HESS 2017-314

Data-based mechanistic model of catchment phosphorus load improves predictions of storm transfers and annual loads in surface waters

Editor comments, 26.10.2017

Dear Dr. M. C. Ockenden

Thank you for the responses to the previous comments and the modification of the manuscript and the Supplementary Material. This has improved the manuscript substantially, clarified many issues and provides very useful information.

There is one single aspect where I see a need for further improvement before the manuscript is ready for publication. It relates to underestimation of the large TP loads (see Fig. S10, S12, S14) and the explanation you provide in your response (p. 4). You mention the length of the time series and the possibility for errors in rainfall estimates as potential causes. However, these arguments are not fully convincing because the discharge is much less affected by patterns of the residuals (compare e.g. Fig. S9 and S10). This suggests that there is an issue how the P mobilisation is simulated for such events.

This observation raises two questions: First, how does it come that a DBM model (provided the available data set) cannot capture this systematic pattern? Would a more complex model capture such a pattern but at the costs of too many parameters? Providing a (tentative) answer to this question can help readers to better understand the potential and limitations of the DBM approach. Second, what could be the physical or chemical process that was not properly captured or what was the simplification that goes with the model? Your thoughts and insights on that aspect may be highly relevant for further studies to include processes that may have been overlooked in the past or to represent them in a better way. I think a few comments on those two aspects would fit well into the first paragraph of 3.5 and further add value to your manuscript.


Sincerely

Christian Stamm
Editor HESS