the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 11 Aug 2023
Assessing the Assimilation of Himawari-8 observations on Aerosol Forecasts and Radiative Effects During Pollution Transport from South Asia to the Tibetan Plateau
Min Zhao
Daisuke Goto
Guangyu Shi
Abstract. The emissions from South Asia (SA) represent a critical source of aerosols on the Tibetan Plateau (TP), and aerosols can significantly reduce the surface solar energy. To enhance the precision of aerosol forecasting and its radiative effects in SA and TP, we employed a four-dimensional local ensemble transform Kalman filter (4D-LETKF) aerosol data assimilation (DA) system. This system was utilized to assimilate Himawari-8 aerosol optical thickness (AOT) into the Weather Research and Forecasting-Chemistry (WRF-Chem) model to depict one SA air pollution outbreak event in spring 2018. Sensitivity tests for the assimilation system have been conducted firstly to tune temporal localization lengths. Comparisons with independent Moderate Resolution Imaging Spectroradiometer (MODIS) and AErosol RObotic NETwork (AERONET) observations demonstrate that the AOT analysis and forecast fields have more reasonable diurnal variations by assimilating all the observations within 12 h window, which are both better than assimilating the hourly observations in the current assimilation timeslot. Assimilation of the entire window of observations with aerosol radiative effect activation significantly improves the prediction of downward solar radiation compared to the free-run experiment. The assimilation of aerosol radiative effect activation led to a reduction in aerosol concentrations over SA, resulting in increased surface radiation, temperature, boundary layer height, and atmospheric instability. These changes facilitated air uplift, promoting aerosol transport from SA to the southeastern TP and leading to an increase in AOT in this region.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(4140 KB)
-
Supplement
(94 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(4140 KB) - Metadata XML
-
Supplement
(94 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Min Zhao et al.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1581', Anonymous Referee #1, 22 Aug 2023
According to literatures, the emissions from South Asia can be transported to the Tibetan Plateau (TP) as critical source regions of aerosols to TP. In this study, an assimilation system with the Himawari-8 aerosol optical thickness (AOT) into the Weather Research and Forecasting-Chemistry (WRF-Chem) model was applied for the deep study of aerosol forecasting and its radiative effects in SA and TP. In general, the study was well organized, and the findings should have some scientific contributions to related studies in future. However, the structure and English writing should be comprehensively polished for easily reading and understanding before it can be accepted. Furthermore, there are some suggestions for the manuscript as follows:
- If the words only appeared one time, the abbreviation is not necessary, such as cloud condensation nuclei (CCN), checking for other abbreviations.
- Lines 35-37, references needed
- Lines 74-80 should be moved to the end of the above paragraph
- Lines 111-116 indicated the scientific contribution of this study, it is better to move them to the end of the last paragraph.
- Line 136, the emission inventory is the necessary and important input data for WRF model. According to the introduction, the inventory of 2010 was used, however, the modelled air pollution event was happened in 2018. There should be some uncertainties. Did the inventory be updated? Some explanation needed here.
- There are too many abbreviations in the manuscript. A list of abbreviations should be added in order to make it more clearly reading and understanding.
- Line 150, what’s the meaning for “J”
- In Section 2, did the authors made the sensitivity and uncertainty analysis? The methods should be mentioned.
- Line 269, some detailed information (for example equations) should be added for the statistical criteria, maybe in SI
- Line 289, what’s the reasons for the underestimation
- Lines 312-313, some references support needed
- Line 378, underestimation or overestimation?
- Lines 426-428, any supporting references?
Citation: https://doi.org/10.5194/egusphere-2023-1581-RC1 -
AC1: 'Reply on RC1', Tie Dai, 16 Oct 2023
Response to the Reviewer #1
General comments:
According to literatures, the emissions from South Asia can be transported to the Tibetan Plateau (TP) as critical source regions of aerosols to TP. In this study, an assimilation system with the Himawari-8 aerosol optical thickness (AOT) into the Weather Research and Forecasting-Chemistry (WRF-Chem) model was applied for the deep study of aerosol forecasting and its radiative effects in SA and TP. In general, the study was well organized, and the findings should have some scientific contributions to related studies in future. However, the structure and English writing should be comprehensively polished for easily reading and understanding before it can be accepted. Furthermore, there are some suggestions for the manuscript as follows:
Response: We really appreciate your great efforts and positive evaluation of the manuscript. In response to your question about the structure and English writing, it has been polished by professional English speaking editors, as evidenced by the following certificate.
Some concerns:
- If the words only appeared one time, the abbreviation is not necessary, such as cloud condensation nuclei (CCN), checking for other abbreviations.
Response: Thank you for your comment. We have deleted the unnecessary abbreviations, such as CCN, GSI, ICs, ERA5.
- Lines 35-37, references needed
Response: Done. Thank you for your comment. We have added the reference (i.e., Li et al., 2020) on aerosol transport to the Tibetan Plateau in lines 35-37.
Li, F., Wan, X., Wang, H., Orsolini, Y. J., Cong, Z., Gao, Y., and Kang, S.: Arctic sea-ice loss intensifies aerosol transport to the Tibetan Plateau, Nature Climate Change, 10, 1037-1044, https://doi.org/10.1038/s41558-020-0881-2, 2020.
- Lines 74-80 should be moved to the end of the above paragraph
Response: Done. Thank you for your comment.
- Lines 111-116 indicated the scientific contribution of this study, it is better to move them to the end of the last paragraph.
Response: Done. Thank you.
- Line 136, the emission inventory is the necessary and important input data for WRF model. According to the introduction, the inventory of 2010 was used, however, the modelled air pollution event was happened in 2018. There should be some uncertainties. Did the inventory be updated? Some explanation needed here.
Response: Thank you for your comment. We mainly consider the impact of South Asia on the Tibetan Plateau, relatively new emission inventories such as the MEIC inventory which has 2018 but only includes the Chinese regions. The MIX Asian inventory, which includes the South Asian regions, has only recently been updated to 2010, so the 2010 MIX Asian inventory is used and there is no current 2018 MIX Asian inventory.
Changes in Manuscript:
Line 136-137:
For anthropogenic emissions, we use the MIX Asian inventory for March 2010 due to the need to include emission sources from South Asia.
- There are too many abbreviations in the manuscript. A list of abbreviations should be added in order to make it more clearly reading and understanding.
Response: Thank you for your comment. We have added a list of abbreviations in Appendix A.
Changes in Manuscript:
Appendix A: The summary of the abbreviations and their corresponding full names in this study.
Abbreviation
Full name
Abbreviation
Full name
SA
South Asia
GOCART
the Goddard Chemical Aerosol Radiation Transport
TP
Tibetan Plateau
NCEP
the National Centers for Environmental Prediction
DA
data assimilation
AHI
Advanced Himawari Imager
AOT
aerosol optical thickness
AE
Ångström exponent
WRF-Chem
Weather Research and Forecasting-Chemistry
JAXA
the Japan Aerospace Exploration Agency
MODIS
Moderate Resolution Imaging Spectroradiometer
DT
the Dark Target
AERONET
AErosol RObotic NETwork
DB
the Deep Blue
BC
black carbon
SYN
synoptic
GOES‐8
Goddard Earth Observing System‐8
BIAS
mean bias
GOCI
Geostationary Ocean Color Imager
RMSE
root-mean-square error
CERES
Clouds and the Earth’s Radiant Energy System
CORR
correlation coefficient
LETKF
local ensemble transform Kalman filter
PDFs
the probability distribution functions
MOSAIC
Model for Simulating Aerosol Interactions and Chemistry
DSRc
downward solar radiation under clear-sky
RRTMG
Rapid Radiative Transfer Model
DSR
downward radiation flux at the surface at the all-sky
OC
Organic Carbon
PBLH
planetary boundary layer height
PM2.5
Particulate matter with an aerodynamic equivalent diameter of less than or equal to 2.5 micrometers in ambient air
T2
2-m temperature
PM10
Particulate matter with an aerodynamic equivalent diameter of less than or equal to 10 micrometers in ambient air
RH2
2-m relative humidity
NMVOC
non-methane volatile organic compounds
q
water vapor mixing ratio
FINN
the Fire INventory from NCAR
T
temperature
MEGAN
the Model of Emissions of Gasses and Aerosols from Nature
- Line 150, what’s the meaning for “J”
Response: Sorry for the unclear "J". The “J” represents the value of the cost function. Data assimilation is essentially the problem of solving “J”. Since the “J” does not appear directly in the subsequent formulas, so we have removed the “J” to avoid confusing the reader.
Changes in Manuscript:
Line 151:
Data assimilation is essentially for solving the minimum value of the cost function.
- In Section 2, did the authors made the sensitivity and uncertainty analysis? The methods should be mentioned.
Response: Thank you for your comment and question. In fact, in our previous study (Dai et al., 2021), the 4D-LETKF sensitivity experiments were conducted on different ensemble sizes (10 members, 20 members, and 40 members) for the assimilation system. The results reveal that the ensemble size has little effect on the assimilation performance. In this study, the effect of these assimilation tuning parameters is not a focus. We use 20 ensemble members and a constant multiplicative covariance inflation factor of 1.1 in all our assimilation experiments.
Dai, T., Cheng, Y., Goto, D., Li, Y., Tang, X., Shi, G., and Nakajima, T.: Revealing the sulfur dioxide emission reductions in China by assimilating surface observations in WRF-Chem, Atmospheric Chemistry and Physics, 21, 4357-4379, https://doi.org/10.5194/acp-21-4357-2021, 2021
Changes in Manuscript:
Line 201-205:
Sensitivity tests on this 4D-LETKF assimilation system with varying parameters (specifically, the ensemble members: 10, 20, and 40) have been conducted in our previous studies (Dai et al., 2021). Since the finding from ensemble size has only little effect on the assimilation performance (Dai et al., 2021), the impact of ensemble size on the data assimilation performance is not a focus of this study. We use twenty ensemble members and a constant multiplicative covariance inflation factor of 1.1 in all our following assimilation experiments.
- Line 269, some detailed information (for example equations) should be added for the statistical criteria, maybe in SI
Response: Thank you for your comment. We have added the equations about the statistical criteria in Appendix B.
Changes in Manuscript:
Appendix B: Statistical criteria.
Three primary statistical metrics were employed to assess the simulation performance of observed data on model aerosol fields. These metrics include the Bias (BIAS), Root Mean Square Error (RMSE), and Correlation Coefficient (CORR). BIAS is defined as the average difference between simulated results and observed values. RMSE quantifies the standard deviation of the differences between simulated results and observed values. CORR measures the correlation between simulated results and observed values. The formulas for these three statistical metrics are as follows:
(B1)
(B2)
(B3)
Here, represents the total number of data points, and are the simulated and observed values, respectively. The symbols and represent the means of simulated and observed values, respectively.
- Line 289, what’s the reasons for the underestimation
Response: Thank you for your question. The underestimations do not improve well relative to the overestimations. Small values of underestimation are always associated with small values of model error. The small value of the model error leads to a larger weighting of the simulated values, so that weak underestimations are less susceptible to assimilation effects.
Changes in Manuscript:
Line 292-294:
This may because smaller underestimations are always associated with smaller model error values, resulting in larger weights of simulated values and reduced sensitivity to assimilation effects.
- Lines 312-313, some references support needed
Response: Done. Thank you for your comment.
Changes in Manuscript:
Line 316-320:
In contrast to self-checking, the AOT analyses in DA_REON_12H are closer to MODIS than those in DA_REON_01H. The AOT analyses are interpolated to the MODIS, which is different in temporal and spatial distribution from Himawari-8. The AOT analyses in DA_REON_12H absorb the entire window of observations, and this asynchronous assimilation corrects the AOT analysis field at all times in each window (Dai et al., 2019).
- Line 378, underestimation or overestimation?
Response: Sorry for the ambiguity of this sentence, we've corrected it..
Changes in Manuscript:
Line 386-388:
Compared to CERES observations, the underestimation occurs in eastern SA, and the overestimation occurs in western SA in FR_REON with a total domain mean bias of 10.69 W m-2, which is due to the uncertainty in the aerosols and other factors such as the complex terrain processing in the model.
- Lines 426-428, any supporting references?
Response: Thank you for your comment. We have added some supporting references.
Changes in Manuscript:
Zhang et al. (2020) and Zheng et al. (2017) pointed out that pollutants from the South Asian can be transported to the Tibetan Plateau, and the transport is much stronger and deeper along the valley.
Zhang, M., Zhao, C., Cong, Z., Du, Q., Xu, M., Chen, Y., Chen, M., Li, R., Fu, Y., Zhong, L., Kang, S., Zhao, D., and Yang, Y.: Impact of topography on black carbon transport to the southern Tibetan Plateau during the pre-monsoon season and its climatic implication, Atmospheric Chemistry and Physics, 20, 5923-5943, https://doi.org/10.5194/acp-20-5923-2020, 2020.
Zheng, J., Hu, M., Du, Z. F., Shang, D. J., Gong, Z. H., Qin, Y. H., Fang, J. Y., Gu, F. T., Li, M. R., Peng, J. F., Li, J., Zhang, Y. Q., Huang, X. F., He, L. Y., Wu, Y. S., and Guo, S.: Influence of biomass burning from South Asia at a high-altitude mountain receptor site in China, Atmospheric Chemistry and Physics, 17, 6853-6864, https://doi.org/10.5194/acp-17-6853-2017, 2017.
-
RC2: 'Comment on egusphere-2023-1581', Anonymous Referee #2, 28 Aug 2023
General Comments:
The aerosols transported from South Asia (SA) are the important sources to the Tibetan Plateau (TP). Given the complex topography and limited observations in these areas, the aerosol simulations are still prone to large uncertainties. This paper shows the effects of the assimilation of Geostationary satellite Himawari-8 observations on aerosol forecasts and direct radiative effects (DRE) during the period of pollutants transporting from SA to the TP. The results demonstrate the AOT analysis and forecast fields can both effectively reproduce diurnal variations by assimilating all the observations within 12h window. Compared to the free-run experiment, assimilation experiments with aerosol radiative effect activation significantly improve the prediction of downward solar radiation and thereby enhance the transport of pollutants to the TP.
In general, the paper is well-written, and the results are nicely presented and well discussed with novel scientific founds. The authors did a credible job in analyzing how different experiment settings. The findings help better understand the transport of pollutants from South Asia to TP. Therefore, I would like to recommend accepting it to ACP after addressing the following minor comments
- Compared to DA_REON_12H experiment, the DA_REON_01H experiment performs better results in self-check due to assimilation of the entire window of observation data, but why a little worse in independent verification? Could you explain it in detail?
Specific comments:
- L83 Please rewrite this sentence
- L112-113 Please rewrite the sentence “The observations assimilated are Himawari-8 AOT and the main assimilated region is SA”.
- L126 It's better to write “Carbon Bond Mechanism Z”
- L126 Do you mean NMVOCs (non-methane volatile organic compounds)?
- L247 Replace " Fig. 1b " with " Fig. 1a "
- L273, L279 and L402, L403 Some figure number need to be consistent, e.g., “Figure 2”vs “Fig. 2”, “Figures 9a”vs “Figs. 9b”
- L281 “high surface albedo” maybe more precise
- L282 Please revise the sentence to make it clearer.
- L312 Replace " to absorb" with " aims to incorporate " to make it more clarity and fluidity
- L319 Which statistical indicators?
- L378 change the “Compare” to “Compared”
- L479 “surface temperature” maybe more precise
Citation: https://doi.org/10.5194/egusphere-2023-1581-RC2 -
AC2: 'Reply on RC2', Tie Dai, 16 Oct 2023
Response to the Reviewer #2
General comments:
The aerosols transported from South Asia (SA) are the important sources to the Tibetan Plateau (TP). Given the complex topography and limited observations in these areas, the aerosol simulations are still prone to large uncertainties. This paper shows the effects of the assimilation of Geostationary satellite Himawari-8 observations on aerosol forecasts and direct radiative effects (DRE) during the period of pollutants transporting from SA to the TP. The results demonstrate the AOT analysis and forecast fields can both effectively reproduce diurnal variations by assimilating all the observations within 12h window. Compared to the free-run experiment, assimilation experiments with aerosol radiative effect activation significantly improve the prediction of downward solar radiation and thereby enhance the transport of pollutants to the TP.
In general, the paper is well-written, and the results are nicely presented and well discussed with novel scientific founds. The authors did a credible job in analyzing how different experiment settings. The findings help better understand the transport of pollutants from South Asia to TP. Therefore, I would like to recommend accepting it to ACP after addressing the following minor comments
Response: We thank the reviewer for the encouraging comments. We have revised the manuscript following your comments.
Mainly concerns:
- Compared to DA_REON_12H experiment, the DA_REON_01H experiment performs better results in self-check due to assimilation of the entire window of observation data, but why a little worse in independent verification? Could you explain it in detail?
Response: Thanks. Considering the analysis fields, it is true that the validation results of DA_REON_01H perform better at the time of the self-test. This is due to the fact that the aerosol field at the hour of assimilation for this set of tests is then examined against itself, which is a much better time match. In the independent test, the analysis field is interpolated to the site of the independent test observations (MODIS or AERONET), which is different in time and space from the assimilated Himawari-8. At this point the asynchronous assimilation, absorbing the entire time window of DA_REON_12H would have performed relatively better. This is because the aerosol field before and after the assimilation time was also assimilated in this experiment.
Changes in Manuscript:
Line 317-320:
The AOT analyses are interpolated to the MODIS, which is different in temporal and spatial distribution from Himawari-8. The AOT analyses in DA_REON_12H absorb the entire window of observations, and this asynchronous assimilation corrects the AOT analysis field at all times in each window (Dai et al., 2019).
Specific comments:
- L83 Please rewrite this sentence
Response: Done. Thank you for your comment.
Changes in Manuscript:
Line 81:
Aerosols are second only to clouds in regional surface solar energy simulations.
- L112-113 Please rewrite the sentence “The observations assimilated are Himawari-8 AOT and the main assimilated region is SA”.
Response: Done. Thank you.
Changes in Manuscript:
Line 116-117:
The assimilated observations are Himawari-8 AOT, and the main assimilated region is SA.
- L126 It's better to write “Carbon Bond Mechanism Z”
Response: Thank you, we've corrected it.
- L126 Do you mean NMVOCs (non-methane volatile organic compounds)?
Response: Thank you for your comment, we've corrected it.
- L247 Replace " Fig. 1b " with " Fig. 1a "
Response: Done. Thank you.
- L273, L279 and L402, L403 Some figure number need to be consistent, e.g., “Figure 2”vs “Fig. 2”, “Figures 9a”vs “Figs. 9b”
Response: Done. Thank you for your comment, uniformly use Fig. or Figs.
- L281 “high surface albedo” maybe more precise
Response: Done. Thank you for your comment.
- L282 Please revise the sentence to make it clearer.
Response: Done. Thank you.
Changes in Manuscript:
Line 285-286:
The assimilated experiments perform very well regarding the AOT analyzed fields, which are more consistent with the Himawari-8 retrievals.
- L312 Replace " to absorb" with " aims to incorporate " to make it more clarity and fluidity
Response: Done. Thank you for your comment.
- L319 Which statistical indicators?
Response: The statistical indicators mainly refer to the bias.
- L378 change the “Compare” to “Compared”
Response: Thank you for your comment, we've corrected it.
- L479 “surface temperature” maybe more precise
Response: Thank you, we've corrected it in Line 489.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1581', Anonymous Referee #1, 22 Aug 2023
According to literatures, the emissions from South Asia can be transported to the Tibetan Plateau (TP) as critical source regions of aerosols to TP. In this study, an assimilation system with the Himawari-8 aerosol optical thickness (AOT) into the Weather Research and Forecasting-Chemistry (WRF-Chem) model was applied for the deep study of aerosol forecasting and its radiative effects in SA and TP. In general, the study was well organized, and the findings should have some scientific contributions to related studies in future. However, the structure and English writing should be comprehensively polished for easily reading and understanding before it can be accepted. Furthermore, there are some suggestions for the manuscript as follows:
- If the words only appeared one time, the abbreviation is not necessary, such as cloud condensation nuclei (CCN), checking for other abbreviations.
- Lines 35-37, references needed
- Lines 74-80 should be moved to the end of the above paragraph
- Lines 111-116 indicated the scientific contribution of this study, it is better to move them to the end of the last paragraph.
- Line 136, the emission inventory is the necessary and important input data for WRF model. According to the introduction, the inventory of 2010 was used, however, the modelled air pollution event was happened in 2018. There should be some uncertainties. Did the inventory be updated? Some explanation needed here.
- There are too many abbreviations in the manuscript. A list of abbreviations should be added in order to make it more clearly reading and understanding.
- Line 150, what’s the meaning for “J”
- In Section 2, did the authors made the sensitivity and uncertainty analysis? The methods should be mentioned.
- Line 269, some detailed information (for example equations) should be added for the statistical criteria, maybe in SI
- Line 289, what’s the reasons for the underestimation
- Lines 312-313, some references support needed
- Line 378, underestimation or overestimation?
- Lines 426-428, any supporting references?
Citation: https://doi.org/10.5194/egusphere-2023-1581-RC1 -
AC1: 'Reply on RC1', Tie Dai, 16 Oct 2023
Response to the Reviewer #1
General comments:
According to literatures, the emissions from South Asia can be transported to the Tibetan Plateau (TP) as critical source regions of aerosols to TP. In this study, an assimilation system with the Himawari-8 aerosol optical thickness (AOT) into the Weather Research and Forecasting-Chemistry (WRF-Chem) model was applied for the deep study of aerosol forecasting and its radiative effects in SA and TP. In general, the study was well organized, and the findings should have some scientific contributions to related studies in future. However, the structure and English writing should be comprehensively polished for easily reading and understanding before it can be accepted. Furthermore, there are some suggestions for the manuscript as follows:
Response: We really appreciate your great efforts and positive evaluation of the manuscript. In response to your question about the structure and English writing, it has been polished by professional English speaking editors, as evidenced by the following certificate.
Some concerns:
- If the words only appeared one time, the abbreviation is not necessary, such as cloud condensation nuclei (CCN), checking for other abbreviations.
Response: Thank you for your comment. We have deleted the unnecessary abbreviations, such as CCN, GSI, ICs, ERA5.
- Lines 35-37, references needed
Response: Done. Thank you for your comment. We have added the reference (i.e., Li et al., 2020) on aerosol transport to the Tibetan Plateau in lines 35-37.
Li, F., Wan, X., Wang, H., Orsolini, Y. J., Cong, Z., Gao, Y., and Kang, S.: Arctic sea-ice loss intensifies aerosol transport to the Tibetan Plateau, Nature Climate Change, 10, 1037-1044, https://doi.org/10.1038/s41558-020-0881-2, 2020.
- Lines 74-80 should be moved to the end of the above paragraph
Response: Done. Thank you for your comment.
- Lines 111-116 indicated the scientific contribution of this study, it is better to move them to the end of the last paragraph.
Response: Done. Thank you.
- Line 136, the emission inventory is the necessary and important input data for WRF model. According to the introduction, the inventory of 2010 was used, however, the modelled air pollution event was happened in 2018. There should be some uncertainties. Did the inventory be updated? Some explanation needed here.
Response: Thank you for your comment. We mainly consider the impact of South Asia on the Tibetan Plateau, relatively new emission inventories such as the MEIC inventory which has 2018 but only includes the Chinese regions. The MIX Asian inventory, which includes the South Asian regions, has only recently been updated to 2010, so the 2010 MIX Asian inventory is used and there is no current 2018 MIX Asian inventory.
Changes in Manuscript:
Line 136-137:
For anthropogenic emissions, we use the MIX Asian inventory for March 2010 due to the need to include emission sources from South Asia.
- There are too many abbreviations in the manuscript. A list of abbreviations should be added in order to make it more clearly reading and understanding.
Response: Thank you for your comment. We have added a list of abbreviations in Appendix A.
Changes in Manuscript:
Appendix A: The summary of the abbreviations and their corresponding full names in this study.
Abbreviation
Full name
Abbreviation
Full name
SA
South Asia
GOCART
the Goddard Chemical Aerosol Radiation Transport
TP
Tibetan Plateau
NCEP
the National Centers for Environmental Prediction
DA
data assimilation
AHI
Advanced Himawari Imager
AOT
aerosol optical thickness
AE
Ångström exponent
WRF-Chem
Weather Research and Forecasting-Chemistry
JAXA
the Japan Aerospace Exploration Agency
MODIS
Moderate Resolution Imaging Spectroradiometer
DT
the Dark Target
AERONET
AErosol RObotic NETwork
DB
the Deep Blue
BC
black carbon
SYN
synoptic
GOES‐8
Goddard Earth Observing System‐8
BIAS
mean bias
GOCI
Geostationary Ocean Color Imager
RMSE
root-mean-square error
CERES
Clouds and the Earth’s Radiant Energy System
CORR
correlation coefficient
LETKF
local ensemble transform Kalman filter
PDFs
the probability distribution functions
MOSAIC
Model for Simulating Aerosol Interactions and Chemistry
DSRc
downward solar radiation under clear-sky
RRTMG
Rapid Radiative Transfer Model
DSR
downward radiation flux at the surface at the all-sky
OC
Organic Carbon
PBLH
planetary boundary layer height
PM2.5
Particulate matter with an aerodynamic equivalent diameter of less than or equal to 2.5 micrometers in ambient air
T2
2-m temperature
PM10
Particulate matter with an aerodynamic equivalent diameter of less than or equal to 10 micrometers in ambient air
RH2
2-m relative humidity
NMVOC
non-methane volatile organic compounds
q
water vapor mixing ratio
FINN
the Fire INventory from NCAR
T
temperature
MEGAN
the Model of Emissions of Gasses and Aerosols from Nature
- Line 150, what’s the meaning for “J”
Response: Sorry for the unclear "J". The “J” represents the value of the cost function. Data assimilation is essentially the problem of solving “J”. Since the “J” does not appear directly in the subsequent formulas, so we have removed the “J” to avoid confusing the reader.
Changes in Manuscript:
Line 151:
Data assimilation is essentially for solving the minimum value of the cost function.
- In Section 2, did the authors made the sensitivity and uncertainty analysis? The methods should be mentioned.
Response: Thank you for your comment and question. In fact, in our previous study (Dai et al., 2021), the 4D-LETKF sensitivity experiments were conducted on different ensemble sizes (10 members, 20 members, and 40 members) for the assimilation system. The results reveal that the ensemble size has little effect on the assimilation performance. In this study, the effect of these assimilation tuning parameters is not a focus. We use 20 ensemble members and a constant multiplicative covariance inflation factor of 1.1 in all our assimilation experiments.
Dai, T., Cheng, Y., Goto, D., Li, Y., Tang, X., Shi, G., and Nakajima, T.: Revealing the sulfur dioxide emission reductions in China by assimilating surface observations in WRF-Chem, Atmospheric Chemistry and Physics, 21, 4357-4379, https://doi.org/10.5194/acp-21-4357-2021, 2021
Changes in Manuscript:
Line 201-205:
Sensitivity tests on this 4D-LETKF assimilation system with varying parameters (specifically, the ensemble members: 10, 20, and 40) have been conducted in our previous studies (Dai et al., 2021). Since the finding from ensemble size has only little effect on the assimilation performance (Dai et al., 2021), the impact of ensemble size on the data assimilation performance is not a focus of this study. We use twenty ensemble members and a constant multiplicative covariance inflation factor of 1.1 in all our following assimilation experiments.
- Line 269, some detailed information (for example equations) should be added for the statistical criteria, maybe in SI
Response: Thank you for your comment. We have added the equations about the statistical criteria in Appendix B.
Changes in Manuscript:
Appendix B: Statistical criteria.
Three primary statistical metrics were employed to assess the simulation performance of observed data on model aerosol fields. These metrics include the Bias (BIAS), Root Mean Square Error (RMSE), and Correlation Coefficient (CORR). BIAS is defined as the average difference between simulated results and observed values. RMSE quantifies the standard deviation of the differences between simulated results and observed values. CORR measures the correlation between simulated results and observed values. The formulas for these three statistical metrics are as follows:
(B1)
(B2)
(B3)
Here, represents the total number of data points, and are the simulated and observed values, respectively. The symbols and represent the means of simulated and observed values, respectively.
- Line 289, what’s the reasons for the underestimation
Response: Thank you for your question. The underestimations do not improve well relative to the overestimations. Small values of underestimation are always associated with small values of model error. The small value of the model error leads to a larger weighting of the simulated values, so that weak underestimations are less susceptible to assimilation effects.
Changes in Manuscript:
Line 292-294:
This may because smaller underestimations are always associated with smaller model error values, resulting in larger weights of simulated values and reduced sensitivity to assimilation effects.
- Lines 312-313, some references support needed
Response: Done. Thank you for your comment.
Changes in Manuscript:
Line 316-320:
In contrast to self-checking, the AOT analyses in DA_REON_12H are closer to MODIS than those in DA_REON_01H. The AOT analyses are interpolated to the MODIS, which is different in temporal and spatial distribution from Himawari-8. The AOT analyses in DA_REON_12H absorb the entire window of observations, and this asynchronous assimilation corrects the AOT analysis field at all times in each window (Dai et al., 2019).
- Line 378, underestimation or overestimation?
Response: Sorry for the ambiguity of this sentence, we've corrected it..
Changes in Manuscript:
Line 386-388:
Compared to CERES observations, the underestimation occurs in eastern SA, and the overestimation occurs in western SA in FR_REON with a total domain mean bias of 10.69 W m-2, which is due to the uncertainty in the aerosols and other factors such as the complex terrain processing in the model.
- Lines 426-428, any supporting references?
Response: Thank you for your comment. We have added some supporting references.
Changes in Manuscript:
Zhang et al. (2020) and Zheng et al. (2017) pointed out that pollutants from the South Asian can be transported to the Tibetan Plateau, and the transport is much stronger and deeper along the valley.
Zhang, M., Zhao, C., Cong, Z., Du, Q., Xu, M., Chen, Y., Chen, M., Li, R., Fu, Y., Zhong, L., Kang, S., Zhao, D., and Yang, Y.: Impact of topography on black carbon transport to the southern Tibetan Plateau during the pre-monsoon season and its climatic implication, Atmospheric Chemistry and Physics, 20, 5923-5943, https://doi.org/10.5194/acp-20-5923-2020, 2020.
Zheng, J., Hu, M., Du, Z. F., Shang, D. J., Gong, Z. H., Qin, Y. H., Fang, J. Y., Gu, F. T., Li, M. R., Peng, J. F., Li, J., Zhang, Y. Q., Huang, X. F., He, L. Y., Wu, Y. S., and Guo, S.: Influence of biomass burning from South Asia at a high-altitude mountain receptor site in China, Atmospheric Chemistry and Physics, 17, 6853-6864, https://doi.org/10.5194/acp-17-6853-2017, 2017.
-
RC2: 'Comment on egusphere-2023-1581', Anonymous Referee #2, 28 Aug 2023
General Comments:
The aerosols transported from South Asia (SA) are the important sources to the Tibetan Plateau (TP). Given the complex topography and limited observations in these areas, the aerosol simulations are still prone to large uncertainties. This paper shows the effects of the assimilation of Geostationary satellite Himawari-8 observations on aerosol forecasts and direct radiative effects (DRE) during the period of pollutants transporting from SA to the TP. The results demonstrate the AOT analysis and forecast fields can both effectively reproduce diurnal variations by assimilating all the observations within 12h window. Compared to the free-run experiment, assimilation experiments with aerosol radiative effect activation significantly improve the prediction of downward solar radiation and thereby enhance the transport of pollutants to the TP.
In general, the paper is well-written, and the results are nicely presented and well discussed with novel scientific founds. The authors did a credible job in analyzing how different experiment settings. The findings help better understand the transport of pollutants from South Asia to TP. Therefore, I would like to recommend accepting it to ACP after addressing the following minor comments
- Compared to DA_REON_12H experiment, the DA_REON_01H experiment performs better results in self-check due to assimilation of the entire window of observation data, but why a little worse in independent verification? Could you explain it in detail?
Specific comments:
- L83 Please rewrite this sentence
- L112-113 Please rewrite the sentence “The observations assimilated are Himawari-8 AOT and the main assimilated region is SA”.
- L126 It's better to write “Carbon Bond Mechanism Z”
- L126 Do you mean NMVOCs (non-methane volatile organic compounds)?
- L247 Replace " Fig. 1b " with " Fig. 1a "
- L273, L279 and L402, L403 Some figure number need to be consistent, e.g., “Figure 2”vs “Fig. 2”, “Figures 9a”vs “Figs. 9b”
- L281 “high surface albedo” maybe more precise
- L282 Please revise the sentence to make it clearer.
- L312 Replace " to absorb" with " aims to incorporate " to make it more clarity and fluidity
- L319 Which statistical indicators?
- L378 change the “Compare” to “Compared”
- L479 “surface temperature” maybe more precise
Citation: https://doi.org/10.5194/egusphere-2023-1581-RC2 -
AC2: 'Reply on RC2', Tie Dai, 16 Oct 2023
Response to the Reviewer #2
General comments:
The aerosols transported from South Asia (SA) are the important sources to the Tibetan Plateau (TP). Given the complex topography and limited observations in these areas, the aerosol simulations are still prone to large uncertainties. This paper shows the effects of the assimilation of Geostationary satellite Himawari-8 observations on aerosol forecasts and direct radiative effects (DRE) during the period of pollutants transporting from SA to the TP. The results demonstrate the AOT analysis and forecast fields can both effectively reproduce diurnal variations by assimilating all the observations within 12h window. Compared to the free-run experiment, assimilation experiments with aerosol radiative effect activation significantly improve the prediction of downward solar radiation and thereby enhance the transport of pollutants to the TP.
In general, the paper is well-written, and the results are nicely presented and well discussed with novel scientific founds. The authors did a credible job in analyzing how different experiment settings. The findings help better understand the transport of pollutants from South Asia to TP. Therefore, I would like to recommend accepting it to ACP after addressing the following minor comments
Response: We thank the reviewer for the encouraging comments. We have revised the manuscript following your comments.
Mainly concerns:
- Compared to DA_REON_12H experiment, the DA_REON_01H experiment performs better results in self-check due to assimilation of the entire window of observation data, but why a little worse in independent verification? Could you explain it in detail?
Response: Thanks. Considering the analysis fields, it is true that the validation results of DA_REON_01H perform better at the time of the self-test. This is due to the fact that the aerosol field at the hour of assimilation for this set of tests is then examined against itself, which is a much better time match. In the independent test, the analysis field is interpolated to the site of the independent test observations (MODIS or AERONET), which is different in time and space from the assimilated Himawari-8. At this point the asynchronous assimilation, absorbing the entire time window of DA_REON_12H would have performed relatively better. This is because the aerosol field before and after the assimilation time was also assimilated in this experiment.
Changes in Manuscript:
Line 317-320:
The AOT analyses are interpolated to the MODIS, which is different in temporal and spatial distribution from Himawari-8. The AOT analyses in DA_REON_12H absorb the entire window of observations, and this asynchronous assimilation corrects the AOT analysis field at all times in each window (Dai et al., 2019).
Specific comments:
- L83 Please rewrite this sentence
Response: Done. Thank you for your comment.
Changes in Manuscript:
Line 81:
Aerosols are second only to clouds in regional surface solar energy simulations.
- L112-113 Please rewrite the sentence “The observations assimilated are Himawari-8 AOT and the main assimilated region is SA”.
Response: Done. Thank you.
Changes in Manuscript:
Line 116-117:
The assimilated observations are Himawari-8 AOT, and the main assimilated region is SA.
- L126 It's better to write “Carbon Bond Mechanism Z”
Response: Thank you, we've corrected it.
- L126 Do you mean NMVOCs (non-methane volatile organic compounds)?
Response: Thank you for your comment, we've corrected it.
- L247 Replace " Fig. 1b " with " Fig. 1a "
Response: Done. Thank you.
- L273, L279 and L402, L403 Some figure number need to be consistent, e.g., “Figure 2”vs “Fig. 2”, “Figures 9a”vs “Figs. 9b”
Response: Done. Thank you for your comment, uniformly use Fig. or Figs.
- L281 “high surface albedo” maybe more precise
Response: Done. Thank you for your comment.
- L282 Please revise the sentence to make it clearer.
Response: Done. Thank you.
Changes in Manuscript:
Line 285-286:
The assimilated experiments perform very well regarding the AOT analyzed fields, which are more consistent with the Himawari-8 retrievals.
- L312 Replace " to absorb" with " aims to incorporate " to make it more clarity and fluidity
Response: Done. Thank you for your comment.
- L319 Which statistical indicators?
Response: The statistical indicators mainly refer to the bias.
- L378 change the “Compare” to “Compared”
Response: Thank you for your comment, we've corrected it.
- L479 “surface temperature” maybe more precise
Response: Thank you, we've corrected it in Line 489.
Peer review completion
Journal article(s) based on this preprint
Min Zhao et al.
Min Zhao et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 278 | 88 | 25 | 391 | 30 | 10 | 12 |
- HTML: 278
- PDF: 88
- XML: 25
- Total: 391
- Supplement: 30
- BibTeX: 10
- EndNote: 12
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(4140 KB) - Metadata XML
-
Supplement
(94 KB) - BibTeX
- EndNote
- Final revised paper