Dear Authors,

I received one review on your revised manuscript.
Given the recommendations in this review and the overall mediocre rating of the present manuscript, I believe that more changes to the manuscript are necessary than the ones from the last revision round.
Relative to the initial submission, the number of figure has not changed and the page number even increased in revised version.
Considering the fact that both reviewers (of the first round) criticised the length of the manuscript, I strongly encourage to condense the content and make it more digestible for future readers.

Kind regrds,
Simon Unterstrasser

That's the review report of the second round:

Although some improvements have been made to the mansucript, I would further encourage the authors to
a) better explain the rationale for selecting the scenarios analysed
b) reduce the number of figures and make the analysis more accessible to the reader
c) better highlight the added value of LES modelling and put findings in relation to existsing studies evalution urban planning scenarios.

L323: not clear. What are “materials used in reality”?
Section 2.7.1 and 2.72 – the description and rational for the design of scenarios is still not entirely clear. While “synthetic” scenarios are designed to understand the impact of parameter uncertainty, the real life scenarios are evaluating planning strategies. However, more details should be provided explaining why the specific scenarios are selected. For example, why would you test a scenario without any vegetation? Is this considered a “worst-case example”? Maybe the term “real life” scenario is not suitable if the setups are actually not all realistic such as when removing roads but not the cars.

Dear authors,

I appreciate the latest text modifications.
Unfortunately, there is one thing that has to be fixed and which I probably overlooked in the beginning.
GMD is strict about reproducibility. It is required that you provide all scripts, code and data that allows other to reproduce your plots.
This includes configuration files of all performed simulations in addition to the source files given in https://data.uni-hannover.de/dataset/palm-6-0-r4093. In theory, anyone should be able to re-do your simulations.
Moreover, all presented plots must be reproducible.
Best practise is to bundle all plot scripts together with the necessary input data (i.e. the output of the PALM simulation that is used in the plot procedures) in a Zenodo data set.
See for example: https://data.uni-hannover.de/dataset/6687444f-a3a7-45b8-8453-a39995b8e4d4

Best wishes,
Simon

Dear Michael and co-authors,

I am happy to accept your paper on urban flows simulated with PALM.

Congratulations and good luck,
Simon Unterstrasser