the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 11 Apr 2022
Insight into seasonal aerosol concentrations, meteorological influence, and transport over the Pan-Third Pole region using multi-sensors satellite and model simulation
Abstract. The Pan-Third Pole (PTP) owns complex geography and demographic features where aerosol roles and their impact cannot be neglected as it jeopardizes both the environment and human health. Therefore, we analyzed spatio-temporal aerosol concentration, the influence of meteorological conditions, and underlying aerosol transport mechanisms over the PTP by leveraging observation, satellite dataset, and model outputs. The observation and model simulation result showed that aerosol concentrations exceeded the world health organization (WHO) and China guideline values in most of the locations. This study revealed distinctive seasonality with the highest and lowest aerosol concentrations during the winter and summer seasons, respectively, which could be favored by meteorological conditions and emissions from biomass burning. In response to higher aerosol concentrations, the maximum aerosol optical depth (AOD) values were observed over the major hotspot regions however, interestingly summer high (AOD > 0.8) was observed over the Indo Gangetic Plain (IGP) in South Asia. The columnar aerosol profile indicated that the higher aerosol concentrations were limited within 1–2 km elevation over the densely populated regions over South Asia and Eastern China. However, the significant aerosols concentrations found to be extended as high as 10 km could potentially be driven by the deep convection process and summer monsoon activities. Regionally, the integrated aerosol transport (IAT) for black carbon (BC) and organic carbon (OC) was found to be maximum over SA. Noticeable OC IAT anomaly (~5 times > annual mean) found during spring that was linked with the biomass burning events. Yet, the dust transportation was found to be originated from the arid land and deserts that prolonged especially during summer followed by spring seasons. This study highlights the driver mechanism in aerosol seasonality, transport mechanism, and further motivates the additional assessment into potential dynamic relation between aerosol species, aerosol atmospheric river, and its societal impact.
- Preprint
(3484 KB) - Metadata XML
-
Supplement
(243 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on acp-2022-199', Anonymous Referee #1, 05 Jun 2022
Review comments on ACP-2022-199
Insight into seasonal aerosol concentrations, meteorological influence, and transport over the Pan-Third Pole region using multi-sensors satellite and model simulation
by Rai et alUsing observation, satellite and model outputs, the authors investigated the Spatio-temporal aerosol concentration, the influence of meteorological conditions, and underlying aerosol transport mechanisms over the Pan-Third Pole region. They demonstrated the seasonal distribution and spatial variability of aerosol across this wider domain and tried to explore model performance. Even though they have mentioned and highlighted the whole PTP region apart from IGP and China, there is nothing demonstrated in this study over this wide domain. At this level, there isn't enough scientific proof or knowledge enhancement to properly connect the underlying facts as well. With the following factors in mind, I feel the manuscript will be enhanced if the authors make the following changes and, more importantly, if they can eliminate numerous unnecessary discussions.
Recommendation: Major revision
1) One would expect, from the title of the manuscript, that the authors had conducted research on the Tibetan Plateau and Himalayas region in addition to outer PTP region. However, they have only some information about India and China. So this is an unnecessarily misleading part of the title, abstract and even introduction of this study; this should be avoided.
2) I'm not sure how authors can make a very crucial assertion like AOD values are maximum across IGP when two models and two observations produce four findings. Despite the fact that WRF-Chem shows that northern China has higher AOD values than IGP, the other three choices show that the signal endures between seasons. As a result, such a remark in the abstract is not appreciable but also misleads readers.
3) In this study, the authors emphasise aerosol concentration over IGP. If a similar quantity of aerosol concentration is evident above China, however, it is rarely emphasised. This is not a scientifically sound practice.
4) A large drop in PM2.5 is seen in major Indian cities, according to Sing et al 2021 (10.1016/j.scitotenv.2020.141461). However, the authors make no mention of this in the manuscript. The recent downturn in China, on the other hand, is explicitly acknowledged.
5) I am not sure if there is anything unknown to unravel the effect of meteorological conditions on the spatiotemporal distribution of aerosol over the Himalayas. Many studies have already documented these factors in these regions. Furthermore, they stated in each part that the results are comparable to past studies. So, it's evident that they're simply re-inventing something that's been done before. Any such exaggeration should be avoided.
6) Again, I'm not sure why the authors place so much focus on the Third Pole location. Seasonal Spatio-temporal fluctuation of aerosols and total AOD is just discernible over IGP or China in the signals from Figures 4 and 5. As a result, neither the title nor the abstract should be expected to reveal this.
7) If the model has a dust aerosol bias, which is indicated as being resolved by masking, how can authors assure that this is not causing problems in surrounding regions through transport?
8) Due to the low-quality figure in figure 6, the circulation pattern is difficult to interpret.
9) If authors are perfectly aware that the bias comes from the chosen domain, why do they continue to use it? It appears that these are unsubstantiated claims with no scientific basis. The conclusion is littered with ambiguous assertions that should be avoided.
Citation: https://doi.org/10.5194/acp-2022-199-RC1 -
AC1: 'Reply on RC1', Mukesh Rai, 06 Oct 2022
1) One would expect, from the title of the manuscript, that the authors had conducted research on the Tibetan Plateau and Himalayas region in addition to the outer PTP region. However, they have only some information about India and China. So this is an unnecessarily misleading part of the title, abstract and even introduction of this study; this should be avoided.
Response: Thank you for the comments. We agreed and made necessary changes and add more information (l353-368).
2) I'm not sure how authors can make a very crucial assertion like AOD values are maximum across IGP when two models and two observations produce four findings. Despite the fact that WRF-Chem shows that northern China has higher AOD values than IGP, the other three choices show that the signal endures between seasons. As a result, such a remark in the abstract is not appreciable but also misleads readers.
Response: We agree. Taking this notion in mind that we have modified the misleading statements. Instead of just using IGP and East China, we have added other regions where AOD endures throughout the simulation period in (l42-43). We also added more information in the result section (l363-368).
3) In this study, the authors emphasize aerosol concentration over IGP. If a similar quantity of aerosol concentration is evident above China, however, it is rarely emphasized. This is not a scientifically sound practice.
Response: Thank you for highlighting this issue. We tried our best to avoid such discrepancies. Now we have modified this section (l353-362).
4) A large drop in PM2.5 is seen in major Indian cities, according to Sing et al 2021 (10.1016/j.scitotenv.2020.141461). However, the authors make no mention of this in the manuscript. The recent downturn in China, on the other hand, is explicitly acknowledged.
Response: Thank you for recommending this paper. We have included the finding from this paper in our manuscript (l112-114).
5) I am not sure if there is anything unknown to unravel the effect of meteorological conditions on the spatiotemporal distribution of aerosol over the Himalayas. Many studies have already documented these factors in these regions. Furthermore, they stated in each part that the results are comparable to past studies. So, it's evident that they're simply reinventing something that's been done before. Any such exaggeration should be avoided.
Response: Our study region comprises dynamic geophysical features including arid, semi-arid, and mountain to flat land. Previous studies either focus on the Southern flank of the Himalayas, the Tibetan Plateau, or South East Asia. Thus, we attempt to showcase how different meteorological parameters play a role in aerosol concentration distribution and transport mechanism in the synoptic scale which owns unique geophysical features.
6) Again, I'm not sure why the authors place so much focus on the Third Pole location. Seasonal Spatio-temporal fluctuation of aerosols and total AOD is just discernible over IGP or China in the signals from Figures 4 and 5. As a result, neither the title nor the abstract should be expected to reveal this.
Response: We thank the reviewer for pointing out the inconsistency in the result presentation of our manuscript. We sought irregularity in our manuscript and presented more results. We added more findings and discussed more in detail other than the Third Pole region (l351-357).
7) If the model has a dust aerosol bias, which is indicated as being resolved by masking, how can authors assure that this is not causing problems in surrounding regions through transport?
Response: Thank you for highlighting this issue. In line (l347), our choice of the word “mask out” was meant to be that our model did not mimic the general feature of AOD values over arid/semi-arid than the reanalysis dataset.
8) Due to the low-quality figure in figure 6, the circulation pattern is difficult to interpret.
Response: Thank you. Now, figure 6 quality is improved.
9) If authors are perfectly aware that the bias comes from the chosen domain, why do they continue to use it? It appears that these are unsubstantiated claims with no scientific basis. The conclusion is littered with ambiguous assertions that should be avoided.
Response: We appreciate the concern raised by the reviewer which is valid. The previous study is largely from South Asia and East Asia. Our motivation is to present the synoptic scale picture of aerosol distribution and transport dynamics over geographically complex regions. Yes, we are aware that bias is there but we intend to use modelling as a tool and showcase the general features of the aerosols and aid more information over the region. This study underscores the bias correction which was obviously at hand but due to computational cost that could not perform. Further studies are needed to improve the model performance, especially over complex topography.
-
AC1: 'Reply on RC1', Mukesh Rai, 06 Oct 2022
-
RC2: 'Comment on acp-2022-199', Anonymous Referee #2, 12 Jun 2022
The title, abstract, and introduction of the study ``Insight into seasonal aerosol concentrations, meteorological influence, and transport over the Pan-Third Pole region using multi-sensors satellite and model simulation'' by M. Rai et al. suggest an investigation of the seasonal properties and the origin of aerosol over the Pan-Third-Pole my means of satellite observations and model simulations. However, the study mainly presents seasonal means of PM2.5, PM10, and AOD at 10 AERONET stations and an unidentified number between 8 to 12 stations in China, a WRF simulation, CAMS data, MERRA-2 reanalysis, and VIIRS satellite measurements for a single year. While the motivation and envisaged scope of the manuscript would fit the scope of ACP the presented data and analyses do not. In my opinion this manuscript is just a collection of arbitrarily selected data. The data sets do not agree quantitatively, and in case of the WRF simulation not even qualitatively. Despite these discrepancies no attempt was made to assess the uncertainties of the simulation and measurement data to identify the source of the discrepancies.
Concerning the scientific quality I have numerous concerns:
In this study 3 different reanalyses were used: NCEP/FNL with WRF, MERRA-2, and GDAS with HYSPLIT. What was the motivation to do this? It is known that reanalyses differ. In my opinion using multiple reanalyses is a source for discrepancies. This should be taken into account when estimating the uncertainties and comparing the results/data sets.
The methods section (Section 2) is rather a description of the models used, but not of the data. Moreover, it is not clear how the models were set up and used and how the data was prepared. Concerning WRF, what is the purpose of the used models? Why have they been selected? Which parameters were used from the MERRA-2 data? What was the ``further treatment'' of the AERONET data? A description of the Chinese stations is completely missing. Figure 1 indicates 12 stations in China, but according to the manuscript only 8 stations were used (l 205). Which stations and why were four left out? I have the impression that the data and methods are not sufficiently described to reproduce the results.
Did the WRF simulation really produce negative PM2.5 concentrations? E.g. in Xi'an 66.5µg m-3 were measured, whereas the WRF simulation underestimated it by 93.6µg m-3 (ll 235-237). This should be clarified in the text.
Concerning the differences between the WRF simulation and the measurement the authors stated repeatedly some speculation (e.g. l243-246, 248-246). To investigate the effect of e.g. the meteorological data, a sensitivity test with a different reanalysis, which is obviously at hand, could be performed.
Figure 5 gives the impression that the WRF simulation does not perform well in reproducing the AOD. In winter and spring it underestimates the AOD and in summer and autumn it overestimates the AOD. In the text mainly the problems in simulating dust are discussed. However, given the large (a difference plot WRF-observations would be helpful) discrepancies between simulation and observations, I have little confidence in the further results derived from the WRF simulation.
I wonder why the seasonal pattern for PM2.5 and PM10 with maximum concentrations in winter and minimum concentrations in summer (Fig 4, first and second column) is the opposite to AOD with the maximum AOD in summer and the minimum AOD in winter (Fig5, top row).
It is still not clear to me how the trajectory analysis was performed. What was the purpose of doing forward and backward trajectories? Where were the forward and backward trajectories started?
While the title suggests that the focus of this study is put on the Pan-Third-Pole this study mainly presents and discusses the high aerosol load in densely populated regions (Indo-Gangetic Plain and East China) known for strong aerosol sources. The possible impacts of aerosol on the Pan-Third-Pole are stated in the introduction, but throughout the manuscript the authors avoid presenting the current knowledge on the sources and transport pathways and comparing it with their findings. They rather name studies (e.g. ll460 - 464; ll467-469) and remain vague. In my opinion the authors should contrast the state-of-the-art knowledge with their findings to make the advancements in their study clear.
Finally the presentation quality is poor. First of all most figures are too small. In most figures the axis labels, labels, and even data points are barely visible (even when zooming in to 200 %). The manuscript was difficult to read due the the numerous unnecessary abbreviations, some of them were not even introduced (e.g. AP, PG, CA, TD), others not used consistently (e.g. SA & South Asia), others introduced multiple times (e.g. VIIRS). It was also difficult to follow the line of argumentation while working through ``authors XYZ stated that'' at the beginning of consecutive sentences (e.g. ll438-444,). Moreover there are many avoidable mistakes (e.g. Fig. 1 ``dark blue shade'' is actually black, ll 209-212 the latitude and longitude of the stations is swapped). Finally the language certainly needs thorough revision.
In my opinion, fixing the above mentioned severe aspects is more than a major revision. Hence I suggest to reject this manuscript.
Citation: https://doi.org/10.5194/acp-2022-199-RC2 -
AC2: 'Reply on RC2', Mukesh Rai, 06 Oct 2022
1) In this study 3 different reanalyses were used: NCEP/FNL with WRF, MERRA-2, and GDAS
with HYSPLIT. What was the motivation to do this? It is known that reanalysis differs. In my opinion, using multiple reanalyses is a source for discrepancies. This should be taken into account when estimating the uncertainties and comparing the results/data sets.
Response: We appreciate the concern raised by reviewer. We acknowledge the fact of the discrepancy in our study. Carrying out model simulation in geographical complex region is challenging itself as our study area possess the diverse geographical features where observation data set are very limited. Thereby we wanted to leverage other available data source to assess, compare, and draw the conclusion in our study. The motivation of using reanalysis data set in our study are to provide meteorological initial boundary condition in model, compare with model and reanalysis data set, and to perform footprint analysis. To assess the model performance, we performed statistical metrics analysis of different pollutants and compared between model, observation, and reanalysis data (l 243-245, 278-288). The detail comparison of AOD observation versus model and reanalysis data sets are provided in Table S2.
2) The methods section (Section 2) is rather a description of the models used, but not of the data. Moreover, it is not clear how the models were set up and used and how the data was prepared. Concerning WRF, what is the purpose of the used models? Why have they been selected? Which parameters were used from the MERRA-2 data? What was the ``further treatment'' of the AERONET data? A description of the Chinese stations is completely missing. Figure 1 indicates 12 stations in China, but according to the manuscript only 8 stations were used (l 205). Which stations and why were four left out? I have the impression that the data and methods are not sufficiently described to reproduce the results.
Response: We have discussed the data used in section 2.4. Regarding the model set up and data preparation, we have discussed in 2.1 section. Further, key parameterization schemes adopted in our study are included in Table S1. The WRF-Chem model is selected because of its dynamic abilities such includes numerical more consistency, meteorology-chemistry feedback, capable of simulating and predicting pollutant concentration with physical and chemical options. As our model is forced to run in diverse geographical region (i.e. flat land, arid, semi-arid, and mountain region) which may require to have more represented parameterization schemes which should have more realistic presentation of different processes like advection, convection, diffusion, dry deposition, wet scavenging, aqueous and gas phase reaction, photolysis rates, other microphysics. We have used AOD and dust from MERRA-2. ‘Further treat’: we mean to say that we have used level 2 AERONET data however, due to data gap we further fill this void using Level 1.5 data. Now, the description of China station are included (l211-214).
3) Did the WRF simulation really produce negative PM2.5 concentrations? E.g. in Xi'an 66.5μg m-3 were measured, whereas the WRF simulation underestimated it by 93.6μg m-3 (ll 235-237). -This should be clarified in the text.
Response: Thank you for pointing out this issue. We have gone through the code and found out there was typo in the code. We have modified and presented the right values in (l243-246)
4) Concerning the differences between the WRF simulation and the measurement the authors stated repeatedly some speculation (e.g. l243-246, 248-246). To investigate the effect of e.g. the meteorological data, a sensitivity test with a different reanalysis, which is obviously at hand, could be performed.
Response: We agree. Taking the notion of model reproducibility over complex terrain, previous study by Rai et al. (2022) explicitly carried out comprehensive model validation using different meteorological reanalysis data sets over the same domain. In that study, one should note that, entire simulated period is categorized into monsoon and non-monsoon periods. Study found that temperature WRF-CRU (Rm = 0.82, Rnm = 0.96) and WRF-ERA5 (Rm = 0.90, Rnm = 0.93) reproduced better than precipitation WRF-TRMM (Rm = 0.80, Rnm = 0.63) and WRF-ERA5 (Rm = 0.30, Rnm = 0.40) respectively. Note: Rm = correlation coefficient in monsoon, Rnm = correlation coefficient in non-monsoon. Additionally, model was able to reproduce reasonably well with NCEP and ERA5 data. Here we have provided additional statistical metrics to showcase the meteorological data relation with model. During monsoon, model produced cold biased with ERA5 while there is the higher precipitation bias with ERA5 than TRMM. Detailed statistical score are presented in Table S4.
5) Figure 5 gives the impression that the WRF simulation does not perform well in reproducing the AOD. In winter and spring it underestimates the AOD and in summer and autumn it overestimates the AOD. In the text mainly the problems in simulating dust are discussed. However, given the large (a difference plot WRF-observations would be helpful) discrepancies between simulation and observations, I have little confidence in the further results derived from the WRF simulation.
Response: We appreciate the reviewer's concern and concur that our model simulation significantly understates the AOD, particularly in the winter and spring. However, we also conducted a comparison of model AOD data with AOD data from a number of additional data sources, including AERONET (limited but accessible locations), MERRA-2, and CAMS reanalysis datasets, in order to assess the confidence in the simulation results. The time series plot of AOD from the model (WRF-Chem) and observation (AERONET), as well as CAMS and MERRA-2, has these displayed in Figure 3. We employed the datasets further for our research with the help of the statistical analysis, which demonstrated good agreement between the dataset and also with the trend and distribution for the majority of the period.
6) I wonder why the seasonal pattern for PM2.5 and PM10 with maximum concentrations in winter and minimum concentrations in summer (Fig 4, first and second column) is the opposite to AOD with the maximum AOD in summer and the minimum AOD in winter (Fig5, top row).
Response: We thank reviewer for the comment. This discrepancy could arouse by several reasons. First, while AOD is the sum of the extinction coefficient times thickness that is integrated over the atmospheric vertical layer, PM represents surface concentration. Our study is set up by default with 40 vertical sigma levels from the level's surface to its peak. Because of the thin depth of the atmospheric layer and the poorer geographic resolution used in our analysis, we believe that the AOD was underestimated despite the high surface concentrations during winter and spring. Second, according to Pan et al. (2015), inadequate representation of anthropogenic and biofuel emission as well as relative humidity (RH) may all contribute to the underestimating of AOD during the winter. This part is overlooked in this study however we envisioned to carry out effect of such factors in AOD underestimation in near future. Surprisingly, higher AOD during summer was found. This is partly due to aerosol long-range transport through low-level jet and tropical easterly jet that persist over especially over South Asia (Ratnam et al., 2021). Another reason could be modulated by enhanced temperature and RH during summer monsoon that intensified hygroscopic growth of aerosols which consequently yielding high AOD. The detail is given in (l379-392).
7) It is still not clear to me how the trajectory analysis was performed. What was the purpose of doing forward and backward trajectories? Where were the forward and backward trajectories started?
Response: We have briefly stated about the trajectory analysis in section 2.3. The air trajectory essentially measures the dynamical processes occurring in the atmosphere. The trajectory can be forward, indicating the impending path taken by the particles, or backward, indicating the historical path the particles had traveled along their trajectory. The calculations of the air trajectories take wind and weather patterns seriously. The purpose of the trajectories analysis was to provide essential information especially aerosol transport mechanism and source-receptor relationship over the study region. We computed trajectories at Langtang station (28.21̊ N, 85.61̊ E; 4900 m a.s.l). In this study, we consider the end point of backward trajectory is the starting point of forward trajectory. 7 day trajectories at 6 hour interval were simulated in the study.
8) While the title suggests that the focus of this study is put on the Pan-Third-Pole this study mainly presents and discusses the high aerosol load in densely populated regions (Indo- Gangetic Plain and East China) known for strong aerosol sources. The possible impacts of aerosol on the Pan-Third-Pole are stated in the introduction, but throughout the manuscript the authors avoid presenting the current knowledge on the sources and transport pathways and comparing it with their findings. They rather name studies (e.g. ll460 - 464; ll467-469) and remain vague. In my opinion the authors should contrast the state-of-the-art knowledge with their findings to make the advancements in their study clear.
Response: Thank you for the suggestions. Most of the past study focuses either in South Asia or East Asia. In this study we presented the synoptic scale analysis on aerosol concentration, transport dynamics, and meteorological influence by leveraging satellite data, reanalysis data, and model over the interest of global important region (i.e. PTP). As the region is susceptible from air pollution and environment impact perspectives. Over this complex geographic region this study started off by model validation, presented spatio-temporal variation of aerosol and AOD, provided vertical profile of aerosol, and carried out the source-apportionment analysis. Exclusively, we extended atmospheric river concept in term of aerosol which is first of its kind if not otherwise stated. In this study we discussed about the inability of model in AOD simulation which further needs to be taken into consideration. Interesting finding from integrated aerosol transport calculation shows significant aerosol transport over South East Asia. To enhance the knowledge of such methods and process inclusion of more variables with a finer resolution of reanalysis products be warranted. Further to resolve the vertical structure of the transport process across complex terrain like the Himalayas and Tibetan Plateau, the finer resolution of model simulation is anticipated with an aerosol-climate feedback mechanism.
-
AC2: 'Reply on RC2', Mukesh Rai, 06 Oct 2022
Status: closed
-
RC1: 'Comment on acp-2022-199', Anonymous Referee #1, 05 Jun 2022
Review comments on ACP-2022-199
Insight into seasonal aerosol concentrations, meteorological influence, and transport over the Pan-Third Pole region using multi-sensors satellite and model simulation
by Rai et alUsing observation, satellite and model outputs, the authors investigated the Spatio-temporal aerosol concentration, the influence of meteorological conditions, and underlying aerosol transport mechanisms over the Pan-Third Pole region. They demonstrated the seasonal distribution and spatial variability of aerosol across this wider domain and tried to explore model performance. Even though they have mentioned and highlighted the whole PTP region apart from IGP and China, there is nothing demonstrated in this study over this wide domain. At this level, there isn't enough scientific proof or knowledge enhancement to properly connect the underlying facts as well. With the following factors in mind, I feel the manuscript will be enhanced if the authors make the following changes and, more importantly, if they can eliminate numerous unnecessary discussions.
Recommendation: Major revision
1) One would expect, from the title of the manuscript, that the authors had conducted research on the Tibetan Plateau and Himalayas region in addition to outer PTP region. However, they have only some information about India and China. So this is an unnecessarily misleading part of the title, abstract and even introduction of this study; this should be avoided.
2) I'm not sure how authors can make a very crucial assertion like AOD values are maximum across IGP when two models and two observations produce four findings. Despite the fact that WRF-Chem shows that northern China has higher AOD values than IGP, the other three choices show that the signal endures between seasons. As a result, such a remark in the abstract is not appreciable but also misleads readers.
3) In this study, the authors emphasise aerosol concentration over IGP. If a similar quantity of aerosol concentration is evident above China, however, it is rarely emphasised. This is not a scientifically sound practice.
4) A large drop in PM2.5 is seen in major Indian cities, according to Sing et al 2021 (10.1016/j.scitotenv.2020.141461). However, the authors make no mention of this in the manuscript. The recent downturn in China, on the other hand, is explicitly acknowledged.
5) I am not sure if there is anything unknown to unravel the effect of meteorological conditions on the spatiotemporal distribution of aerosol over the Himalayas. Many studies have already documented these factors in these regions. Furthermore, they stated in each part that the results are comparable to past studies. So, it's evident that they're simply re-inventing something that's been done before. Any such exaggeration should be avoided.
6) Again, I'm not sure why the authors place so much focus on the Third Pole location. Seasonal Spatio-temporal fluctuation of aerosols and total AOD is just discernible over IGP or China in the signals from Figures 4 and 5. As a result, neither the title nor the abstract should be expected to reveal this.
7) If the model has a dust aerosol bias, which is indicated as being resolved by masking, how can authors assure that this is not causing problems in surrounding regions through transport?
8) Due to the low-quality figure in figure 6, the circulation pattern is difficult to interpret.
9) If authors are perfectly aware that the bias comes from the chosen domain, why do they continue to use it? It appears that these are unsubstantiated claims with no scientific basis. The conclusion is littered with ambiguous assertions that should be avoided.
Citation: https://doi.org/10.5194/acp-2022-199-RC1 -
AC1: 'Reply on RC1', Mukesh Rai, 06 Oct 2022
1) One would expect, from the title of the manuscript, that the authors had conducted research on the Tibetan Plateau and Himalayas region in addition to the outer PTP region. However, they have only some information about India and China. So this is an unnecessarily misleading part of the title, abstract and even introduction of this study; this should be avoided.
Response: Thank you for the comments. We agreed and made necessary changes and add more information (l353-368).
2) I'm not sure how authors can make a very crucial assertion like AOD values are maximum across IGP when two models and two observations produce four findings. Despite the fact that WRF-Chem shows that northern China has higher AOD values than IGP, the other three choices show that the signal endures between seasons. As a result, such a remark in the abstract is not appreciable but also misleads readers.
Response: We agree. Taking this notion in mind that we have modified the misleading statements. Instead of just using IGP and East China, we have added other regions where AOD endures throughout the simulation period in (l42-43). We also added more information in the result section (l363-368).
3) In this study, the authors emphasize aerosol concentration over IGP. If a similar quantity of aerosol concentration is evident above China, however, it is rarely emphasized. This is not a scientifically sound practice.
Response: Thank you for highlighting this issue. We tried our best to avoid such discrepancies. Now we have modified this section (l353-362).
4) A large drop in PM2.5 is seen in major Indian cities, according to Sing et al 2021 (10.1016/j.scitotenv.2020.141461). However, the authors make no mention of this in the manuscript. The recent downturn in China, on the other hand, is explicitly acknowledged.
Response: Thank you for recommending this paper. We have included the finding from this paper in our manuscript (l112-114).
5) I am not sure if there is anything unknown to unravel the effect of meteorological conditions on the spatiotemporal distribution of aerosol over the Himalayas. Many studies have already documented these factors in these regions. Furthermore, they stated in each part that the results are comparable to past studies. So, it's evident that they're simply reinventing something that's been done before. Any such exaggeration should be avoided.
Response: Our study region comprises dynamic geophysical features including arid, semi-arid, and mountain to flat land. Previous studies either focus on the Southern flank of the Himalayas, the Tibetan Plateau, or South East Asia. Thus, we attempt to showcase how different meteorological parameters play a role in aerosol concentration distribution and transport mechanism in the synoptic scale which owns unique geophysical features.
6) Again, I'm not sure why the authors place so much focus on the Third Pole location. Seasonal Spatio-temporal fluctuation of aerosols and total AOD is just discernible over IGP or China in the signals from Figures 4 and 5. As a result, neither the title nor the abstract should be expected to reveal this.
Response: We thank the reviewer for pointing out the inconsistency in the result presentation of our manuscript. We sought irregularity in our manuscript and presented more results. We added more findings and discussed more in detail other than the Third Pole region (l351-357).
7) If the model has a dust aerosol bias, which is indicated as being resolved by masking, how can authors assure that this is not causing problems in surrounding regions through transport?
Response: Thank you for highlighting this issue. In line (l347), our choice of the word “mask out” was meant to be that our model did not mimic the general feature of AOD values over arid/semi-arid than the reanalysis dataset.
8) Due to the low-quality figure in figure 6, the circulation pattern is difficult to interpret.
Response: Thank you. Now, figure 6 quality is improved.
9) If authors are perfectly aware that the bias comes from the chosen domain, why do they continue to use it? It appears that these are unsubstantiated claims with no scientific basis. The conclusion is littered with ambiguous assertions that should be avoided.
Response: We appreciate the concern raised by the reviewer which is valid. The previous study is largely from South Asia and East Asia. Our motivation is to present the synoptic scale picture of aerosol distribution and transport dynamics over geographically complex regions. Yes, we are aware that bias is there but we intend to use modelling as a tool and showcase the general features of the aerosols and aid more information over the region. This study underscores the bias correction which was obviously at hand but due to computational cost that could not perform. Further studies are needed to improve the model performance, especially over complex topography.
-
AC1: 'Reply on RC1', Mukesh Rai, 06 Oct 2022
-
RC2: 'Comment on acp-2022-199', Anonymous Referee #2, 12 Jun 2022
The title, abstract, and introduction of the study ``Insight into seasonal aerosol concentrations, meteorological influence, and transport over the Pan-Third Pole region using multi-sensors satellite and model simulation'' by M. Rai et al. suggest an investigation of the seasonal properties and the origin of aerosol over the Pan-Third-Pole my means of satellite observations and model simulations. However, the study mainly presents seasonal means of PM2.5, PM10, and AOD at 10 AERONET stations and an unidentified number between 8 to 12 stations in China, a WRF simulation, CAMS data, MERRA-2 reanalysis, and VIIRS satellite measurements for a single year. While the motivation and envisaged scope of the manuscript would fit the scope of ACP the presented data and analyses do not. In my opinion this manuscript is just a collection of arbitrarily selected data. The data sets do not agree quantitatively, and in case of the WRF simulation not even qualitatively. Despite these discrepancies no attempt was made to assess the uncertainties of the simulation and measurement data to identify the source of the discrepancies.
Concerning the scientific quality I have numerous concerns:
In this study 3 different reanalyses were used: NCEP/FNL with WRF, MERRA-2, and GDAS with HYSPLIT. What was the motivation to do this? It is known that reanalyses differ. In my opinion using multiple reanalyses is a source for discrepancies. This should be taken into account when estimating the uncertainties and comparing the results/data sets.
The methods section (Section 2) is rather a description of the models used, but not of the data. Moreover, it is not clear how the models were set up and used and how the data was prepared. Concerning WRF, what is the purpose of the used models? Why have they been selected? Which parameters were used from the MERRA-2 data? What was the ``further treatment'' of the AERONET data? A description of the Chinese stations is completely missing. Figure 1 indicates 12 stations in China, but according to the manuscript only 8 stations were used (l 205). Which stations and why were four left out? I have the impression that the data and methods are not sufficiently described to reproduce the results.
Did the WRF simulation really produce negative PM2.5 concentrations? E.g. in Xi'an 66.5µg m-3 were measured, whereas the WRF simulation underestimated it by 93.6µg m-3 (ll 235-237). This should be clarified in the text.
Concerning the differences between the WRF simulation and the measurement the authors stated repeatedly some speculation (e.g. l243-246, 248-246). To investigate the effect of e.g. the meteorological data, a sensitivity test with a different reanalysis, which is obviously at hand, could be performed.
Figure 5 gives the impression that the WRF simulation does not perform well in reproducing the AOD. In winter and spring it underestimates the AOD and in summer and autumn it overestimates the AOD. In the text mainly the problems in simulating dust are discussed. However, given the large (a difference plot WRF-observations would be helpful) discrepancies between simulation and observations, I have little confidence in the further results derived from the WRF simulation.
I wonder why the seasonal pattern for PM2.5 and PM10 with maximum concentrations in winter and minimum concentrations in summer (Fig 4, first and second column) is the opposite to AOD with the maximum AOD in summer and the minimum AOD in winter (Fig5, top row).
It is still not clear to me how the trajectory analysis was performed. What was the purpose of doing forward and backward trajectories? Where were the forward and backward trajectories started?
While the title suggests that the focus of this study is put on the Pan-Third-Pole this study mainly presents and discusses the high aerosol load in densely populated regions (Indo-Gangetic Plain and East China) known for strong aerosol sources. The possible impacts of aerosol on the Pan-Third-Pole are stated in the introduction, but throughout the manuscript the authors avoid presenting the current knowledge on the sources and transport pathways and comparing it with their findings. They rather name studies (e.g. ll460 - 464; ll467-469) and remain vague. In my opinion the authors should contrast the state-of-the-art knowledge with their findings to make the advancements in their study clear.
Finally the presentation quality is poor. First of all most figures are too small. In most figures the axis labels, labels, and even data points are barely visible (even when zooming in to 200 %). The manuscript was difficult to read due the the numerous unnecessary abbreviations, some of them were not even introduced (e.g. AP, PG, CA, TD), others not used consistently (e.g. SA & South Asia), others introduced multiple times (e.g. VIIRS). It was also difficult to follow the line of argumentation while working through ``authors XYZ stated that'' at the beginning of consecutive sentences (e.g. ll438-444,). Moreover there are many avoidable mistakes (e.g. Fig. 1 ``dark blue shade'' is actually black, ll 209-212 the latitude and longitude of the stations is swapped). Finally the language certainly needs thorough revision.
In my opinion, fixing the above mentioned severe aspects is more than a major revision. Hence I suggest to reject this manuscript.
Citation: https://doi.org/10.5194/acp-2022-199-RC2 -
AC2: 'Reply on RC2', Mukesh Rai, 06 Oct 2022
1) In this study 3 different reanalyses were used: NCEP/FNL with WRF, MERRA-2, and GDAS
with HYSPLIT. What was the motivation to do this? It is known that reanalysis differs. In my opinion, using multiple reanalyses is a source for discrepancies. This should be taken into account when estimating the uncertainties and comparing the results/data sets.
Response: We appreciate the concern raised by reviewer. We acknowledge the fact of the discrepancy in our study. Carrying out model simulation in geographical complex region is challenging itself as our study area possess the diverse geographical features where observation data set are very limited. Thereby we wanted to leverage other available data source to assess, compare, and draw the conclusion in our study. The motivation of using reanalysis data set in our study are to provide meteorological initial boundary condition in model, compare with model and reanalysis data set, and to perform footprint analysis. To assess the model performance, we performed statistical metrics analysis of different pollutants and compared between model, observation, and reanalysis data (l 243-245, 278-288). The detail comparison of AOD observation versus model and reanalysis data sets are provided in Table S2.
2) The methods section (Section 2) is rather a description of the models used, but not of the data. Moreover, it is not clear how the models were set up and used and how the data was prepared. Concerning WRF, what is the purpose of the used models? Why have they been selected? Which parameters were used from the MERRA-2 data? What was the ``further treatment'' of the AERONET data? A description of the Chinese stations is completely missing. Figure 1 indicates 12 stations in China, but according to the manuscript only 8 stations were used (l 205). Which stations and why were four left out? I have the impression that the data and methods are not sufficiently described to reproduce the results.
Response: We have discussed the data used in section 2.4. Regarding the model set up and data preparation, we have discussed in 2.1 section. Further, key parameterization schemes adopted in our study are included in Table S1. The WRF-Chem model is selected because of its dynamic abilities such includes numerical more consistency, meteorology-chemistry feedback, capable of simulating and predicting pollutant concentration with physical and chemical options. As our model is forced to run in diverse geographical region (i.e. flat land, arid, semi-arid, and mountain region) which may require to have more represented parameterization schemes which should have more realistic presentation of different processes like advection, convection, diffusion, dry deposition, wet scavenging, aqueous and gas phase reaction, photolysis rates, other microphysics. We have used AOD and dust from MERRA-2. ‘Further treat’: we mean to say that we have used level 2 AERONET data however, due to data gap we further fill this void using Level 1.5 data. Now, the description of China station are included (l211-214).
3) Did the WRF simulation really produce negative PM2.5 concentrations? E.g. in Xi'an 66.5μg m-3 were measured, whereas the WRF simulation underestimated it by 93.6μg m-3 (ll 235-237). -This should be clarified in the text.
Response: Thank you for pointing out this issue. We have gone through the code and found out there was typo in the code. We have modified and presented the right values in (l243-246)
4) Concerning the differences between the WRF simulation and the measurement the authors stated repeatedly some speculation (e.g. l243-246, 248-246). To investigate the effect of e.g. the meteorological data, a sensitivity test with a different reanalysis, which is obviously at hand, could be performed.
Response: We agree. Taking the notion of model reproducibility over complex terrain, previous study by Rai et al. (2022) explicitly carried out comprehensive model validation using different meteorological reanalysis data sets over the same domain. In that study, one should note that, entire simulated period is categorized into monsoon and non-monsoon periods. Study found that temperature WRF-CRU (Rm = 0.82, Rnm = 0.96) and WRF-ERA5 (Rm = 0.90, Rnm = 0.93) reproduced better than precipitation WRF-TRMM (Rm = 0.80, Rnm = 0.63) and WRF-ERA5 (Rm = 0.30, Rnm = 0.40) respectively. Note: Rm = correlation coefficient in monsoon, Rnm = correlation coefficient in non-monsoon. Additionally, model was able to reproduce reasonably well with NCEP and ERA5 data. Here we have provided additional statistical metrics to showcase the meteorological data relation with model. During monsoon, model produced cold biased with ERA5 while there is the higher precipitation bias with ERA5 than TRMM. Detailed statistical score are presented in Table S4.
5) Figure 5 gives the impression that the WRF simulation does not perform well in reproducing the AOD. In winter and spring it underestimates the AOD and in summer and autumn it overestimates the AOD. In the text mainly the problems in simulating dust are discussed. However, given the large (a difference plot WRF-observations would be helpful) discrepancies between simulation and observations, I have little confidence in the further results derived from the WRF simulation.
Response: We appreciate the reviewer's concern and concur that our model simulation significantly understates the AOD, particularly in the winter and spring. However, we also conducted a comparison of model AOD data with AOD data from a number of additional data sources, including AERONET (limited but accessible locations), MERRA-2, and CAMS reanalysis datasets, in order to assess the confidence in the simulation results. The time series plot of AOD from the model (WRF-Chem) and observation (AERONET), as well as CAMS and MERRA-2, has these displayed in Figure 3. We employed the datasets further for our research with the help of the statistical analysis, which demonstrated good agreement between the dataset and also with the trend and distribution for the majority of the period.
6) I wonder why the seasonal pattern for PM2.5 and PM10 with maximum concentrations in winter and minimum concentrations in summer (Fig 4, first and second column) is the opposite to AOD with the maximum AOD in summer and the minimum AOD in winter (Fig5, top row).
Response: We thank reviewer for the comment. This discrepancy could arouse by several reasons. First, while AOD is the sum of the extinction coefficient times thickness that is integrated over the atmospheric vertical layer, PM represents surface concentration. Our study is set up by default with 40 vertical sigma levels from the level's surface to its peak. Because of the thin depth of the atmospheric layer and the poorer geographic resolution used in our analysis, we believe that the AOD was underestimated despite the high surface concentrations during winter and spring. Second, according to Pan et al. (2015), inadequate representation of anthropogenic and biofuel emission as well as relative humidity (RH) may all contribute to the underestimating of AOD during the winter. This part is overlooked in this study however we envisioned to carry out effect of such factors in AOD underestimation in near future. Surprisingly, higher AOD during summer was found. This is partly due to aerosol long-range transport through low-level jet and tropical easterly jet that persist over especially over South Asia (Ratnam et al., 2021). Another reason could be modulated by enhanced temperature and RH during summer monsoon that intensified hygroscopic growth of aerosols which consequently yielding high AOD. The detail is given in (l379-392).
7) It is still not clear to me how the trajectory analysis was performed. What was the purpose of doing forward and backward trajectories? Where were the forward and backward trajectories started?
Response: We have briefly stated about the trajectory analysis in section 2.3. The air trajectory essentially measures the dynamical processes occurring in the atmosphere. The trajectory can be forward, indicating the impending path taken by the particles, or backward, indicating the historical path the particles had traveled along their trajectory. The calculations of the air trajectories take wind and weather patterns seriously. The purpose of the trajectories analysis was to provide essential information especially aerosol transport mechanism and source-receptor relationship over the study region. We computed trajectories at Langtang station (28.21̊ N, 85.61̊ E; 4900 m a.s.l). In this study, we consider the end point of backward trajectory is the starting point of forward trajectory. 7 day trajectories at 6 hour interval were simulated in the study.
8) While the title suggests that the focus of this study is put on the Pan-Third-Pole this study mainly presents and discusses the high aerosol load in densely populated regions (Indo- Gangetic Plain and East China) known for strong aerosol sources. The possible impacts of aerosol on the Pan-Third-Pole are stated in the introduction, but throughout the manuscript the authors avoid presenting the current knowledge on the sources and transport pathways and comparing it with their findings. They rather name studies (e.g. ll460 - 464; ll467-469) and remain vague. In my opinion the authors should contrast the state-of-the-art knowledge with their findings to make the advancements in their study clear.
Response: Thank you for the suggestions. Most of the past study focuses either in South Asia or East Asia. In this study we presented the synoptic scale analysis on aerosol concentration, transport dynamics, and meteorological influence by leveraging satellite data, reanalysis data, and model over the interest of global important region (i.e. PTP). As the region is susceptible from air pollution and environment impact perspectives. Over this complex geographic region this study started off by model validation, presented spatio-temporal variation of aerosol and AOD, provided vertical profile of aerosol, and carried out the source-apportionment analysis. Exclusively, we extended atmospheric river concept in term of aerosol which is first of its kind if not otherwise stated. In this study we discussed about the inability of model in AOD simulation which further needs to be taken into consideration. Interesting finding from integrated aerosol transport calculation shows significant aerosol transport over South East Asia. To enhance the knowledge of such methods and process inclusion of more variables with a finer resolution of reanalysis products be warranted. Further to resolve the vertical structure of the transport process across complex terrain like the Himalayas and Tibetan Plateau, the finer resolution of model simulation is anticipated with an aerosol-climate feedback mechanism.
-
AC2: 'Reply on RC2', Mukesh Rai, 06 Oct 2022
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,071 | 280 | 44 | 1,395 | 81 | 37 | 45 |
- HTML: 1,071
- PDF: 280
- XML: 44
- Total: 1,395
- Supplement: 81
- BibTeX: 37
- EndNote: 45
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1