Open-Source Code for Radium-Derived Ocean-Groundwater Modeling: Project Open RaDOM

Round 1

Reviewer 1 Report

A very interesting manuscript that could be applied in various studies of groundwater (mixing models), beyond the issue of submarine groundwater discharge. 

1. I suggest changing the title: The title of this manuscript should be more explicit. If I only read the title, I don't understand what the topic being studied is (flux, submarine groundwater discharge).

2. The authors might consider this article:

https://doi.org/10.1016/j.jhydrol.2012.06.020 (Methodological study of submarine groundwater discharge from a karstic aquifer in the Western Mediterranean Sea)

3. Lines163-164 The authors could add examples of these conservative ions (tracers) that could be applied in this case (I can think of chlorides)

4. Lines 196 -289 (examples from the literature). I would recommend including a comparative analysis between the results from the literature and the results obtained with this code using the same data from the literature. That comparison could be presented in a table.

Author Response

First, I would like to thank the reviewer for taking the time to read my manuscript and comment on it.  I realize scientists get asked to review a lot of papers these days, and their time is valuable. 

A very interesting manuscript that could be applied in various studies of groundwater (mixing models), beyond the issue of submarine groundwater discharge. 

1. I suggest changing the title: The title of this manuscript should be more explicit. If I only read the title, I don't understand what the topic being studied is (flux, submarine groundwater discharge).

The title is something I have gone back and forth on for awhile, wondering if it was specific enough.  I’ve added “Groundwater” to the title since this is specific to groundwater applications. 

2. The authors might consider this article:

https://doi.org/10.1016/j.jhydrol.2012.06.020 (Methodological study of submarine groundwater discharge from a karstic aquifer in the Western Mediterranean Sea)

This is a good example of using thermal anomalies to detect SGD.  I’ve added it as a reference for that method.

3. Lines163-164 The authors could add examples of these conservative ions (tracers) that could be applied in this case (I can think of chlorides)

I don’t actually think chlorides would work well in this model since its typically lower in concentration in groundwater than ocean water, but barium or silica could work.  I will add those with some references. 

4. Lines 196 -289 (examples from the literature). I would recommend including a comparative analysis between the results from the literature and the results obtained with this code using the same data from the literature. That comparison could be presented in a table.

I’ve added a whole new section called “Case Studies” that does exactly this.  This was recommended by another reviewer.  Unfortunately for the box model results are unclear as I had to guess at some of the end-members, but the mixing model results turned out well.  I believe the box model results are still valid in that they show how the models can be used, even if they aren’t representative because I had to guess some of the end-members.  I will be adding these scripts to the supplemental materials as well as to the online workspace (https://rstudio.cloud/content/3560079).  I would suggest the reviewer as part of their review.  There’s instructions (with screen shots) for running the scripts in the supplemental materials.  Once you work through those instructions it should be pretty clear how to run any of the case study scripts I’ve provided.  Those case study scripts are in the folder named “Literature Examples” in the workspace. 

Reviewer 2 Report

In this paper, analytical methods used for SGD flux calculation or evaluating the SGD contribution is discussed, however, the application of numerical modeling methods are not included. The author should provided more detail about the numerical modeling researches to explain the innovation of the proposed method. In addition, case studies and corresponding calculation results of the Project Open RaDOM should be introduced. 

Author Response

In this paper, analytical methods used for SGD flux calculation or evaluating the SGD contribution is discussed, however, the application of numerical modeling methods are not included. The author should provided more detail about the numerical modeling researches to explain the innovation of the proposed method. In addition, case studies and corresponding calculation results of the Project Open RaDOM should be introduced. 

First, I would like to thank the reviewer for taking the time to read my manuscript and comment on it. 

I have added this information to the “General Approach” Section.  I am utilizing an already existent numerical technique, which an entire book and many papers exit about.  So, I have kept the explanation short and to the point of what users of this code realistically need to know:

“All models are solved numerically using the Lawson-Hanson algorithm for non-negative solutions [18,19].  This ensures only positive solutions or solutions of 0 can be returned by the model, eliminating unreal solutions (like negative residence times). This algorithm also allows for an over-constrained model in which the number of tracers exceeds the number of unknowns.  This makes it possible to integrate all tracers into a single model instead of iteratively solving for pairs of tracers yielding multiple different results.  The Lawson-Hanson algorithm takes an iterative approach in which the solution is updated each iteration until the solutions begin to vary little from one iteration to another, con-verging on the solution and terminating the script, at which point the results are output. In this way the results are approximated.

Residuals are output each time the model is run so that goodness of fit can be evaluated.  The closed the residuals are to 0 the better the fit of the model to the data, and the more the results can be trusted.” 

I’ve added a whole new section called “Case Studies” that integrates data from previously published papers into Project RaDOM scripts.  This was recommended by another reviewer.  Unfortunately for the box model results are unclear as I had to guess at some of the end-members for each example, but the mixing model results turned out well.  I believe the box model results are still valid in that they show how the models can be used, even if they aren’t representative because I had to guess some of the end-members.  I will be adding these scripts to the supplemental materials as well as to the online workspace (https://rstudio.cloud/content/3560079).  I would suggest the reviewer as part of their review.  There’s instructions (with screen shots) for running the scripts in the supplemental materials.  Once you work through those instructions it should be pretty clear how to run any of the case study scripts I’ve provided.  Those case study scripts are in the folder named “Literature Examples” in the workspace.