the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Jun 2026
A Lagrangian framework to simulate Sargassum transport and growth
Abstract. Blooms of the seaweed Sargassum have been reported in the Tropical Atlantic Ocean and Caribbean Sea since 2011. Large-scale inundation events of this seaweed in coastal regions have a negative impact on both the economy and the environment. To predict the timing, location and quantity of Sargassum strandings, model frameworks that link open ocean distributions to coastal regions are required. Here, we develop an open-source customizable Lagrangian simulation framework and growth model for Sargassum that combines transport by currents, wind and waves with a biological modelling framework that includes dynamic growth limitation depending on temperature, nitrate availability and salinity. The framework can use satellite detections of the Sargassum Watch System (SaWS) to initialise virtual Sargassum particles in the Tropical Atlantic Ocean. We demonstrate that the combination of physical transport and biological growth strongly affects Sargassum distribution, with substantial variability on spatio-temporal scales. We show that temperature is the strongest growth-limiting parameter, and in particular that elevated surface temperatures, together with low salinity from the Amazon River plume, play a crucial role in Sargassum decline.
- Preprint
(4452 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 04 Aug 2026)
- RC1: 'Comment on egusphere-2026-3097', Anonymous Referee #1, 01 Jul 2026 reply
-
RC2: 'Comment on egusphere-2026-3097', Julien Jouanno, 10 Jul 2026
reply
This paper presents an open-source, customizable Lagrangian framework for Sargassum transport and growth, implemented within the Parcels model. It couples physical transport by currents, windage and a depth-integrated Stokes drift with a biological growth model limited by temperature, nitrate and salinity, and it can be initialised from SaWS satellite detections. The development of an open, community-usable tool of this kind is a useful contribution, and the implementation of a depth-integrated Stokes drift is of interest. I have a number of comments that I believe would strengthen the manuscript.
Major comments
1. Positioning relative to previous studies. The authors do not clearly delineate what is new. Several components already exist in the literature: the temperature/nitrate/salinity growth limitation is largely taken from previous work, and the Lagrangian growth–decay approach is close in spirit to Podlejski et al. (2024), which reaches a similar conclusion on the role of temperature (by construction of the temperature dependence of the growth function I guess). It would help the reader to state more explicitly what the genuinely novel contributions are. I take these to be the implementation within Parcels and the depth-integrated Stokes drift.
2. Stokes drift and its interaction with windage. The depth-integrated Stokes drift is an important aspect of the transport that, to my knowledge, has been overlooked by the Sargassum community. I think this aspect is currently under-exploited in the manuscript, and developing it further could substantially increase the interest of the study. The windage coefficient (1%) has traditionally been used to account for unresolved Stokes drift, so resolving Stokes explicitly while retaining a 1% windage may double-count the wave contribution. Since the fields and machinery are already in place, a sensitivity analysis, running with and without the explicit Stokes term, and with windage across a 0–3% range, evaluated against the observed distribution, would justify the parameter choices and quantify what the explicit Stokes term adds beyond windage. This would also help constrain the raft depth extent, currently fixed at 1 m without a demonstrated sensitivity.
3.Model validation. The transport validation rests on a single visual comparison (Fig. 2) between the July 2024 SaWS distribution after 31 days and the 1 August detection. Over a one-month integration the two distributions are very similar, which limits what can be concluded from this figure. Since basin-wide monthly runs for all of 2024 are already produced (Section 3.2) and SaWS imagery is available year-round, a seasonally-resolved and quantitative skill assessment would be more convincing. Metrics such as those used in Berline et al. (2019) or Bonner et al. (2024) would allow an objective evaluation and would help refine the windage and Stokes-drift choices raised above. In short, my main recommendation would be to make the validation more quantitative.
4. Nitrogen limitation. Growth is limited by surface nitrate alone, which is at odds with conclusions reached by several authors indicating that both nitrogen and phosphorus limit growth (Lapointe et al. 1986, McGillicuddy et al., 2023, among others). There is also growing evidence for the role of diazotrophy, which can locally relieve nitrogen limitation when temperature, iron and phosphorus conditions are favourable. I therefore read the argument at L283–290 differently from the authors: Jung et al. (2025) attribute Sargassum growth to diazotrophy (N₂ fixation) driven by excess phosphorus from equatorial upwelling, which implies that the relevant nitrogen source is not only the ambient nitrate. Using this study to justify neglecting phosphorus and modelling only nitrate therefore appears to run counter to its conclusion. I would suggest revisiting this reasoning, and possibly discussing whether a phosphorus- or diazotrophy-informed nitrogen source would better represent Sargassum growth patterns.
5. Relevance of the Lagrangian approach. A discussion of the strengths and limitations of representing Sargassum with Lagrangian particles, relative to Eulerian alternatives, would be a useful addition. Points worth addressing include the treatment of raft dispersion and fragmentation, and the behaviour of the simple velocity-based stranding criterion at the coast.
Minor comments
L21: Wang et al. (2019) would be a more appropriate citation.
L36: Maybe cite Jouanno et al. 2023, in which a seasonal forecast was developed with skill assessment.
L40–42: It is not clear what "biophysical feedback" means here, please clarify. Regarding the potential influence of size, mass, and density, I agree in principle, but I am not sure any of the cited references have actually quantified or demonstrated such an impact.
L47–49: In Jouanno et al. (2025) it was shown that the GASB offers more favourable conditions for Sargassum growth, in terms of temperature, light, and nutrients, than the Subtropical Gyre.
L114: I find this formulation somewhat unusual. What is the rationale for the exponent of 2? I understand that Hanisak and Samuel use this form because it is the most appropriate way to approximate a growth rate from two measurements distant in time, but for numerical treatment I would find it more accurate to write B = B(t) × (f − G)·Δt.
L144: I suspect this is largely dependent on the forcing used. Surface NO₃ can differ by more than an order of magnitude between biogeochemical models.
L157: Please specify the basic characteristics here, in particular, whether assimilation is included or not.
L189–L190: I do not see a comparison with the information obtained from this cruise.
L193–198: For a one-month integration, the distributions are very similar, so it is difficult to assess the model's efficiency. Could you provide a more quantitative metric? See, for example, Berline et al. (2019). This would allow you to refine the choices of windage and of the vertical integration of Stokes drift.
Citation: https://doi.org/10.5194/egusphere-2026-3097-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 20 | 7 | 2 | 29 | 1 | 0 |
- HTML: 20
- PDF: 7
- XML: 2
- Total: 29
- BibTeX: 1
- EndNote: 0
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Major concern:
You validated the particle-tracking simulations by comparing the simulated dispersal areas with satellite observations. However, your model also incorporates larval growth and biomass change. How were these components validated?
It is unclear how your model improves upon existing approaches. I would suggest comparing simulations using transport alone versus transport coupled with the growth model. Or some other comparisons.
This manuscript does not appear to have been generated by AI, yet the writing still reads as somewhat informal in places. For example, there are too many paragraphs in this paper have only one or two sentences.
I would suggest separating the Conclusions and Discussion into two distinct sections. The Discussion should provide mechanistic interpretations of the results and place them in the context of previous studies, followed by a clear acknowledgment of the study's limitations and its implications for fisheries management. A thorough discussion in the context of previous studies is essential; otherwise, your work risks being disconnected from the broader field. The Conclusions, by contrast, should be concise and self-contained, presenting only the key findings of the paper without citing references or introducing new interpretation.
Minor revision
L4: what do you mean “customizable”?
Abstract: You need to rewrite this section.
L30: The significance of this study is mentioned in only one sentence throughout the Introduction, and also the last sentence of this Section, which I feel is inadequate and should be expanded.
L31-35, I would suggest rewriting these three sentences, as the third does not follow logically from the previous two.
L55: I suggest you list the two challenges clearly, which would improve readability.
Section 2.1: Is this a two-dimensional model? If so, please state this explicitly rather than leaving readers to infer it.
L77: delete “for example”
Equation 2: vStokes(xz=0,t) is the wave-induced Stokes drift velocity at the ocean surface, then what is vStokes(x,t)? At what depth is it evaluated?
Equation 4, what’s the value of vwind10? I guess it is not a fixed value, right?
Equation 1: what’s the value of vocean, how do you get it?
L131: You have lost the other half of the parentheses.
L194: “The simulation captures the areas of Sargassum convergence in the Central Atlantic (55°W-15°W in Figure 2) especially well.” Can you give a quantitative value of this?
L202: kN should be italic type.
L240: There are too many paragraphs in this paper have only one or two sentences.
L242: “temperature growth factor (left column in Figure 7) is much lower…”. How much lower is it?
L244: Wthat’s the value of optimal temperature?
Section Results: Please label all the geographic locations mentioned in the text on the figure. Readers should not have to refer back to the map to identify these locations.
Conclusion and Discussion
L264: “For example…”, the context after for example did not explain the sentence before “Our results indicate that the combination of physical transport and biological growth significantly affects Sargassum distribution.” There is no apparent logical relationship between these two sentences.
L269-275: What does this paragraph want to say? The main point of this paragraph is unclear