After I had a close look at your manuscript I dare say that I found your study deals with a highly relevant topic, presents a rather sophisticated data set and provides very interesting insights on the potential of in-situ remediation of contaminated sites. The study is hence without doubt worth to be published in HESS.

Nevertheless, I think in line with the reviewer Marnik Vanclooster, that the representativeness of single point observations in different depths for water and reactive solute dynamics in the entire vadose profile is not at all straightforward, particularly as the VMS does not observe along a single vertical flow path, but across several parallel ones at different depths. This implies that a plot of concentrations at different depths into a single profile and particularly the fit interpolation between these the concentrations is misleading because it suggests that these observation points are connected by a vertical flow paths, and concentrations in depths between these points can be estimated using a depth dependent interpolation. This would be correct a) in case the perchlorate concentrations are homogenous in a given depth, and b) the flow and transport process during irrigation would be a rather uniform. At least the latter is against my personal experience (and of many others), and the FTDR data also indicate a rather in homogenous wetting process. Note that the increase of the green FTDR values (supposed to be in a shallower depth) strongly differ from the increases of the read graphs. My first idea is that this is due to bypassing, after a while I realized, that those observations are not on the same flow path. I appreciate that you point already pointed this out in the manuscript, but still the Figure might pull the reader on the wrong track (at least I stepped into the pit). By the way, please add a legend to this Figure relation the color codes to the depths and position of the FTDR.

In summary I think that particularly the point about the representativeness of the VMS data for the entire domain needs to be carefully addressed and discussed within the revised manuscript. This does not necessarily imply to use a model if the authors have a different idea, but using a model is certainly way to address this issue. I appreciate that the authors already corroborated the reproducibility of the dynamic reactions in time in their response to Reviewer II, but this does not imply that observations are also representative in space at a given depth. In this context the authors might consider water balance closure as means to address this issue – for instance by balancing the storage changes and the irrigation amounts. A closed water balance would indicate a representative monitoring.
I also wonder to which extent the increased perchlorate concentrations in the larger depths are caused by leaching from above locations – as the total irrigation amount was 2.5 times the annual precipitation amount.
Looking forward to receive the revised manuscript,

Erwin Zehe

Dear Dr. Dahan,

I am delighted to let you know, that both reviewers recommend publication of your study, after very minor revisions, which I am sure you are willing to implement.

Allow me to say, that I am happy to have such a strong contribution to vadose zone hydrology in our journal.

Best regards,

Erwin Zehe