the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 15 May 2025
Tracking Marine Debris in Northwest Spain: Assessing Wind Influence with a Lagrangian Transport Model
Abstract. Marine debris is responsible for major problems in our oceans, causing serious environmental degradation, detrimental health effects and economic losses in sectors related to the marine environment. In this study we investigate the influence of wind forcing on the transport, accumulation and beaching of floating particles in the Ría de Arousa, an estuary on the northwest coast of the Iberian Peninsula. Using Lagrangian simulations of particle tracking under different wind drag coefficients (1 %, 3 % and 5 %), we evaluate the spatial and seasonal patterns of particle concentration, residence time and deposition on the coast. Our results show that wind plays a crucial role in modulating particle behavior. Low wind-driven conditions favor greater near-shore accumulation and longer residence times, especially in the northern and inner regions of the estuary. As wind influence increases, particle dispersion intensifies, leading to lower overall accumulation and weakening of correlations between river discharge and coastal deposition. Seasonal differences are also studied, with higher concentrations observed in the north during winter and in the south during summer.
- Preprint
(2621 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-2057', Anonymous Referee #1, 09 Jul 2025
- CC1: 'Reply on RC1', Martiño Rial-Osorio, 05 Sep 2025
-
RC2: 'Comment on egusphere-2025-2057', Anonymous Referee #2, 02 Oct 2025
Preprint: https://doi.org/10.5194/egusphere-2025-2057
Review:
Tracking Marine Debris in Northwest Spain: Assessing Wind Influence with a Lagrangian Transport Model
Martiño Rial-Osorio,Vicente Pérez-Muñuzuri, and Sara Cloux
The article addresses a topic of considerable environmental and scientific relevance: the transport and accumulation of plastics in estuarine systems under wind influence. The study combines Lagrangian simulations with varying windage coefficients to evaluate seasonal accumulation patterns, residence times, and correlations with river discharge in the Ría de Arousa (NW Spain). The work is well-structured, and the results are consistent with previous literature, providing valuable insights into the processes governing the retention of marine debris in estuarine environments.
Nonetheless, before publication, several aspects require clarification and further detail, particularly with respect to methodological details and contextual framing. Overall, I recommend the manuscript for publication, subject to minor revisions.
Some points that I believe require clarification or improvement:
1. Introduction
In the manuscript, the distinction between macro- and microplastics is presented in centimetres. However, the standard unit, used in the literature, is typically millimetres, as indicated in the references cited by the authors themselves (Min et al., 2020; Wayman & Niemann, 2021). I recommend aligning this definition with the units in the literature, to avoid terminological inconsistencies.
2.1 Study Area
The criteria used to divide the study area into the sections shown in Figure 1 are not clearly explained. Does this segmentation arise from hydrodynamic differences, socio-economic characteristics, or another specific parameter? The authors should clarify the rationale behind this segmentation to justify its relevance.
While I understand the choice to present simulations for winter (DJF) and summer (JJA), as these represent extreme conditions, it would be advisable to also include wind roses for the intermediate seasons (MAM and SON) in Figure 2, in order to provide a more comprehensive view of the seasonal variability of the wind patterns analysed in the study. Additionally, since the results also refer to these intermediate seasons, their inclusion would be pertinent.
Considering that the entire study is based on particle discharge proportional to the flow of the Ulla River, it would be important to provide, in this section, information on the average, minimum, and maximum river discharge values.
2.2 Input Data
Although the Numerical Methods section specifies that simulations were performed using 2D surface fields, I consider that this information should also appear in the Input Data section. This would prevent the reader, upon first encountering the description of the hydrodynamic model, from assuming that vertical processes or particle sinking were also considered. A simple statement indicating that only 2D surface fields were used would make the methodology clearer from the start.
2.3 Numerical Methods
The adoption of an 80% beaching probability is a central aspect of the study, yet the manuscript does not provide the rationale for this choice. It would strengthen the work if the authors explained this choice in more detail. I understand that a high retention rate may be associated with rocky coasts and intertidal zones, where stranded plastics tend to remain longer due to lower remobilization. However, in the correlation section (Table 1), the analysis focuses mainly on sandy beach segments, where higher remobilization would be expected. Therefore, it would be useful to clarify this apparent discrepancy and discuss the choice of the 80% value for the different types of coast included in the study.
A key element of the work is the analysis using different windage values (cw = 1%, 3%, and 5%). However, this choice could be better supported with additional references or, ideally, observational data. Providing such justification would increase the robustness and relevance of the adopted parameterization (e.g., https://doi.org/10.1016/j.envint.2020.105655 and https://doi.org/10.1038/s41598-018-22939-w).
Results and Conclusion:
Figure 3 is not entirely informative. The current colour palette limits the ability to discern differences between areas of higher particle accumulation, which reduces the figure’s interpretative value.
The authors state that particles tend to concentrate upstream due to the Coriolis force. While this is correct, it would be helpful to explicitly relate it to the resulting residual circulation. Including a figure illustrating the residual circulation would significantly enhance the discussion and provide clearer mechanistic insight.
There are also concerns regarding the discussion of the results, which primarily focuses on a single winter and a single summer scenario with different windage values. Given that, particle release is substantially higher in winter, Figure 5 appears largely redundant. If the goal is to assess whether beaching is more frequent in summer or winter and under varying windage conditions, it would be more scientifically robust to compare scenarios with similar particle release rates across seasons. Alternatively, considering particle age—the time from release to beaching—across all scenarios would provide a more meaningful and grounded comparison.
Addressing these points would greatly strengthen the manuscript by clarifying the drivers of particle accumulation and providing a more rigorous basis for seasonal and windage-related comparisons.
The definition and discussion of residence time appears to be somewhat unclear. From my perspective, this parameter seems strongly influenced by the size of the areas considered and the number of particles released. In principle, a larger number of particles or a more extensive area would tend to increase residence time, but this relationship is not necessarily true from a hydrodynamic standpoint.
Moreover, the calculations also appear to depend on the initial time of the simulation and the number of particles released. For example, if the simulation begins in January, the residence time estimated for this first month will inevitably be shorter than in March—not due to hydrodynamic conditions or wind, but simply because fewer particles are present to be account for.
The interpretation of Figure 6 is also not entirely clear. In the innermost zone of the estuary, residence time may be higher due to proximity to the source. In the intermediate zone, however, since particles that return to this area are counted again, the probability of this occurring is naturally higher, particularly, given the way the estuary has been subdivided. This hypothesis could be further supported by including a figure showing the Eulerian residual circulation of velocities, which would provide better context for the results.
Additionally, according to the methodological criterion that particles leaving the study area are considered “dead,” the downstream region is excluded from the residence time calculation, which limits interpretation for this outer area.
Therefore, it would be important to clarify how residence time was computed or, alternatively, to explicitly highlight these limitations in the discussion to avoid potentially biased conclusions.
I would encourage the authors to explicitly discuss the limitations of the implemented methodology. For instance, it is not clear whether a horizontal resolution of 300 m is sufficient, considering the dimensions of the zones chosen as accumulation areas. Similarly, the potential effects of three-dimensional processes should be addressed: to what extent might a 3D simulation alter the results compared to the current approach?
The role of waves in particle distribution within the estuary, as well as in the beaching process, also warrants consideration, since they may significantly influence particle trajectories and residence times.
Finally, it would be valuable to discuss the generalizability of the methodology: how could it be applied to other study areas, and are there comparable studies in different regions that support or challenge the current approach? Addressing these points would strengthen the manuscript by providing a clearer context and by acknowledging the methodological constraints.
Citation: https://doi.org/10.5194/egusphere-2025-2057-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 664 | 56 | 14 | 734 | 8 | 28 |
- HTML: 664
- PDF: 56
- XML: 14
- Total: 734
- BibTeX: 8
- EndNote: 28
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This paper investigates how the distribution of floating plastic debris in the Ria de Arausa that originates from Ulla River is affected by the windage term used in Lagrangian simulations. Using this model, they find that a lower wind drag results in higher nearshore accumulation and larger retention rates. Furthermore, they find that the beaching pattern is highly heterogenous, where over half of the beached particles stranded in only 10 of the 242 coastal segments of the Ria. The number of beached particles at these segments is correlated with the river flow rate, where they find highest correlation for the sections closest to the release point of the particles.
I think in general this set up of this study allows for interesting analysis, however I feel the analysis presented here is not complete. In addition, the way model settings are chosen and why there are valid is not sufficiently explained currently. There is no discussion on the limitations of the model used here and what this implies for the obtained results. Due to these points, I think mayor revisions are needed before this work is published. I provide the following comments and questions about the manuscript, which hopefully the authors will find useful.
Comments/questions per section:
Numerical model:
Fig 1: How were the inner/intermediate and outer region chosen? From the plot, it seems that they are not equally spaced relative to the river mouth.
Fig 2: Why only show the wind roses of DJF and JJA? In the result section also the periods MAM and SON are considered and based on the results of Fig 6 it seems that the SON period has interesting wind characteristics. In addition, it would be good to not only show the wind but also the (average) Eulerian currents, as the particle distribution is strongly driven by this current. This would help with understanding the results.
Is a homogenous beaching of 80% everywhere realistic? One would expect that the beaching is a function of the specific coastal features. I think the choices that went into setting this beaching parameter have to be part of the paper.
The effect of stokes drift is not separately included. Is the assumption here that this is captured by windage? How big is the effect of swell in this region? I feel this has to be discussed in the paper.
Why were windage factors of 1%, 3% and 5% chosen, and how realistic is this range compared to the windage of real buoyant plastics? I think the work would benefit from comparing these values by typical windage factors measured in experimental settings (for example for surface drifters) to show the relevance of this range.
Why was the river input represented as a point quite far into the Ria, and not further back where the actual river mouth is located? Why were the particles not released over a line that spans the entire width of the estuary at the release point? With the current release point, there are coastal segments located behind the river input location, including segments 6 and 10. Which raises whether beaching of particles measured at these segments is realistic for plastic that originates from the river.
I think the section on how particles were simulated should contain more information. Were particles simulated for a specific time (1 season/1 year) or were they simulated until they were beached or left the domain? And if particles were simulated for the entire 5 years, how does this affect the analysis of the accumulation pattern done? As in the 5th year, the concentration of particles then will have increased.
Results:
Fig 3: the way the data is plotted makes it hard to compare, maybe a different color bar with more shades (like magma or viridis) and/or using a discrete colormap can help to see differences. In addition, it might be useful to plot the relative concentration change for the 5% relative to the 1% instead. It is not clear whether the concentration shown here is for particles that might be in a beach cell but are yet definitely beached? Or are particles that are permanently beached also considered?
You mention that the concentration of the northern part of the estuary is higher due to the Coriolis force. I think it would be better to say due to the currents/ or explain how the Coriolis force affects the currents in the estuary. For this it would be helpful to plot the Eulerian currents (and the seasonal variations therein if present) as well (see comment for fig 2). For plastics there can also be a direct Coriolis force working on the particle because of their inertia (see the Coriolis force in the Maxey-Riley equation for example on the work of Beron-Verra et al. 2019) thus the wording Coriolis force might be confusing.
How are the northern and southern regions defined? Are they right, resp. left compared as viewed from the river mouth in direction of the Ria? And is the Island then part of the northern or southern region? Or is it an actual north/south distribution, which seems a bit arbitrary as it does not align with the shape and orientation of the Ria? In general, the south has more area located further from the river mouth, thus that would automatically result in less beaching as particles already beached close to the river mouth cannot beach there.
Fig 4 (and the discussion thereof in line 120-128) The discussion states that “during warmer months (MAM and JJA) beaching along the southern coast increases”. However, the absolute value of particles beached decreases compared to the DFJ period as well for the south? Maybe of the total particles beached, the relative amount of particle beached in the south increases, but this is not a relevant quantity to highlight? Also, how does the number of beached particles relate to the total input for each period? I think this relative number would give more insight on if particles are more likely to beach in summer or winter or that there are just more particles released into the system in winter but beaching stays constant.
Fig 5 and table 1 and discussion on the correlation: Is the correlation the best quantity to measure? If you put a lot of particles in the system, you expect that a lot of particles beach a period T after you put them into the system. This period T will be a function of the total time a particle needs to reach the area. Thus, I do not think figure 5 is the most interesting figure to show. Also in table 1, it is expected that the time delay grows with distance and thus the direct correlation will become less. In addition, the regions selected are not suitable to reveal any non-trivial spatial pattern (especially as 6 and 10 are sort of located at/in the river mouth for the location of particle release here). I think a more useful quantity to study is the cross correlation and then measuring the time lag. Or since you do Lagrangian analysis you can calculate the age of the particle (i.e. how long ago they were released) at the point in time that they reach land/beach. This would be an analysis that can be done for all coastal segments, revealing if there are any patterns different from what you expect (further away from the river mouth the particles are older). You could consider normalizing the age with the distance from the coastal segment to the particle release location.
Figure 6. It seems that the inner, intermediate and outer regions are not equally spaced and/or of the same size. This would affect the results. I would expect that for the outer region there are particles that do not reenter the Ria and thus have a residence time of infinity. Or are particles removed from the simulation if they leave the simulation domain? Which is not similarly shaped to the inner and intermediate region set here. Are beached particles considered in the calculation of the residence time?
Conclusions:
You state that wind also disperses particles more widely, but you do not show/highlight this in your result section.
Maybe highlight that the release location strongly affects the patterns of beaching in the Ria as your results are very different from Cloux et al. 2022 fig 4.
There should be a discussion on the limitations of the model used here and how these might affect the results presented here.