Review Reports

Round 1

Reviewer 1 Report

General comments:

A major issue in this study is that all the analyses were carried out at only three receptors. In similar previous studies (e.g., Pirovano et al., 2015; Ciarelli et al., 2017), they used 9 and 11 sites to evaluate the proposed methods, respectively. However, in this study, are results from only three receptors enough to convince your audiences the superior performance of the proposed HMS? The authors need to handle this issue (or well discuss this limitation) before it can be accepted.

Specified comments:

---Page 3, Figure 1: Please improve the flowchart, it could be drawn better than the current one.

---Page 3, line 107: "driven by ECMWF analysis" What does ECMWF mean? No explanation for this term.

---Page 4, Figure 2: This map is not professional. Please draw this map in a professional manner (e.g., using ArcMap or QGIS, and insert north arrow and a better scale bar). Besides, it would be good if you could add three zoomed-in pictures showing detailed surrounding environment of the three receptors in SI.

---Page 12, line 386: "in Figure Error! Reference source not found.7 that allows appreciating"

---Page 1 line 11, Page 2, line 80 & Page 16 line 470: This manuscript mentioned many times that: ".....in the reconstruction of the spatial distribution of fine particles (PM2.5) and elemental carbon (EC) concentration...." and "This work is focused on air quality modelling......." and ".....was set up for air quality assessment and source apportionment....". Could you please also add some modeling results (e.g., annual mean PM2.5 spatial distribution map or annual mean EC spatial distribution map) at least in SI.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments on the manuscript entitled as “Combined Eulerian-Lagrangian hybrid modelling system for PM2.5 and elemental carbon source apportionment at the urban scale in Milan” by Lonati et al.

September 17, 2020

 

General comment

The source appointment of the pollutants in the urban area is quite important to improve the horizontal distribution and the budget of emissions of precursors, and it would be useful for the precise estimation of air quality and its impact on the human health in megacities. The target of this paper well adapts to the scope of the journal. However, I can not find the evaluation of some meteorological parameters (wind direction, boundary layer height, precipitation etc.) and model configurations (wet deposition, vertical mixing within PBL etc.), which are quite important for the transport of materials, in the manuscript and related papers. I think the authors should add some information in the revised manuscript.

 

Specific comment

1. Page 3, line 108: There are several types of “ECMWF analysis” products. The authors should specify the name of reanalysis data (e.g. ERA-interim, ERA-5, etc.). I am also wondering why the authors did not use 1-hourly data of ERA-5.
2. Page 4, lines 133-134: I can not find the information for the treatment of the wet deposition process in AUSTAL2000 in this manuscript and Pepe et al., 2016. I think a wet deposition process is important for some of PM, such as sulfate, nitrate, and hydrophilic BC. The target of this manuscript is the source appointment, and the turbulent mixing within the PBL might also quite important for the precise estimation of the transport pathway near the surface. The authors should add information for the treatment of vertical mixing in AUSTAL2000 in the revised manuscript. I think the evaluation of wind direction, PBL height and precipitation, which are not shown in this manscript and Pepe et al., 2016,  might be useful to show a more concrete conclusion.
3. Page 5, lines 186-187: The authors say that “it should be noted that the contribution of local precursor emissions to local concentration of secondary PM is substantially negligible”. It seems it does not agree with the results of Jiang et al., 2019 (Atmos. Chem. Phys., 19, 15247–15270, 2019), which showed that the contribution of SOA over the Po valley region is estimated to be 40%(winter) to 80%(summer) using CAMx. The author should add an explanation for the possible differences by neglecting SOA.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have addressed most of my concerns.

However, I still have one suggestion for this manuscript.

The authors explained 'a 1.7 km2 urban..........enough to test........'.

The three receptors are indeed enough to validate this specific area (1.7 km2).

However, can this small area represent the complex spatial variations of air pollution within the whole urban? The urban has many sub-areas, and the distribution pattern in each area could be totally different. 

I am not saying your model has any problem. But there is a potential risk of over representation that the model's performance (designed for the whole urban) was only validated by three receptors in a small area. 

Therefore, the authors need to state this issue as a limitation of this study in either the discussion or conclusion chapters (two or three sentences are enough).

Author Response

We thank the reviewer for this remark and we agree that the validation through just three receptors, even though reasonable for the small area of our case study, may be questionable for model application at the whole urban area scale.

In order to point out this limitation, the following paragraph has been added in the conclusion:

"In this work, the enhanced performance of the HMS has been assessed for three selected receptors, located in a relatively small area of Milan city centre and representative of different exposure to emission sources (park receptor: urban background; traffic receptor: heavy urban traffic; Duomo receptor: urban traffic, commercial and residential emissions). However, this promising result is someway limited by the few receptors considered and the limited extension of the modelled area by the Lagrangian component. Therefore, these first outcomes need to be further confirmed by future work where, extending the computational domain of the Lagrangian component, a larger number of receptors (e.g: including residential areas outside city center, city outskirts, main access roads to the city) can be considered, in order to account for the complex spatial variability of air pollution within the whole urban area. Future work could also include the development of alternative local scale modelling layers, based either on Lagrangian modelling approach, such as puff models, as well as on geostatistical-based methods, both allowing the extensions of the analysis to larger portions of the urban area."