I think that the manuscript underwent improvements and got more focused, which is good. There are some structural problems (results, discussion, conclusions are in the introduction, see specific comments below) and language problems. In addition some additional clarification would help the reader.
Going back to original publication by Kollet and Maxwell (2006), I feel the paper's important point and novelty was not on a numerical approach (common or dual node or something else). The important point was the realization that the kinematic wave equation can be merged into the top flux boundary condition resulting in a free surface boundary condition at the land surface, very similar to the one used to obtain analytical solutions for a pumping test in unconfined aquifers. I feel this was really the message and as such it can be implemented in a model (numerical, perhaps even semi-analytical and analytical) in a consistent fashion to mathematically close the problem of subsurface flow as the authors wrote. Unfortunately, the numerical implementation is not very clear from the paper, but the sensitivity to the vertical discretization, which clearly is a disadvantage in case of excess infiltration, was transparently reported in the numerical test cases. Yet, this does not detract from the theoretical and practical appeal of the free surface overland flow boundary condition proposed in this paper. I feel that is something that should be pointed out in the ensuing revision of this manuscript. Note also the term “common node” approach was not introduced by Kollet and Maxwell.
I am surprised by the (sometimes) large difference in non-linear iterations, because in principle the common node and dual nodes approach are very similar (which is also the reason why they can be easily switched in the code as described in the manuscript). Which additional nonlinearity is introduced in case of the common node approach? A more satisfactory attempt of explaining this phenomenon would be needed in my opinion.
Provided the general comments above and more specific comments below, I am recommending minor revisions.
68: Is Kollet and Maxwell (2006) and appropriate reference for MODHMS?
100 - 127: These are results and discussion/summary/conclusions and do not belong in the introduction section.
137: ...the mesh...
220: What is fp in the equation?
232: ...dua nodes. The infiltration…
292: I am confused and not sure what the authors wants get at. When qr = I then qr = Kz, thus pss = 0, because there is a unit gradient at the right at the land surface. If pss < 0 then qr < I, contrary to equation 8 because qs is not zero. In case of qr > I, again qs is not equal zero and pss < 0 is wrong in this case. The physics and transient pressure behavior near the surface right at the onset of ponding is more complicated and has been discussed extensively in the literature (Kutilek and Nielsen). Storativity concepts need to be introduced if I remember correctly. If the author wants to look at these processes in the some limit, then this should be done in a mathematically rigorous fashion in my opinion.
376:
380: necessarily
441: Kollet and Maxwell also showed that in their original publication, no?
In figure 4 and 5, it is very difficult to distinguish between the dn and cn results.
Make sure to carefully check language and grammar, especially for typos. |