Review of “A simple approach to represent precipitation-derived freshwater fluxes into nearshore ocean models: an FVCOM4.1 case study” by Rutherford et al.

General comments:

This manuscript studies an important technical issue of how to represent freshwater fluxes accurately and efficiently in high-resolution nearshore ocean models. For coarse-resolution models, freshwater inputs are normally incorporated by aggregating individual contributions from various watersheds into one source and specifying it at one or two model grid points. This approach is definitely undesirable for high-resolution models in which explicit depiction of disparate spatial scales is at high priority. Using Quatsino Sound as a case study site and employing a relatively simple rain-based hydrological model, the authors perform a series of sensitivity tests to answer the two research questions they proposed. The topic is interesting and scientifically important within the geoscientific model development. The results derived here can be extended to other coastal areas and, thus, are worthy of publication. However, I would suggest the authors to clarify certain points and make some statements in the text more accurate.

Specific comments:

1) I would suggest that the title of the paper be changed to “A simple approach to represent precipitation-derived freshwater fluxes into nearshore ocean models: a case study of Quatsino Sound, British Columbia”.

This method is not unique to FVCOM. If you change to another coastal ocean circulation model, the proposed rain-based hydrological model can still be used. In the meantime, the conclusions of the paper, such as those listed in the Abstract, is only valid for Quatsino Sound.

2) This paper only considers the very simple rain-based hydrological model. Actually, within the framework of FVCOM, a more accurate way to estimate the effect of precipitation-derived freshwater fluxes on fjord salinity dynamics can be done. In this approach, nearshore ocean model domain will be enlarged to encompass all the watersheds in the study area. Then using FVCOM's wetting and drying capability to simulate over land flow due to rain events using FVCOM’s precipitation and evaporation forcing. Have the authors tried this approach? Of course, it requires a lot more computational time.

Do the authors consider evaporation or evapotranspiration, in addition to precipitation?

3) Lines 170-173, do you consider the freshwater falling on the surface of the numerical domain (i.e., the fjord system)? If not, the simulated salinity field will be biased.

4) Lines 228-249, the authors should give a definition of “mean” or “average” in the paragraphs here. Otherwise, it will take the readers a lot of time to try to figure that out.

5) Lines 297-298, “All other sensitivity tests had metrics in between those of the Marble River Only and All Rivers simulations”. This is definitely a wrong statement, which is not consistent with the numbers (e.g., Willmott Score) quoted in the text. Fig. 9 is another source to check with.

6) Lines 310-313, to make sure this statement is correct, you can either use the general vertical coordinate in FVCOM simulation, and/or greatly increase the number of the vertical layers. Have you tried these?

Technical corrections:

1) Line 128, add “on” before October 14^th, 2021

2) Line 131, add “on” before October 14^th, or (it would be better) add “at 00:00 am on October 14^th”

3) Line 162, delete “in each watershed”

4) Line 169, Equation (3). Should it be A_HRDPS,j  the denominator and A_WS,j the numerator?

5) Line 319, add “as” after “as long”

6) Figure 4 caption, “in equation 2” or in equation 3?

7) Figure 6, for right-hand side panels I would suggest to change the color scale to blue color only because, I guess, no positive difference in surface salinity exists in the result.

8) Figure 7 caption, is this called a histogram? I can understand that Figure 8 is called a histogram, but not this one.

9) Figure 9, a Table may be a better choice than a Figure

General comments:

The paper, “A Simple Approach to represent precipitation-derived Freshwater Fluxes into Nearshore Ocean Models: an FVCOM4.1 Case Study,” presents a straightforward solution that significantly improves the representation of hydrological influence in ocean circulation models. The authors demonstrate a keen understanding of the simplicity of this solution that would be beneficial to oceanographers when improving the results of their model. For example, the decision to utilize the grid of the atmospheric model HRDPS aligned well with this principle. While not perfect, this approach eliminates the need for remapping operations and mitigates other potential sources of error, such as water conservation.

Furthermore, the authors explore various gaps in the knowledge of the river network, including the watershed area and river mouth location. They provide valuable insights into where efforts should be concentrated to enhance the efficiency of coastal circulation models. Despite these strengths, the authors remain cognizant of the existing gaps in their rivers model.

Specific comments:

• The river proxy currently relies solely on precipitation data. Incorporating evaporation data (P-E) could enhance the proxy’s accuracy in predicting the volume of water in the river system.
• The frequency of the river proxy used in the FVCOM model is unclear. While some analyses were conducted at a high frequency (as shown in Figure 4), others were done on a monthly basis (Figure 5). Clarification on the frequency of these inputs implemented in the model would be beneficial.
• Insufficient time was dedicated to validating the CIOPS-W model in the region under study. This raises concerns about whether the CIOPS-W model itself could be a source of errors from the initial states and the boundary conditions used in their coastal circulation model.
• The paper frequently mentions that the river proxy is rain-based and does not account for snow coverage. Conducting a climatological overview of the region, particularly regarding winter snow coverage or snowfall, could help assess the extent of this limitation in the current model.
• Validation of the properties of deeper water should be considered. While the 0-50m depth is directly affected by river flow, this could also influence the content of deeper water through estuarine circulation.
• The modelling of water storage will primarily delay and temper peak precipitation, rather than limit the total volume of water flowing into the river system (line 176).
• A comparison with the Marble River at a higher time resolution than monthly could provide more detailed insights (line 196). One suggestion could be to combine Figures 4 and 5 for this purpose.
• The paper states that the All Rivers configuration performs better, but then suggests that only 60-75% of the total freshwater source is necessary (line 324). An explanation of this apparent contradiction would be helpful.