The oxygen concentration DO has decreased in the subsurface waters of the Gulf and Lower St. Lawrence Estuary where the latter has developed hypoxic conditions. The reason for the deteriorated oxygen conditions seems to be an increase of remineralization of organic matter OM and a reduction of DO in the inflowing deepwater. The latter was earlier determined by an about 40/60 mixture of cold oxygen rich Labrador Current Water and warmer, oxygen poor NADW but this ratio of mixture has changed to about 0/100.  

The idea is to supply oxygen gas to the inflowing deepwater in the Cabot Strait for artificial re-oxygenation of the Laurentian Channel bottom waters. For this one must know the response in DO in the Laurentian Channel bottom waters to a specified supply of oxygen gas in the Cabot Strait. To this end, a one-dimensional (1D) advection – diffusion model has been applied. To get confidence in applying any model, model results must be verified using observations from the area where it is applied.

The rate of change of DO along the flow path depends on (i) the rate of oxygen consumption OUR by mineralization of OM, (ii) the rate of oxygen supply by inflow through Cabot Strait, which is determined by the advective flow speed u and the concentration of DO of the inflowing water (boundary concentration), and (iii) the rate of supply of oxygen by turbulent vertical diffusion from the oxygen rich Cold Intermediate Layer overlying the deepwater. Turbulent vertical diffusion is known to take place essentially at the bottom boundary, where most of the vertical mixing occurs due to breaking internal waves, often driven by internal tides generated at sloping bottoms and steps in the bottom.  Vertical mixing at bottom boundaries creates horizontal buoyancy gradients that drive transversal circulation, not described by a 1D model, that distributes the effects of mixing to the whole water body. Changes of DO due to changes in the rate of change of turbulent vertical mixing are hard to show. If the turbulent mixing is driven by the internal tide, it may change if the vertical stratification changes. This could be discussed in the manuscript.

The 1D model is tuned using historical data, and data from the large-scale tracer experiment TReX in the Bay of St Lawrence.  It is required that the model can describe the distribution of DO along its path. If it can, one may have confidence in model results when changing the boundary concentration of DO at Cabot Strait.

Horizontal diffusion, but not vertical diffusion, is included in the model. The argument for discarding vertical diffusion is that is has a time scale of 30 years while the horizontal time scale is 5 years. However, the vertical time scale is estimated using the vertical diffusivity K_z = 1x10^-5 m²s^-1. The horizontal mean K_z is maybe larger because one may expect high values at the boundaries (hot mixing spots). With K_z = 1x10^-4 m²s^-1, the vertical time scale would be only 3 years. This should be discussed in the manuscript because vertical diffusion possibly may provide a significant contribution to the DO budget of the deepwater.

The model describes quite well the observed year to year changes in DO in the deepwater using only known changes of the DO concentration in the Cabot Strait. The model uses a constant UOR. One would expect that UOR might be greater in the inner part of the St. Lawrence River Estuary due to possibly greater production of OM here due to nutrient supply by the river. It would be interesting if the authors could discuss the sensitivity of model results to the assumption of a constant UOR. It would also be interesting to know if there are large variations in the annual supply of nutrients from the St. Lawrence River, and the expected annual supply of OM to the deepwater.

The physics of the model, i.e. the current speed u and the horizontal diffusivity K_H, has been calibrated using results from the transient plume of the tracer experiment TReX. Since the duration of the tracer experiment was shorter than the residence time of the deepwater it was necessary to include horizontal diffusivity to describe the observed spreading of the tracer. For a quasi-steady description of DO it might possibly be more important to include vertical diffusion since vertical diffusion might contribute to the DO budget of the deepwater. The authors should discuss this and estimate the uncertainty of model results due to the explicit ignorance of vertical diffusion.

I am sorry for the delayed review, which is due to personal circumstances. I really enjoyed reading this manuscript.

Anders Stigebrandt

General comments

This study reports the spatial and temporal variation of dissolved oxygen (DO) dissolved inorganic carbon (DIC) in the lower St. Lawrence estuary and the Gulf of St Lawrence based on observational and model data, and simulated the scenarios of pure oxygen injection to mitigate the hypoxia status. The logic is clear and it is well written. I have only one concern. The simulation is based on steady state. However, when the pure oxygen is injecting, the stratification is destroyed. This may lead to the sequence that the model parameters may no longer be appropriate. Additionally, is there oxygen leakage during the injection?

Specific comments

Lines 47-49: The number of “(“ and “)” are different, i.e. two “(” but only one “)”.

Lines 358-359: From the DIC data of 2021, 2022 and 2023, I don’t agree that the DIC accumulation rate is 18.3 ± 2.5 μmol kg⁻¹ yr⁻¹.

Line 408: Should “(Tanioka and Matsumoto, 2020)” be “Tanioka and Matsumoto (2020)”?

Dear Authors, 

This study, entitled “The coupled oxygen and carbon dynamics in the subsurface waters of the Gulf and Lower St. Lawrence Estuary and Implications for Artificial Oxygenation,” is a study with lots of efforts based on more than a decade of observational data. The authors employed along-channel surveys and carbon isotope (δ¹³C) analysis to investigate the relationship between dissolved oxygen (DO), dissolved inorganic carbon (DIC), and their isotopic signatures in the study region. The authors have addressed a broad range of perspectives relevant to the topic.

The work further examines the role of decomposition processes in shaping the DO-DIC dynamics and develops a numerical model to estimate the oxygen residence time in this coastal system. 

Overall, this is a high-quality paper. However, the structure of the manuscript—especially the integration of Results and Discussion—could be improved. The authors are encouraged to separate the discussion from the results to enhance clarity and allow for a more focused interpretation of the findings. In addition, the scientific significance and conceptual framing of the study could be more explicitly stated, especially in the Abstract and Conclusion.

________________________________________

Major Comments

1.    Format and Structure

The formatting between sections 1 and 1.1 is unusual and should be revised for consistency.

Sections 3.4, 3.5, and the last paragraph of the Conclusion could be revised into a dedicated Discussion section. Currently, the discussion is fragmented and lacks cohesion. A clear and standalone discussion would improve the paper's overall impact.

Consider incorporating a paragraph that synthesizes past findings and highlights how this study builds on previous work. This could serve as a conceptual model-like summary and strengthen the narrative arc from historical data analysis to future outlook.

2.    Purpose and Abstract

While the final paragraph of the paper presents a clearer articulation of the study’s aim, this clarity is not reflected in the Abstract. The Abstract currently focuses too much on methods and lacks a high-level synthesis of the findings and implications.

Revise the Abstract to emphasize better the study’s scientific significance, including the conceptual contribution and potential applications.

3. Discussion

Some explanations—for example, differences in coefficients used in prior studies—are mentioned but not thoroughly discussed. Please elaborate on the potential consequences of these coefficient differences on the model output or interpretation.

A more comprehensive discussion could highlight the novelty of this long-term dataset, its value for understanding biogeochemical cycling, and its implications for future monitoring or management.

Line 424. This appears to be the key assumption of the entire study. Could the authors comment on any potential side effects or unintended consequences associated with this assumption? 

Model Limitations are mentioned in lines 493-497. Although the authors include error estimates, the limitations of the numerical model are not clearly discussed. Given that short-term variations (e.g., tidal or diurnal changes) may exceed long-term trends in certain locations, the authors should discuss how these short-term dynamics may impact model accuracy and interpretation.

Figures

Many figures are overly crowded. For instance, Y-axis labels are dense, and contour lines frequently overlap with annotations.

These issues are relatively easy to address and would significantly improve the readability and professional appearance of the figures.