Spatial-Temporal Variation Characteristics of Vertical Dust Flux Simulated by WRF-Chem Model with GOCART and AFWA Dust Emission Schemes (Case Study: Central Plateau of Iran)

Round 1

Reviewer 1 Report

This study is about spatial and temporal variation characteristics of Vertical dust flux simulated by WRF-Chem model in the Central plateau of Iran. There are two dust events are represented here. I have some detailed comments listed below to help improving this article.

1, It would good to list the correlation coefficient between AFWA scheme with the horizontal visibility even the coefficient is low or even negative. Are both spatial and temporal correlation addressed in the study? For AFWA, neither spatial or temporal correlated to the visibility?

2, Figure1, the city name is not clear, please enlarge the map and the city names.

3, Figure3, the city names are not clear.

4, Figure5, again, the city names are not clear.

5, How about the relationship between AOD and the dust concentration?

Author Response

Thank you very much for your very accurate review. Your comments were very interesting and helpful. This article was taken from my PhD student thesis. I used your comments as much as possible to improve the article. Thank you for your good review.

1, it would good to list the correlation coefficient between AFWA scheme with the horizontal visibility even the coefficient is low or even negative.

We now also list the correlation coefficient between AFWA scheme with the horizontal visibility in the text.

Are both spatial and temporal correlation addressed in the study? Yes

 For AFWA, neither spatial or temporal correlated to the visibility? No

2, Figure1, the city name is not clear, please enlarge the map and the city names.

Done. To better show the city names, we numbered them and explained them in the title below

 3, Figure3, the city names are not clear.

Done. To better show them, we numbered them and explained them in the title below

4, Figure5, again, the city names are not clear.

Done. To better show them, we numbered them and explained them in the title below

5, How about the relationship between AOD and the dust concentration?

AOD on these dates had missing values ​​due to the presence of clouds in the area, as seen in the images below. We now explain this in the text.

 

The presence of clouds in the study area has caused missing values ​​in AOD

Author Response File: Author Response.docx

Reviewer 2 Report

See attached file.

Comments for author File: Comments.pdf

Author Response

Thank you very much for your very accurate review. Your comments were very interesting and helpful. This article was taken from my Ph.D. student thesis. I used your comments as much as possible to improve the article. Thank you for your good review.

Major comments

1. Why did the authors choose the Iran region to test the model?

The Middle East is one of the five regions of the world which has the most dust production (Rezazadeh 2013). Iran is exposed to various dust systems due to its location in the dry and semi-arid belt of the world (Rashki 2013). Dust phenomenon is one of the most important problems in Iran. The existence of Loot Desert and the Central Desert and climatic conditions have caused this phenomenon to occur frequently. Iran is on the world's dry belt. The depopulation of some villages is a result of this phenomenon in this area In recent years, the phenomenon has intensified due to severe droughts. Therefore, to identify this phenomenon, the identification of dust sources and transport is of great importance. This is now explained in more detail in the manuscript Introduction.

 

2. How did they identify the selected case studies?

Using criteria based on low horizontal visibility, and its duration and spatial extent, dust events at the regional level have been selected. These events were selected using the statistics of 36 synoptic stations and over the period of 2006 to 2018. The most important factors to identify sever dust storms were low horizontal visibility and dust expansion in the study area.

 

3. Why did they use the Ackerman index?

 

The Ackerman index was used to confirm the widespread of dust in the region during the selected storm periods. This index has been shown to have a good performance for detecting dust storms around the Central Plateau of Iran.

 

4. Why did they use the concentration/horizontal visibility relationship to validate the model instead of comparing modeled vs. measured concentrations or modeled vs. measured Aerosol Optical Depth (AOD - either from inside measurement or from satellite)?

 

Horizontal visibility decreases as the dust concentration increases. Therefore, the estimated concentration using the model was compared with the horizontal view at some stations (Shao et al 2007). However, this was not the only validation, and MERRA2 data on dust concentration were also used.

AOD on these dates had missing values ​​due to the presence of clouds in the area, as shown below.

The presence of clouds in the study area has caused missing values ​​in AOD

 

Minor comments

1. Flux units: μg/m²s

Title changed to: Spatial-temporal variation characteristics of Vertical Dust Flux simulated by WRF-Chem model with GOCART and AFWA dust emission schemes (case study: Central Plateau of Iran)

2. Added summary (lines 63 to 90)
3. We added the references (Alizadeh Coobari 2013, 2014)
4. Reference added
5. Thomson et al 2006 removed
6. Azizi et al 2012 removed
7. The importance of dust storms in Iran was explained.
8. The abbreviation MODIS is explained as Moderate Resolution Imaging  Spectroradiometer
9. We added the following explanation:

In this study, the results of GOCART and AFWA wind erosion schemas were validated using horizontal visibility and simulated surface concentration. One of the most common methods of validation is the use of AERONET and AOD data. AOD had missing values on the mentioned dates and the reason for this was the presence of clouds in the area. Also, there was no AERONET data on these two dates in the region. Therefore, it has been forced to use horizontal visibility and surface concentration. As the surface concentration increases, the amount of horizontal visibility decreases. MERRA2 reanalysis dust concentration data were also used for validation.

10. Dust storms were described in the study area

 

11. The materials and method section was modified and the requested items were added
12. HYSPLIT calculations were calculated online for two events with an accuracy of 0.5 degrees for 24 hour back trajectories.

 

13. Table 1 moved to Section 2.3.2

 

14. The output of the model has been compared for two different schemas and has not been used for verification.

 

15. Data for Zahdan and Zabol stations were not available for the two selected dates.

 

16. In addition to these two schemas, there is also the Shao schematic. But to keep the volume of content in this paper manageable, additional schemes were not included. Based on previous work that we cite, these two schemas have been widely used worldwide and in the Iran region and were therefore selected to be examined.

The first domain was 27 km and the second domain was 9 km. The time interval between networks was 180 seconds. GFS data with a 6-hour time interval was used for initial and boundary conditions. The model output was also saved every 3 hours.

 

17. If f (roughness) is equal to 1, it means a smooth surface. This value increases with increasing rock, vegetation, and other roughness elements.

 

18. The reason for choosing February 2015 and 2018 was due to heavy dust on these dates, as well as to better compare two events within the same season.
19. The figures were corrected.

 

20. Extra figures have been removed.
21. The order of the shapes was modified.
22. The title of the figures has been corrected.
23. Suggested sources were used.
24. Acknowledgment was corrected
25. References were corrected.
26. The spatial pattern legend is now set to maximum and minimum.
27. The time-scale (Y-Axis) was changed in the graphs given to coincide.

 

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

This version is fine. 

Author Response

Dear Reviewer

Thank you so much.

Best

Authors

Reviewer 2 Report

See attached file.

Comments for author File: Comments.pdf

Author Response

• Major comment:

1. The study falls into the scope of Applied Sciences journal as it presents numerical simulations of desert dust in a region where these aerosols are of primary importance for human population and ecosystems. The aim of the paper is interesting as it gives an insight on the sensitivity of desert dust emission modelling to the dust emission scheme.

Response: Thank you so much

 

2. The authors made many corrections that I asked in my first review, and I thank them for that. However, there are still lots of improvements to be done before the manuscript may be suitable for publication.In particular, in the “aims and scope” section of the journal, it is stated that “The full experimental details must be provided so that the results can be reproduced.” In the present form of the manuscript, a lot of details allowing to reach this goal is still missing (see the minor comments below).

Response : We, the authors, make every effort to make the most of your comments to improve the article.

 

3. After reading again the paper, as results from the two presented case studies are identical, the authors should focus on one event; the second one could be presented in Appendix.

Response: Given that the amount of dust simulated  by both schemas ( GOCART and AFWA) in two events is different , and the number and location of the dust centers are different in both cases, I think if you agree, either Two cases are currently available.

 

4. Finally, there are again several figures that are presented, but not discussed, so that they must be removed. For all these reasons, I recommend the publication of this paper in Applied Sciences but only after the questions raised and comments listed below had been addressed.

Response: Done, Thank you so much for your recommend

• Minor comments:

1. The authors didn’t understand correctly the first correction I asked. The units of the vertical dust flux is still incorrect: it is not “µg/m²s”, but “µg/(m² s)” (parenthesis are mandatory) or “µg m-2 s-1”.

Response: done

 

1. Introduction

Main comment (again): a lot of studies have already used WRF-Chem to model dust cycle. The authors

should summarize l. 70 to 88 focusing on the key points. In the new version of the manuscript, more

references were added, but the interesting points raised by the different studies were removed. I write

again my previous suggestion: As an example, the authors should write something like that: “Several

studies have demonstrated the suitability of the WRF-Chem model to simulate the desert dust cycle

(e.g., Nabavi et al., 2017, Chen et al., 2018). However, attention must be paid on the chosen land

surface model (Rizza et al., 2018), model vertical resolution (Teixeira et al., 2016)…” At the end, the

authors should cite the numerical studies that have already been conducted in the studied region using

WRF-Chem concluding that it was able to reproduce the observed dust loads and spatial patterns.

Response: Some references were summarized. I also used your comments. Some references were deleted and a number of references were added. I tried to apply your opinion

Minor comments:

1. 48-49: references are still (ex l. 49 to 51): references are still missing to justify the “decrease in soil

fertility” and “damage to crops and natural vegetation”.

Response: This text has been deleted

1. 56: a reference is needed: where does this number come from?

Response: Approximately 88% of Iran is located in arid and semi-arid regions (Vaghefi et al 2019)

Vaghefi, S.A., Keykhai, M., Jahanbakhshi, F. et al. The future of extreme climate in Iran. Sci Rep 9, 1464 (2019). https://doi.org/10.1038/s41598-018-38071-8)

1. 89-98 (ex l. 94 to 100): I clarify my previous comment: what do you conclude from that? You can’t

just make a list of references.

Response: You are right. corrected.

2. Material and methods

2.2 Selecting dust events

On which time period was this selection done? Please clarify. Once done, this subsection should be

presented as follows: i) synoptic data, ii) MODIS data, and iii) determination of dust storm origin using

HYSPLIT.

Response: Done

 

2.2.1 Synoptic data:

1. 135-137 (ex l. 128-129): the given explanation is still not sufficient. I copy my previous comment:

Was there a minimum number of stations that must have recorded a horizontal visibility less than

1000 m to class a day as “dust storms”? If yes, how did the authors choose this threshold number? Did

the authors use the “weather codes” presented in Tab. 2 (the one l. 213) to class a day as “dust

storms”? If yes, please move this Table 2 in this subsection, and clarify. If no, this Table 2 must be

removed.

 

Response:No. Because the desert and plain area and the number of stations are limited. This mode was only used to extend the event, and the farther away the dust source was because the dust concentration decreased, the higher the horizontal visibility, but it could still be a measure of the extent of the storm.

Table 2 was moved and the necessary explanations were given about the meteorological codes for the selection of dust storms event.

2.2.2 MODIS data

1. 145 (ex l. 130): The reference corresponding to the selected MODIS product must be included (MODIS

Characterization Support Team (MCST), 2017):

MODIS Characterization Support Team (MCST), 2017. MODIS 1km Calibrated Radiances Product.

NASA MODIS Adaptive Processing System, Goddard Space Flight Center, USA:

http://dx.doi.org/10.5067/MODIS/MOD021KM.061

Response: Done

Moreover, the selected product must be described in the text of the manuscript: what are its temporal

and spatial resolutions? Collection?

Response: added in Text

1. 152: please clarify in the manuscript that equation 1 is the definition of the Ackerman Index. For

instance, add the following: “The Ackerman Index is defined as follow:”.

At the end of this section, the authors must include the answer they gave to my question: “Why did

the authors choose to use the Ackerman Index instead of the classically used Aerosol Optical Depth to

detect dust storms?”

Response: Done

2.2.3 Determination of dust storm origin using HYSPLIT

The references to Rolph et al. (2017) must be added:

Rolph, G.; Stein, A.; Stunder, B. Real-time Environmental Applications and Display sYstem: READY.

Environ. Modell. Softw. 2017, 95, 210–228, doi:10.1016/j.envsoft.2017.06.025.

Response: Done

The following pieces of information are still missing and must be added in the manuscript: Which

database did the authors use to compute the backtrajectories? At which resolution? When did the

backtrajectory analysis begin? How long did it last? Why are the origins of backtrajectories different

between the two events?

Response: Done(added in the manuscript)

2.3 WRF-Chem model

2.3.1 Dust schemes in WRF / Chem

1. 151-152: the other available dust schemes available in WRF-Chem must be cited in the text and then

you have to justify why you chose to use GOCART and AFWA only in this study.

Response: UOC scheme (Shao 2001, 2004 and 2011). In this study, according to the volume of the study and also according to the previous research (Hossein Hamzeh et al. 2019), GOCART and AFWA schemas were considered suitable for this region.

Hossein Hamze. N, Karami. S, Ranjbar.A, 2019. Simulation of a severe dust storm with different dust emission schemes. E3S Web of Conferences 99, 02013 (2019) CADUC 2019,  https://doi.org/10.1051/e3sconf/20199902013

2.3.1.1 GOCART dust scheme

1. 176: This is not “Passy”, but “Passi”. To be corrected.

Response: Done

Equation 3: the authors that Fp = 0 otherwise. What does “p” in “Fp” stands for?

Response:   Done  (Fp=0)     Where F_p is (μg m−2 s−1) represents the emission flux for size bin p

Table 4. Dust particle diameter distribution by category in WRF - Chem model.

Type of dust particle Effective particle diameter ( Particle diameter range ( Particle density (g/cm3) bin1 0.73 0 – 1.0 2500 bin2 1.4 1.0 – 1.8 2650 bin3 2.4 1.8 – 3.0 2650 bin4 4.5 3.0 – 6.0 2650 bin5 8 6.0 – 10.0 2650

 

1. 177: This is not “C”, but “CG”. To be corrected.

Response: Done

 

1. 178: rephrase.

Response: Done

1. 179-180: “which is the function of particle size, air density and soil moisture”. What does it refer to?

Response: is the threshold velocity (m s⁻¹ ) below which dust emission does not occur and is a function of particle size, air density and surface moisture(kumar et al 2014).

1. 180-181: Ginoux et al. (2001) never stated that. This sentence is totally wrong and must be deleted.

Response: Done

1. 183-184 (ex. l. 165-167): to be rephrased.

Response. Deleted

2.3.1.2 Dust Scheme AFWA

1. 190: please correct “ ∗” to “∗”.

Responces: Done

1. 191-192 (ex l. 176-177): why is the f(roughness) set to 1 for southwest Asia? What will be the

consequences on the simulated dust flux?  

Response: Because the great deserts of Saudi Arabia, Iraq, Syria, Iran and etc. have been created in this region. Vegetation in these areas is very limited and the land surface is bare and wind erosion systems are active in these areas .

And as the values) f(roughness ))decrease, the amount of emission flux also decreases. Because Value decreases with increasing amounts of large rocks, cobbles, vegetation, etc (Jones et al 2012).

Dust emission is restricted to areas with roughness length Z₀ ≤ 5m (typically barren lands and sparsely vegetated surfaces)

1. 194-196: please, give the units of each parameters. What is the value of C? Please correct “U*

Response: Done.

Where C is an empirical constant (0.129),  is the density of air parcel, (g cm⁻³),  is the acceleration due to gravity (cm s^-2),  and  (in units of cm s⁻¹)

2.3.2 WRF-Chem Model Configuration

1. 206: the acronym USGS has to be defined.

Response: Done

1. 202-208 (ex l. 200-201): the answers given by the authors to the following questions must be included

in the text: at which time step are the computations done? At which time step is the lateral nudging

done? At which time step are the model results saved? To be specified.

Response: Done

2.4 Verification of results

1. 220: the acronyms AERONET and AOD have to be defined. The following reference has to be added

to AERONET:

Holben, B.N.; Eck, T.F.; Slutsker, I.; Tanré, D.; Buis, J.P.; Setzer, A.; Vermote, E.; Reagan, J.A.;

Kaufman, Y.J.; Nakajima, T.; et al. AERONET—A federated instrument network and data archive for

aerosol characterization. Remote Sens. Environ. 1998, 66, 1–16.

Response: Done

1. 223: give the references of the MERRA2 re-analysis.

Response: Done

3. Results

3.1. Dust storm selecting

1. 227-228: this sentence is not sufficient. Once again, the first thing to be discussed here is: on the

studied period (2006-2018), how many dust storms did the authors identify? Then, among the

identified dust storms, the authors must explain why they selected the cases of 21 February 2015 and

14 February 2018. Once this have been done, the authors can present and discuss Figure 2. Considering

the answer of the authors, I am very surprised that only 5 dust storms were identified over the 12

studied years.

Response: For the period 2006 to 2018, the number of dust events in the study area was high But not all of them are used for simulation. Because it is not possible. Therefore, based on three indicators (the extent of dust storm (based on dust codes in synoptic stations), dust intensity (the rate of decrease in horizontal visibility below 1000 or 500 meters) and durability (dust duration) 5 events (2 events in summer, 2 events in winter and 1 event in spring) were identified to implement the model. Two winter events were used for this paper to ensure that the seasonal conditions are the same in both dates.

Figure 3: there are many stations where horizontal visibility are > 1000 m. This questions the

methodology used by the authors to define a dust storm.

Response: The further away from the dust source, the lower the dust concentration and the higher the horizontal visibility. However, there is a reason for the spread of the dust phenomenon, and it may not indicate the intensity of the storm.

3.2 Detection with MODIS Images

Figure 4 must be discussed.

Response: Done

1. 239: insufficient.

Response: Done

4. Discussion and conclusions
5. 412-416: discussion should include a comparison with the studies by Alizadeh Choobari et al. (2013;

2014).

 

Response. Done

References:

The reference “Lio et al. (2006)” cited l. 47 (ex. l. 48) is still missing.

Response: deleted

Figures:

When the domain presented in the maps is larger than the one presented in Figure 1, please add the

country borders and the shore lines.

Response: Done

Figure 3: check label of x-axis.

Response: Done

Figures 6, 8, 10, and 12: what do numbers stand for? Please add a table to give the correspondence.

Response: Referee 1 commented that I should display the names larger. To do this, I entered the watershed codes and named them in the guide. According to your comment, I have now put it on the table.

Author Response File: Author Response.docx