Dear authors,

As you have seen, all 3 reviewers appreciate the general intention of your manuscript, but they also stress that it is not yet developed to the point. They point out the fact that, although simplicity is frequently required and even desired in hydrology (in particular for communicating results and interpretations to a wider audience, including policy makers), the model, as it currently formulated, may be too much of a simplification. The reviewers agree that it will be necessary to document the "road" to MIR in a more systematic way, illustrating and discussing which processes can be omitted and why, and which need to be in a model, as at this point MIR is a largely unsubstantiated claim. This could for example be done in a stepwise analysis, by one by one adding simple(!) representations of different biogeochemical (e.g. time-varying concentrations in the stores; more explicit processes linked to P or N, etc.) and hydrological (e.g. wetlands) processes. From this analysis it can then more objectively be inferred what the MIR really is to still allow for valid interpretations.
In addition I would like to encourage the authors to link their work a bit stronger to a wide body of recent literature on the topic, e.g. Benettin (2013; WRR), Botter (2010,WRR; 2011,GRL), Harman (2014,GRL; 2014,WRR), Hrachowitz (2013; HESS); van der Velde (2010,2012; WRR)

Please make sure, in addition, to adress *all* reviewers concerns in detail in the revised manuscript and to also provide a point-by-point overview of the responses.

Best regards,

Markus Hrachowitz

Dear authors,

all reviewers agree that the revised version of your manuscript showed considerable improvements, which is in particular true for the justification/explanation of your approach. However, the reviewers also point out several points that still need some attention (see reviewer reports).
In addition, I would encourage the authors to rethink/discuss some of the points below:

- Putting wastewater treatment plants and groundwater in the same box may reduce complexity of the model, but also makes it much less useful for water management scenarios.

- The fit between modelled and observed discharge is still hard to see (fitting a time series of 7 years into one window doesn't work).

- The bad fit between observed and modelled total phosphorus remains. The authors solved this by removing the scatter plot (fig. 5), which showed too much scatter, and refer to the time series plots. Besides the fact that "hiding" unsuitable results (even if it was not done on purpose) does not inspire much confidence, here again, time series of several years compressed into a small window already look well if you have the seasonal cycle right.

- The introduction is still too long and not focussed enough on the motivation for this study.

- Several things were explained to the reviewers (for example choices on model complexity), but not in the paper itself. Readers may wonder the same things.

- Overall, the paper does not give the impression of really careful work. The lay-out of the figures could be improved and the text can be more structured and shortened. The parts that were added after the review are not embedded in the original text and little text was removed or shortened.

Looking forward to a revised version,

best regards,

Markus Hrachowitz

Dear authors,

Thank you very much for your detailed and thoughtful responses, appropriately addressing the reviewers' remaining comments. Apart from a few technical corrections as detailed below I am thus happy to accept the manuscript for publication.

Technical corrections:

(1) As explained in detail in the author information section of the HESS webpage, multi-letter variables violate the style convention of HESS. Mathematical symbols and abbreviations should therefore be specified with one normal sized letter and if necessary further specified using sub- or superscripts. Please correct that.

(2) p.2, l.42 and reference list: should read as "Hrachowitz et al., 2013"

Best regards,

Markus Hrachowitz