The paper “Optimization of an Urban-like Monitoring Network for Retrieving an Unknown Point Source Emission” by Hamza Kouichi et al. has been re-assessed in light of the accepted Ngae et. al (2019) manuscript and the previous published articles Kumar et al. (2015a, 2015b, 2016). The current case was discussed with other topical editors and also with executive editors in addition to referee opinions.

The paper presents a further analysis of the renormalisation technique to the MUST dataset using the CFD model fluidyn-PANACHE (already published). The main result consist in reducing the network size form 40 to 13 and 10 sensors.
The points yet to be addressed belong to four categories:

1 - Technical: pertinence of using the same measurements for optimisation and validation and other minor points
2 - Originality of the results, analysis and conclusions.
3 - Applicability in practice of the results, analysis and conclusions
4 - Code description and availability

One reviewer has a point that has been partially answered by the authors, namely that the optimisation evaluation is done using the MUST set of measurements and this makes it more likely that the resulting sensor configuration performs well reconstructing the source( that “the same measurements shouldn’t be used for the optimisation and for the reconstructions”). A robust reply to this criticism is the core result of the QJRMS manuscript, that can therefore not be used as a reply in the context of the current review.

Some additional reviews considered that the distinct novelty of the manuscript under consideration in not put forward clearly enough with respect to the manuscripts Ngae et. al (2019) and Kumar et al. (2015a, 2015b, 2016). On the other hand, the scope can be adjusted in underlying that the current work addresses only part of the problem. In this case the text has to be adapted accordingly. The current presentation (and the title) suggests a very general treatment of the problem. This is contrasts with the fact that part of the results seem to be published in Ngae et. al (2019) and Kumar et al. (2015a, 2015b, 2016). In general it has to be highlighted so it becomes immediately clear to an average reader what is new in this manuscript with respect to the other two.

Otherwise, it may be put forward how in the view of the authors the results and conclusions can be applied in practice.

Remember that the article cannot be published if the authors do not make their code available. There appears to not even be a code availability section. This is not only referring to the CFD code, which is not the core of the article and has been published before. The authors have to make the pieces of code used in the core results, including the renormalisation and the optimisation. If possible, the scripts used for preparing the figures presenting the main results have to be provided as well. The executive editors clearly state: "The paper must be rejected if the authors refuse to comply with requests to make the code accessible within the requirements of GMD."

I suggest the authors to revise the manuscript in a way that makes the criticisms irrelevant.
This includes reworking the introduction (and also the discussion) narrowing the scope in order to describe more accurately what the paper delivers and how is is related to the previously published works. I suggest to explain and discuss more explicitly the results and conclusions of the works previously published by the group (i.e. Ngae et. al (2019) and Kumar et al. (2015a, 2015b, 2016).) and precise the relationship of the results and conclusions presented in the current manuscript.
The main outstanding point is however the code availability. The paper cannot be published without it. A possible way forward in the context of GMD is to put the accent in the description of the code (and reproducibility) as it seems that none of the previously published works the code is sufficiently documented.


minor/specific points:

p 1

l1 “sensors measuring the polluting substances” > sensors measuring polluting substances
l2 “environment with a view to estimate an unknown” > environment in order to estimate an unknown
l3 “ The methodology was presented by coupling the” > The methodology is presented by coupling the
optimal network was analyzed by > optimal network is analyzed by

l10 “In 80% trials, emission rates with the 10 and 13 sensors networks were estimated within a factor of two which are also comparable to 75% from the original network.” this phrase is not clear enough for the abstract, please clarify.

l11 The last phrase: “This study presents an application of the renormalization data-assimilation theory for determining the optimal monitoring networks to estimate a continuous point source emission in an urban-like environment.” is background/introcution information, and should be combined with the first-second phrase.

l17 “by the concern authority.” The concerned authorities?

p 2

l1 “The problem to optimize a monitoring network is common and consists in reducing the size of a network of sensors at the level of a city, county, or a neighborhood while retaining its properties.”
This statement contrasts with the following aspects of monitoring networks optimization: First deployment, Updating an existing network, reducing the size of an existing network, increasing the size of an existing network.
Please clarify specifying the actual contribution of this work in the general context.

“This study will be focused with an objective to reconstruct an unknown continuous point source’s release in an urban-like environment.“
This contrasts with the stated narrower focus of reducing an existing network of sensors for source location and intensity estimation.
This observation is relevant because the inverse methodology was already introduced in Ngae et. al (2019) and Kumar et al. (2015a, 2015b, 2016).
These papers are about inverse modelling on fluidyn‐PANACHE fields in the MUST case applying the renormalisation inversion technique. It has to be clear from the beginning what is the new contribution of this work that adds to what has already been published or has been submitted and accepted (namely the network reduction?).

l6 “an optimally monitoring sensors network” something seems missing in this grammatical construction.

l9 “This study presents a methodology for the sensor’s locations choice, leading to the best network for the estimation of 10 an unknown point point source in a geometrically complex urban environment.” You have to

The “the” article in “The SA algorithm was designed for the statistical physics” is spurious > “The SA algorithm was designed in the context of statistical physics?”

p3

l13 “advanced search algorithm*s* like genetic algorithm “
l18 replace “..., etc.” with a concrete reference or omit.

p3

l8 “ This study deals with a case of reducing the number of sensors in order to obtain an optimal network from an existing network.” This sentence is key, because it describes precisely what is done in the study. This information has to be clear much earlier and in the abstract.

Sections 2 and 3: provide the working code. This is a fundamental GMD requirement.

Section 5: You don’t have to provide the CFD code if it has been published elsewhere, but if the code described in 2 and 3 requires sample input files, those have to be included, at least for testing purposes.

Section 6: Ideally scripts for the results should be provided for reproducibility.



References

Kumar, P., Feiz, A. A., Ngae, P., Singh, S. K., & Issartel, J. P. (2015a). CFD simulation of short-range plume dispersion from a point release in an urban like environment. Atmospheric Environment, 122, 645–656. http://doi.org/10.1016/j.atmosenv.2015.10.027

Kumar, P., Feiz, A. A., Singh, S. K., Ngae, P., & Turbelin, G. (2015b). Reconstruction of an atmospheric tracer source in an urban-like environment. Journal of Geophysical Research, 120(24), 12,589-12,604. http://doi.org/10.1002/2015JD024110

Kumar, P., Singh, S. K., Feiz, A. A., & Ngae, P. (2016). An urban scale inverse modelling for retrieving unknown elevated emissions with building-resolving simulations. Atmospheric Environment, 140, 135–146. http://doi.org/10.1016/j.atmosenv.2016.05.050

Ngae, P., Kouichi, H., Kumar, P., Feiz, A.-A., & Chpoun, A. (2019). Optimization of an urban monitoring network for emergency response applications: An approach for characterizing the source of hazardous releases. Quarterly Journal of the Royal Meteorological Society. http://doi.org/10.1002/qj.3471