clément Haëck, 16 Mar 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-1489', Anonymous Referee #1, 18 Feb 2023
The manuscript by Haeck et al., presents an analysis of the amplification of surface chlorophyll along weak and strong fronts (detected from satellite SST) in the Western North Atlantic region. The results show significant amplification of chlorophyll at the location of SST fronts, which, when averaged over the domain where fronts are found, increases the total chlorophyll by up to 5% on average over broad biomes. Furthermore, spring blooms are observed to occur earlier — by up to two weeks — along fronts in the subpolar biome.
Heterogeneity in the marine environment is a topic of great interest from physical to biological oceanography. Fronts and eddies usually draw a lot of interest, but the actual impacts on marine ecosystems remain poorly quantified. The manuscript by Haeck et al. thus sheds light on a topic of current relevance.
I found the manuscript very well written and the analysis thorough and sound. This is clearly the result of a lot of work, and the results are not only useful (because they address long standing questions), but also very stimulating. I can imagine a global extension of the approach, which would further advance our understanding of physical-biogeochemical coupling in the ocean, and inform ecological thinking. I recommend publication of the manuscript with only minor suggestions.
As a general criticism, I somewhat disagree with the negative spin on the results by the Authors. While 5 % amplification over large scales may not seem that large, it may still have important implications for ecology; after all, chlorophyll amplification can be significant along fronts (E.g. Fig. 8f,h); early onset of blooms along front may also be important for organisms phenology. Furthermore, the study quantifies only one aspect of the effects of front on phytoplankton — i.e., is the direct chlorophyll amplification; as discussed in the text, also because of the ephemeral nature of fronts, nutrients upwelled along fronts may be dispersed more broadly and contribute to an average amplification of chlorophyll that may not be directly co-located with fronts. Of course, this effect is hard, if not impossible to detect purely from remote-sensing. But in few instances in the abstract and conclusions the Authors could be more specific in stating that they quantify the specific effect of biomass amplification at fronts, not other possibly more widespread effects. Additionally, SST is an imperfect proxy for the expected phytoplankton response along fronts, which may lag more than a week relative to the excess of recently upwelled nutrients that may co-occur with cooler SSTs. While these caveats are discussed in the manuscript, they could be somewhat better highlighted.
Specific comments:
Line 11, “the global enhancement of Chlorophyll-a due to fronts” maybe clarify as “the global enhancement of Chlorophyll-a along fronts”
Line 16, “misleading”: I would use more nuance here, and avoid this term.
Line 17, “budget”: maybe “biomass amplification” or “chlorophyll amplification” would be more appropriate, since technically a budget (which implies some balance of different sources and sinks) has not been evaluated. See also “budget” in line 55 — at least clarify what the term means in this context.
Line 71: this sentence seems a bit obvious; it could be removed.
Line 122, “All pixels where water depth is less than 1500m are masked to exclude the continental shelf.”. This needs at least a sentence to justify the removal of shelf waters.
“Front detection” section. This is well described and builds nicely on previous work. I would only advise to clarify better why some choices were made and how different choices may affect the results; showing or stating that results are not very sensitive to specific thresholds or changes from previous methods would suffice.
Lines 204-206: I was somewhat confused by how the lag “L” was defined; maybe add a sentence to clarify its definition.
Lines 246-249: This seems an important point; lack of co-location (in space or time) of SST fronts and Chl maxima may be the consequence of interesting dynamical reasons, e.g. related to the timescales of phytoplankton response vs. the physical lifetime of a front. This could be discussed.
Lines 262-264: this entire paragraph doesn’t seem necessary; it could be removed to streamline the paper. In general the paper is on the long side, so some streamlining may help.Lines 306-307: Very interesting to observe the negative effect of fronts in summer, likely related to subduction as discussed later. This is a nice result.
Lines 315-319, “To quantify …”: this part may belong to Methods; it could also be clarified.
Lines 351-353: the other effect not quantified here is the broader supply and re-distribution of nutrients that may be caused by fronts; i.e., additional upwelled nutrients may not remain confined to the front that upwelled them, and be able to fertilize phytoplankton more broadly.
Line 385, “what is generally thought”: maybe add a reference, or it risks to be a straw man argument.
Also, related to the greater impact of fronts in bloom regions relative to oligotrophic regimes, the study by Yamamoto et al. (2018, Journal of Geophysical Research) provided (based on a mesoscale-eddy permitting model) evidence that the majority of nutrient supply to the euphotic zone of the oligotrophic gyres occurs by lateral eddy fluxes rather than vertical eddy fluxes. This certainly relates to the argument of deep nutriclines in these regions.
Lines 420-424: a clear sense of why blooms occur early along fronts is a bit hidden in this explanation, maybe clarify a bit.
Line 429: I think the results of early blooms on fronts may be important for phenology of zooplankton, and potential their predators, which could be more explicitly discussed.
Line 444, “due to fronts”: again here, it may be better to add more nuance, since only a direct effect of fronts on co-located chlorophyll was quantified here; indirect effects (e.g. nutrient redistribution) may be also important.
Technical comments:Line 14: “the the” —> “the”
Line 64, “spacial” —> “spatial”
Line 70, “built” —> “build”
Line 82, “contrasted” —> “contrasting”
Line 115: “tends to provide an underestimate of” —> “tends to underestimate”
Line 139: “Gulf stream” —> “Gulf Stream”
Line 276: “and this throughout” —> “and this holds throughout”
Line 326, “one months” —> “one month”
Caption of fig. 10, and other parts of the text: the use of “rest.” May not be completely clear, maybe use the full word (respectively?)
Line 429, “budget” —> “budgets”Citation: https://doi.org/10.5194/egusphere-2022-1489-RC1 -
AC1: 'Reply on RC1', Clément Haëck, 16 Mar 2023
RC1: The manuscript by Haeck et al., presents an analysis of the amplification of surface chlorophyll along weak and strong fronts (detected from satellite SST) in the Western North Atlantic region. The results show significant amplification of chlorophyll at the location of SST fronts, which, when averaged over the domain where fronts are found, increases the total chlorophyll by up to 5% on average over broad biomes. Furthermore, spring blooms are observed to occur earlier — by up to two weeks — along fronts in the subpolar biome.
Heterogeneity in the marine environment is a topic of great interest from physical to biological oceanography. Fronts and eddies usually draw a lot of interest, but the actual impacts on marine ecosystems remain poorly quantified. The manuscript by Haeck et al. thus sheds light on a topic of current relevance.
I found the manuscript very well written and the analysis thorough and sound. This is clearly the result of a lot of work, and the results are not only useful (because they address long standing questions), but also very stimulating. I can imagine a global extension of the approach, which would further advance our understanding of physical-biogeochemical coupling in the ocean, and inform ecological thinking. I recommend publication of the manuscript with only minor suggestions.AC1: We thank the reviewer for their positive appreciation of our work
RC: As a general criticism, I somewhat disagree with the negative spin on the results by the Authors. While 5 % amplification over large scales may not seem that large, it may still have important implications for ecology; after all, chlorophyll amplification can be significant along fronts (E.g. Fig. 8f,h); early onset of blooms along front may also be important for organisms phenology. Furthermore, the study quantifies only one aspect of the effects of front on phytoplankton — i.e., is the direct chlorophyll amplification; as discussed in the text, also because of the ephemeral nature of fronts, nutrients upwelled along fronts may be dispersed more broadly and contribute to an average amplification of chlorophyll that may not be directly co-located with fronts. Of course, this effect is hard, if not impossible to detect purely from remote-sensing. But in few instances in the abstract and conclusions the Authors could be more specific in stating that they quantify the specific effect of biomass amplification at fronts, not other possibly more widespread effects. Additionally, SST is an imperfect proxy for the expected phytoplankton response along fronts, which may lag more than a week relative to the excess of recently upwelled nutrients that may co-occur with cooler SSTs. While these caveats are discussed in the manuscript, they could be somewhat better highlighted.
AC: We thank the reviewer for their constructive criticism. We agree on all points raised and will discuss them in the revised manuscript:
- Regarding the negative spin on the results: in the context of previous observational and modeling studies that have shown very large amplifications at ocean fronts, our initial intention was to put them into a broader context, and to show that these large local amplifications contribute modestly to regional scale amplification. We agree we might have fallen too much on the negative side, and we propose a few modifications in the text to give some more insights on the widespread effects. For instance, we propose to change the last sentence of the abstract into something more nuanced (see answer to specific comment line 16).
- that 5% amplification of surface Chl might lead to greater amplification in terms of ecological implications (Stock et al. 2014, Lotze et al. 2019). We will mention these two papers to illustrate that point in our discussion (section 4.4):
- Stock CA, Dunne JP, John JG. 2014. Drivers of trophic amplification of ocean productivity trends in a changing climate. Biogeosciences. 11(24):7125–35
- Lotze HK, Tittensor DP, Bryndum-Buchholz A, Eddy TD, Cheung WWL, et al. 2019. Global ensemble projections reveal trophic amplification of ocean biomass declines with climate change. Proc. Nat. Ac. Sc. 1:201900194
We will also add in the plain language summary: “Nevertheless these relatively modest effects in term of surface chlorophyll might have more significant impacts for the rest of the ecosystem.” - that nutrient enrichments due to fronts might lead to phytoplankton biomass enhancement outside of fronts.
We agree and this result is actually seen for example in our Figure 2, as the Chl-a filament expands outside the area of elevated HI in the north west part of the domain. We will use this example to discuss that our estimate is a lower limit because the enrichments due to fronts may be seen outside of fronts. We will expand the discussion on how these enrichments outside of fronts are integrated in our analysis and how much they represent by showing an additional supplementary figure with the Chl-a distributions. We will expand the caveat section of the discussion. - that the mismatch between Chla and HI (seen in figure 2) may be due to the time evolution of the flow, to diffusion, and to the time lag between nutrient supplies and Chl increase.
These aspects will be discussed when Figure 2 is described, and we will come back to them in the final discussion.
RC: Specific comments:
Line 11, “the global enhancement of Chlorophyll-a due to fronts” maybe clarify as “the global enhancement of Chlorophyll-a along fronts”AC: We will rephrase the sentence as it is not exactly the enhancement along fronts, but really, the effect of fronts when integrated over a large region:
“… the local enhancement over fronts, the global enhancement that it implies at the regional scale, and...”RC: Line 16, “misleading”: I would use more nuance here, and avoid this term.
AC: We agree and propose to rephrase as follows: “Nevertheless, our results suggest that the spectacular impact of fronts at the local level is more limited when considered at the regional level”.
RC: Line 17, “budget”: maybe “biomass amplification” or “chlorophyll amplification” would be more appropriate, since technically a budget (which implies some balance of different sources and sinks) has not been evaluated. See also “budget” in line 55 — at least clarify what the term means in this context.
AC: We agree and will remove the term budget everywhere in the paper, and replace it with amplification.
RC: Line 71: this sentence seems a bit obvious; it could be removed.
AC: We agree and will shorten it
RC: Line 122, “All pixels where water depth is less than 1500m are masked to exclude the continental shelf.”. This needs at least a sentence to justify the removal of shelf waters.
AC: We agree and propose to add the following sentence: “In addition to these three biomes, other biomes are found on the continental shelf but were excluded from this study that focuses on open ocean conditions”.
RC: “Front detection” section. This is well described and builds nicely on previous work. I would only advise to clarify better why some choices were made and how different choices may affect the results; showing or stating that results are not very sensitive to specific thresholds or changes from previous methods would suffice.
AC: Indeed, on line 167 we say that we performed sensitivity tests on the parameter choices with no further comments. And it is only from line 280 that we say that our results are very weakly sensitive to the choice of parameters (with a reference to supplementary Fig. A1 which actually shows it). We will bring these two paragraphs together.
RC: Lines 204-206: I was somewhat confused by how the lag “L” was defined; maybe add a sentence to clarify its definition.
AC: We agree and we propose to change this entire paragraph in the method section, by incorporating also your other comment regarding lines 315-319 that are moved to the method section:
“Finally, the time series of the Chl-a median in the subpolar biome is characterized by a spring bloom, of which we measured the timing (onset date) both in the fronts and in the background, for each year. Because the spring bloom onsets propagates from south to north, these timings were inferred over latitudinal band of width 5°. Our estimates are based on averaged statistics over Eulerian time series, assuming that the bloom evolves coherently in the background (resp. over fronts) within each latitudinal bands in the subpolar biome. Thus we are assuming that front and non front pixels can be pooled apart to follow the bloom evolution over the two contrasting environments. This assumption is suggested by high-resolution models of the bloom (e.g. Lévy et al., 2005; Karleskind et al., 2011), but it is imperfect as the bloom evolves along Lagrangian trajectories. Nevertheless, we found very consistent results with this approach. We define the lag in bloom onset L as the bloom onset day in the fronts minus the bloom onset day in the background.
To extract the onset date, we filtered the Chl-a median time series with a low-pass Butterworth filter of order 2 and cutting frequency 1/20days−1. The filtered time-series displayed strong variations in their phenology from year to year, but a bloom was always discernible. We considered data from February to July, which allowed us to isolate the spring bloom and exclude the autumn bloom. First, we detected the maximum value of Chl-a in this time window, then defined the bloom onset as the time of maximum Chl-a derivative prior to the time of maximum Chl-a.
We defined the uncertainty in bloom onset date as the standard deviation of all days for which the Chl-a derivative is above 90% of its maximum value. We defined the uncertainty in lag L as the square root of the sum of squared uncertainties in fronts and background. Finally, we defined the mean lag (over the 20 years of data and for each latitudinal band) as the weighted averaged lag, with the weights equal to the inverse of the uncertainties; and we defined the mean lag uncertainty as the weighted standard deviation of all lag values around the mean lag, with the weights equal to the inverse of the uncertainties.“RC: Lines 246-249: This seems an important point; lack of co-location (in space or time) of SST fronts and Chl maxima may be the consequence of interesting dynamical reasons, e.g. related to the timescales of phytoplankton response vs. the physical lifetime of a front. This could be discussed.
AC: Yes we agree, this relates to one of your earlier comments, and we will discuss it here. We propose to add the following paragraph, and to provide an additional supplementary figure to show how the distributions of Chl differ between fronts and the background (attached here):
“the mismatch between the exact location of fronts at a given time and the exact location of the Chl-a response at the same time may be due to different factors, such as the very dynamic nature of fronts which can lead to the chaotic advection of phytoplankton (and/or nutrients) outside of fronts, the time-scale needed for phytoplankton to respond to nutrient supplies at fronts, or trophic interactions at the front (as shown, for instance, in the modelling study of Mangolte et al. 2022, their Figure 1d). This implies that there might be remnant of large Chl-a values due to frontal activity outside of fronts. Nevertheless, our results show that on average, the distribution of Chl-a at a given time differs over fronts and outside of fronts (supplementary figure), and also depends on the strength of fronts.“
RC: Lines 262-264: this entire paragraph doesn’t seem necessary; it could be removed to streamline the paper. In general the paper is on the long side, so some streamlining may help.
AC: We agree – the paragraph will be removed.
RC: Lines 306-307: Very interesting to observe the negative effect of fronts in summer, likely related to subduction as discussed later. This is a nice result.
AC: Thanks. We feel that this result is maybe not highlighted enough and will make it appear more clearly in the abstract and in the discussion of the revised version.
RC: Lines 315-319, “To quantify …”: this part may belong to Methods; it could also be clarified.
AC: Indeed, lines 315-321 will be moved to method and clarified. This relates to your comment regarding the computation of the lag L line 204-206. We will rewrite entirely the paragraph that goes from line 196-206 that explains the computation of the bloom timing (see above)
RC: Lines 351-353: the other effect not quantified here is the broader supply and re-distribution of nutrients that may be caused by fronts; i.e., additional upwelled nutrients may not remain confined to the front that upwelled them, and be able to fertilize phytoplankton more broadly.
AC: Indeed, in addition, we will complete this paragraph and add one reference:
“Moreover, our assessment of the effect of fronts on phytoplankton, based on surface Chl-a, is probably a lower estimate. First, it only measures the excess Chl-a at the surface but the biological signal due to the upwelling of nutrients at fronts often does not reach the surface and is more intense at sub-surface (Mourino et al., 2004; Ruiz et al., 2019). Secondly, the excess Chl-a due to fronts is not always located exactly over fronts because of chaotic advection during the time needed for phytoplankton to respond to nutrient enrichments at fronts (Abraham, 1998).“- Abraham E. 1998. The generation of plankton patchiness by turbulent stirring. Nature. 391(6667):577–80
RC: Line 385, “what is generally thought”: maybe add a reference, or it risks to be a straw man argument.
AC: Indeed, we will remove this statement and add:
“Finally, an important outcome of this study is that the amplification in phytoplankton biomass at fronts can be stronger in bloom regimes than in oligotrophic regimes. Indeed, in eutrophic regions, where nutrients are not limiting productivity, fronts have generally been shown to subduct excess nutrients (Oschlies, 2002; Gruber at al., 2011) and excess biomass (Lathuiliere et al., 2010) rather than to increase biomass as in oligotrophic regions. This effect of a decreased biomass, suggestive of subduction, is observed here in the northern parts of the subpolar biome between May and September (Fig. 8b-d). But we also find that in the subpolar biome, the Chl-a excess over fronts can reach 150% during the bloom (Fig. 8f), and 50% during summer (Fig. 8h), while in the permanent subtropical biome it never exceeds 10% (Fig. 6b).”
- Oschlies A. 2002. Can eddies make ocean deserts bloom. Glob. Biogeochem. Cyc. 16(4):1106
- Gruber N, Lachkar Z, Frenzel H, Marchesiello P, Münnich M, et al. 2011. Eddy-induced reduction of biological production in eastern boundary upwelling systems. Nature Geoscience. 4(11):787–92
- Lathuiliere C, Lévy M, Echevin V. 2010. Impact of eddy-driven vertical fluxes on phytoplankton abundance in the euphotic layer. J. Plankton Res. 33:827–31RC: Also, related to the greater impact of fronts in bloom regions relative to oligotrophic regimes, the study by Yamamoto et al. (2018, Journal of Geophysical Research) provided (based on a mesoscale-eddy permitting model) evidence that the majority of nutrient supply to the euphotic zone of the oligotrophic gyres occurs by lateral eddy fluxes rather than vertical eddy fluxes. This certainly relates to the argument of deep nutriclines in these regions.
AC: Yes, we agree and this is a very important point. We thank you and the second reviewer for raising it. We will change our introduction, discussion and conclusion to highlight that increased Chl-a over fronts is not necessarily the result of vertical advection but may also be related to lateral transport and particularly to the nutrient stream (as our results actually show evidence for it, and we saw it thanks to your feedback).
• We will replace “vertical nutrient supply” by “nutrient supply” in the first sentence of the abstract
• In the 2nd line of the abstract, we will add “linked to local nutrient upwelling and/or to horizontal transport”
• Abstract, line 15, we will add “the local enhancement of Chl-a was particularly strong in the latitude band of the Gulf Stream, likely reflecting the lateral supply of nutrients through the nutrient stream”
• We will revise the second paragraph of the introduction. line 42 we will add: “In addition, western boundary currents like the Gulf Stream (Peligry et al., 1996) and the Kuroshio (Long et al., 2022) are also the locus of intense nutrient transport over long distances, or nutrient streams. We will also remove lines 43 to 46
• Introduction line 78, we will also refer to the nutrient stream as a potential process
• Method section line 132: we will add that the Gulf Steam conveys nutrients and that the nutrient stream falls within the seasonal subtropical biome
• Discussion section 4.2 will be strongly modified to discuss the idea that the stronger local increase over strong fronts in the intermediate region likely reflects the lateral transport of nutrients by the nutrient stream. We will also discuss lateral eddy fluxes of nutrients which are more important at the gyre boundaries (Yamamoto et al. 2018, Spingys et al., 2021). We will make it clear that the local increase of Chl-a over fronts is not only the consequence of local upwelling of nutrients but also of lateral eddy fluxes of nutrients.- Long Y, Guo X, Zhu X-H, Li Z. 2022. Nutrient streams in the North Pacific. Prog. Oceanogr. 202:102756
- Pelegrí JL, Csanady GT, Martins A. 1996. The North Atlantic nutrient stream. Journal of oceanography. 52(3):275–99
- Yamamoto A, Palter JB, Dufour CO, Griffies SM, Bianchi D, et al. 2018. Roles of the Ocean Mesoscale in the Horizontal Supply of Mass, Heat, Carbon, and Nutrients to the Northern Hemisphere Subtropical Gyres. J. Geophys. Res. Ocean. 123(10):7016–36
- Spingys CP, Williams RG, Tuerena RE, Garabato AN, Vic C, et al. 2021. Observations of Nutrient Supply by Mesoscale Eddy Stirring and Small-Scale Turbulence in the Oligotrophic North Atlantic. Global Biogeochemical Cycles. 35(12):e2021GB007200RC: Lines 420-424: a clear sense of why blooms occur early along fronts is a bit hidden in this explanation, maybe clarify a bit.
AC: We propose this new formulation:
“Several field and modeling studies have shown that frontal dynamics, by tilting existing horizontal density gradients, increase the vertical stratification of surface mixed layers (Taylor and Ferrari, 2011), which can lead to the stratification of the mixed layer prior to seasonal stratification. Given that the surface spring bloom is triggered by increased stratification, this effect can cause earlier local phytoplankton blooms over fronts compared to surrounding areas”RC: Line 429: I think the results of early blooms on fronts may be important for phenology of zooplankton, and potential their predators, which could be more explicitly discussed.
AC: Indeed, we propose to add this comment:
“This effect alone is unlikely to affect primary productivity, but may impact phytoplankton competition at the onset of the bloom season, affect their predators, and the phenology of zooplankton.“RC: Line 444, “due to fronts”: again here, it may be better to add more nuance, since only a direct effect of fronts on co-located chlorophyll was quantified here; indirect effects (e.g. nutrient redistribution) may be also important.
AC: Yes, we propose to use “associated with fronts” instead of “due to fronts”
RC: Technical comments:
Line 14: “the the” —> “the” Line 64, “spacial” —> “spatial” Line 70, “built” —> “build” Line 82, “contrasted” —> “contrasting” Line 115: “tends to provide an underestimate of” —> “tends to underestimate” Line 139: “Gulf stream” —> “Gulf Stream” Line 276: “and this throughout” —> “and this holds throughout” Line 326, “one months” —> “one month” Caption of fig. 10, and other parts of the text: the use of “rest.” May not be completely clear, maybe use the full word (respectively?) Line 429, “budget” —> “budgets”
Citation: https://doi.org/10.5194/egusphere-2022-1489-RC1AC: Thanks again for your time in providing these very useful and detailed comments
Attached: new supplementary figure showing the distributions of Chl-a by season, biome, and front type.
-
AC1: 'Reply on RC1', Clément Haëck, 16 Mar 2023
-
RC2: 'Comment on egusphere-2022-1489', Anonymous Referee #2, 01 Mar 2023
This is the review of “Satellite data reveal earlier and stronger phytoplankton blooms over fronts in the Gulf Stream region”, by Haëck et al.
The authors build on the methods from Levine et al. 2016 to investigate the impact of fronts—as diagnosed by a heterogeneity index (HI) based on sea surface temperature—on surface chlorophyll-a—as estimated by remote sensing—in different biomes in the Western North Atlantic. Namely, a permanently subtropical, a seasonally subtropical and a subpolar biome. Based on the HI, they estimate the occurrence of weak (5<HI<10) and strong (HI>10) fronts in each biome. Regions with HI less than 5 were considered to be background. The authors noted that high HI values were associated with quasi-permanent fronts while the weak fronts were more ephemeral. Data resolution (4 km) allows the authors to estimate HI in a square of around 24 km (49 grid points).
The authors find that fronts increase surface phytoplankton by about 5% over the region, which they claim is a “much weaker than expected” impact ( although the authors are not explicit about the sources of the high expectations). Secondly, the authors claim that the nutrient supply at fronts more than doubled chlorophyll concentrations during the spring bloom Thirdly, the authors claim that the spring bloom over fronts occur one to two weeks earlier than in background regions.
I find the study interesting. It is well written, well organized and the methodology is sound. I do, however, have a few comments which I believe should be clarified before publication.
Although it is not explicitly stated, the hypothesis the authors are testing is : SST fronts impact surface ocean chlorophyll via submesoscale vertical advection due to unbalanced motions. By stating that the the impact of SST fronts is “much weaker than expected”, they seem to reject that hypothesis.
The premise is that the impact of SST fronts on surface chlorophyll occurs via submesoscale motions (e.g. ageostrophic secondary circulations due to frontogenesis.). By supplying nutrients into a nutrient-depleted surface layer via submesoscale vertical advection, biomass would increase. The problem with this premise is that it is only valid in the permanently oligotrophic region. In the seasonally subtropical and the subpolar biomes, where fronts are more frequent and stronger, lateral advection plays a much more important role due to large-scale climatological horizontal gradients in surface properties. In the Gulf Stream region, not only is the current know to be a “nutrient stream” (Pelegri and Csanady, 1991), with robust horizontal transport of nutrients, but also the front is maintained by the thermal wind shear. There certainly is imbalance and ageostrophic motion, but I am not sure if that is truly discernible from the submesoscale instabilities the authors seem to be mostly concerned about.
In addition, it is unclear whether or not the impact of mesoscale eddies on chlorophyll is removed from the analysis by considering a 24 km square. It is known that the curvature induced by mesoscale eddies impact chlorophyll-a distribution (Siegel et al. 2011). While the analysis was performed on a larger scale in the case of Siegel et al. 2011, it is unclear if the mesoscale signal is removed from the current analysis. Could, for example, some of the detected frontal regions be associated with the edges of mesoscale eddies?
With regards to the permanently stratified, there’s observational evidence (Johnson et al. 2010) that episodic injections, due to meso- or submesoscale processes induce vertical advection of nutrients but that these nutrients get consumed before reaching the surface, so the signal may not be captured by SST in these regions. In addition, there may be photo-inhibition which prevents phytoplankton to be near the surface in oligotrophic regions, even though they may be impacted by frontal motions. The authors partially mention that in lines 351-353, but the studies they cite make use of the QG-omega equation to estimate the vertical velocities, which may be on the mesoscale end of the frontal structures.
Lastly, it is not clear how the authors reconcile the “considerable evidence that submesoscale motions influence nutrient and light environments” with the “smaller than expected” impact of fronts on chlorophyll-a. In addition, conclusions from this work seem to differ from those of Lie and Levine 2016, which showed a 40% increase in the winter. Is it only due to the different resolution in the SST and chlorophyll-a products?
Specific comments:
Title: “earlier and stronger” than what?
Lines 29-31: “largely explained by consistent physical forcing and environmental conditions” - not clear what the authors are referring to. Aren’t the environmental conditions a consequence of the physical forcing?
Line 31 : “wind-driven circulation” : the authors could be more explicit in saying that negative wind stress curl induces downwelling in the subtropical gyres while the opposite occurs on subplolar gyres.
LIne 149: the authors seem to differentiate the “atmospherically and topographically”-controlled fronts from “submesoscale” fronts. This distinction is not so straightforward as submesoscale
Figure 2c: I suggest a qualitative or discreet colorbar for the different categories of HI. In other words, what matters to the reader are the intervals, not the continuous values.
Line 197: typo. “Onsets propagates”. ~Remove “s”.
LIne 221: It is also consistent with baroclinic instability, which is not a submesoscale process.
Line 237: typo : add “s” to “front”.
Lines 243-244: not clear. Are you saying that strong fronts are counted as weak fronts during summer? Isn’t the methodology robust enough to have a consistent classification?
Line 249: I assume this does not affect the interpretation of regions affected by fronts right? Otherwise, it would required a more detailed analysis of points with high chlorophyll-a in regions of low HI.
References
Johnson, K., Riser, S. & Karl, D. Nitrate supply from deep to near-surface waters of the North Pacific subtropical gyre. Nature 465, 1062–1065 (2010). https://doi.org/10.1038/nature09170
Pelegrí, J. L., and Csanady, G. T. (1991), Nutrient transport and mixing in the Gulf Stream, J. Geophys. Res., 96( C2), 2577– 2583, doi:10.1029/90JC02535.
Siegel, D. A., Peterson, P., McGillicuddy, D. J., Maritorena, S., and Nelson, N. B. (2011), Bio-optical footprints created by mesoscale eddies in the Sargasso Sea, Geophys. Res. Lett., 38, L13608, doi:10.1029/2011GL047660.
Citation: https://doi.org/10.5194/egusphere-2022-1489-RC2 -
AC2: 'Reply on RC2', Clément Haëck, 16 Mar 2023
RC2: This is the review of “Satellite data reveal earlier and stronger phytoplankton blooms over fronts in the Gulf Stream region”, by Haëck et al.
The authors build on the methods from Levine et al. 2016 to investigate the impact of fronts—as diagnosed by a heterogeneity index (HI) based on sea surface temperature—on surface chlorophyll-a—as estimated by remote sensing—in different biomes in the Western North Atlantic. Namely, a permanently subtropical, a seasonally subtropical and a subpolar biome. Based on the HI, they estimate the occurrence of weak (5<HI<10) and strong (HI>10) fronts in each biome. Regions with HI less than 5 were considered to be background. The authors noted that high HI values were associated with quasi-permanent fronts while the weak fronts were more ephemeral. Data resolution (4 km) allows the authors to estimate HI in a square of around 24 km (49 grid points).
The authors find that fronts increase surface phytoplankton by about 5% over the region, which they claim is a “much weaker than expected” impact (although the authors are not explicit about the sources of the high expectations). Secondly, the authors claim that the nutrient supply at fronts more than doubled chlorophyll concentrations during the spring bloom. Thirdly, the authors claim that the spring bloom over fronts occur one to two weeks earlier than in background regions.
I find the study interesting. It is well written, well organized and the methodology is sound. I do, however, have a few comments which I believe should be clarified before publication.
Although it is not explicitly stated, the hypothesis the authors are testing is: SST fronts impact surface ocean chlorophyll via submesoscale vertical advection due to unbalanced motions. By stating that the the impact of SST fronts is “much weaker than expected”, they seem to reject that hypothesis.
The premise is that the impact of SST fronts on surface chlorophyll occurs via submesoscale motions (e.g. ageostrophic secondary circulations due to frontogenesis). By supplying nutrients into a nutrient-depleted surface layer via submesoscale vertical advection, biomass would increase. The problem with this premise is that it is only valid in the permanently oligotrophic region. In the seasonally subtropical and the subpolar biomes, where fronts are more frequent and stronger, lateral advection plays a much more important role due to large-scale climatological horizontal gradients in surface properties. In the Gulf Stream region, not only is the current know to be a “nutrient stream” (Pelegri and Csanady, 1991), with robust horizontal transport of nutrients, but also the front is maintained by the thermal wind shear. There certainly is imbalance and ageostrophic motion, but I am not sure if that is truly discernible from the submesoscale instabilities the authors seem to be mostly concerned about.AC2: We are very grateful to the reviewer for their positive assessment of our work, and especially for their sound criticism that we should have examined and discussed the role of lateral advection more thoroughly. This was indeed missing and is an important part of the story. We agree that particularly in the two biomes that are more affected by the Gulf Stream, it is very likely that the signal that we detect is due to the lateral supply of nutrients by the Gulf Stream. This view can explain why the signal is maximum and so much stronger than elsewhere at around 40°N in our figure 9a. We propose to add this in our discussion and highlight it in our results and abstract (see also our answer AC1 to referee 1 on this same point).
RC: In addition, it is unclear whether or not the impact of mesoscale eddies on chlorophyll is removed from the analysis by considering a 24 km square. It is known that the curvature induced by mesoscale eddies impact chlorophyll-a distribution (Siegel et al. 2011). While the analysis was performed on a larger scale in the case of Siegel et al. 2011, it is unclear if the mesoscale signal is removed from the current analysis. Could, for example, some of the detected frontal regions be associated with the edges of mesoscale eddies?
AC: We believe that considering a 24 km square removes a large part of the mesoscale signal from the analysis since mesoscale eddies in this region have diameters roughly between 50 and 200 km. But indeed, mesoscale and submesoscale dynamics are often strongly intertwined, and eddy edges can be considered as small frontal areas and can behave as such. We will add a paragraph to discuss what is captured by our analysis in the method section.
RC: With regards to the permanently stratified, there’s observational evidence (Johnson et al. 2010) that episodic injections, due to meso- or submesoscale processes induce vertical advection of nutrients but that these nutrients get consumed before reaching the surface, so the signal may not be captured by SST in these regions. In addition, there may be photo-inhibition which prevents phytoplankton to be near the surface in oligotrophic regions, even though they may be impacted by frontal motions. The authors partially mention that in lines 351-353, but the studies they cite make use of the QG-omega equation to estimate the vertical velocities, which may be on the mesoscale end of the frontal structures.
AC: Indeed, in Mourino et al., they lacked the full resolution to estimate the submesoscale W, but we wanted to mention this reference because of the vertical profiles of Chl (their Fig4) which showed Chl-a max at 100m, with finescale biomass enhancement at 75m. In Ruiz et al. (2019) their observations are complemented by a high-resolution model with strong submesoscale W where the Chl-a enhancement is also at subsurface (close to 40 m). In these two cases, the enhancement cannot be seen with surface data. We will extend the discussion in lines 351-353 to include these points. In addition, going back to an “old” submesoscale model that looked at nutrient injection in the permanently stratified oligotrophic gyre, the enhancement was much more pronounced at subsurface (Levy et al., 2001, Fig. 9f) but was nevertheless also seen at the surface.
- Lévy M, Klein P, Treguier A. 2001. Impact of sub-mesoscale physics on production and subduction of phytoplankton in an oligotrophic regime. J. Mar. Res. 59(4):535–65RC: Lastly, it is not clear how the authors reconcile the “considerable evidence that submesoscale motions influence nutrient and light environments” with the “smaller than expected” impact of fronts on chlorophyll-a. In addition, conclusions from this work seem to differ from those of Lie and Levine 2016, which showed a 40% increase in the winter. Is it only due to the different resolution in the SST and chlorophyll-a products?
AC: We will remove “smaller than expected”. What we intended to say here, is that the impact of the regional scale are smaller than the impact at the scale of the front. We propose the following rephrasing for the last sentence of our abstract, to explain how we reconcile these two apparently conflicting views:
“Nevertheless, our results suggest that the spectacular impact of fronts at the local level is more limited when considered at the regional level”.
We will also be more cautious in our revision, and this answers some of the concerns of the first reviewer as well, on the fact that the “small impact” may in fact 1) be underestimated and 2) have stronger impacts on the ecosystem as a whole.
Also, indeed, the impact of 40% find by Liu and Levine is larger than our findings in the permanent subtropical biome. And it is a very good question to ask why. It could be indeed be due to the resolution of the products, to the size of the window to compute the HI index (10 km for Liu and Levine versus 30 km here), or to the difference between the two regions. Thus we would need to compare the two products in the two regions (Atlantic versus Pacific) for a range of window sizes to understand what drives the difference. We did some tests with their product, but on a few images, and that did not seem to play a role. We also compared the results for different window sizes (Supplementary Fig): in the PSB in winter, there is indeed some degree of sensitivity, larger than elsewhere, mainly because the Chl-a values are very small, and the median value is very sensitive there to the area covered by fronts. But a more in depth study comparison, based on a statistically significant number of cases would be required. This would go beyond our objectives here but we intend to test it in the future, as we intend to extend our approach to larger zones of the global ocean. We will make sure that this comparison is properly highlighted in the paper.RC: Specific comments:
Title: “earlier and stronger” than what?AC: The title says “earlier and stronger phytoplankton blooms over fronts” so implicitly, “than outside of fronts”. We felt it would be a bit heavy to add this precision to the title, and we have hence explained it more clearly the abstract.
RC: Lines 29-31: “largely explained by consistent physical forcing and environmental conditions” - not clear what the authors are referring to. Aren’t the environmental conditions a consequence of the physical forcing?
AC: Not necessarily, if one thinks in terms of the nutrient distributions for instance, they are related to the physical forcings but also to their sources and sinks. We will make it clear that by environmental conditions, we are thinking of nutrients.
RC: Line 31 : “wind-driven circulation”: the authors could be more explicit in saying that negative wind stress curl induces downwelling in the subtropical gyres while the opposite occurs on subplolar gyres.
AC: Indeed, we will.
RC: LIne 149: the authors seem to differentiate the “atmospherically and topographically”-controlled fronts from “submesoscale” fronts. This distinction is not so straightforward as submesoscale
AC: This differentiation comes from the fact that there are indeed various ways to define submesoscale, either in terms of a specific scale range or in terms of Rossby number. Here we think of submesoscale as flows characterized by Rossby number of order one, and these flows encompass what we call “ephemeral submesoscale fronts”, and “persistent fronts” (such as the Gulf Stream here), which also have Rossby number of order one but have larger scale. We will make this clearer in the introduction.
RC: Figure 2c: I suggest a qualitative or discreet colorbar for the different categories of HI. In other words, what matters to the reader are the intervals, not the continuous values.
AC: In Fig 2c, we have overlaid the contours for HI=5 and HI=10 to the color bar. The same contours are overlaid in Fig 2a and 2b. This way, we both show the intervals, and we also show the continuous values. Indeed, part of our discussion is to say that there is a continuity in the relationship between HI and Chl, as shown in Figure 3.
RC: Line 197: typo. “Onsets propagates”. ~Remove “s”.
AC: Thanks
RC: Line 221: It is also consistent with baroclinic instability, which is not a submesoscale process.
AC: We will remove this sentence, and add that “Submesoscale flows are typically created by mesoscale eddy stirring or emerge spontaneously from baroclinic instability in surface or bottom boundary layers”
RC: Line 237: typo : add “s” to “front”.
AC: Thanks
RC: Lines 243-244: not clear. Are you saying that strong fronts are counted as weak fronts during summer? Isn’t the methodology robust enough to have a consistent classification?
AC: Indeed this was not clear, all the more that there is in fact also a weak reduction in the number of fronts in the subpolar biome in summer. We propose to remove this comment which was unclear and not supported by further analysis.
RC: Line 249: I assume this does not affect the interpretation of regions affected by fronts right? Otherwise, it would required a more detailed analysis of points with high chlorophyll-a in regions of low HI.
AC: See our answer to the first reviewer on that specific point. As a more detailed analysis, we propose to add one figure in supplementary material which shows the distributions of Chl-a (in addition to their median value) within each region and for each season. This figure shows that the quantity of high Chl-a points in regions of low HI (i.e. in the background) is much less than in the regions of large HI (the fronts), and that the median value captures that.
RC: References
Johnson, K., Riser, S. & Karl, D. Nitrate supply from deep to near-surface waters of the North Pacific subtropical gyre. Nature 465, 1062–1065 (2010). https://doi.org/10.1038/nature09170
Pelegrí, J. L., and Csanady, G. T. (1991), Nutrient transport and mixing in the Gulf Stream, J. Geophys. Res., 96( C2), 2577– 2583, doi:10.1029/90JC02535.
Siegel, D. A., Peterson, P., McGillicuddy, D. J., Maritorena, S., and Nelson, N. B. (2011), Bio-optical footprints created by mesoscale eddies in the Sargasso Sea, Geophys. Res. Lett., 38, L13608, doi:10.1029/2011GL047660.
Citation: https://doi.org/10.5194/egusphere-2022-1489-RC2AC: Thanks again for your time in providing these very useful and detailed comments.
Joined in AC1: new supplementary figure showing the distributions of Chl-a by season, biome, and front type.
Citation: https://doi.org/10.5194/egusphere-2022-1489-AC2
-
AC2: 'Reply on RC2', Clément Haëck, 16 Mar 2023
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 320 | 135 | 15 | 470 | 4 | 4 |
- HTML: 320
- PDF: 135
- XML: 15
- Total: 470
- BibTeX: 4
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Cited
1 citations as recorded by crossref.
Clément Haëck
Marina Lévy
Inès Mangolte
Laurent Bopp
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1271 KB) - Metadata XML