the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 Jun 2024
AI-NAOS: An AI-Based Nonspherical Aerosol Optical Scheme for Chemical Weather Model GRAPES_Meso5.1/CUACE
Abstract. The AI-based Nonspherical Aerosol Optical Scheme (AI-NAOS) is a newly developed aerosol optical module that improves the representation of aerosol optical properties for radiative transfer simulations in atmospheric models. It incorporates the nonsphericity and inhomogeneity (NSIH) of internally mixed aerosol particles through a deep learning method. Specifically, the AI-NAOS considers black carbon (BC) as fractal aggregates and models soil dust (SD) as super-spheroids, encapsulated partially or completely with hygroscopic aerosols such as sulfate, nitrate, and aerosol water. To obtain AI-NAOS, a database of the optical properties for the models was constructed using the invariant imbedding T-matrix method (IITM), and deep neural networks (DNN) were trained based on this database. In this study, the AI-NAOS was integrated into the mesoscale version 5.1 of Global/Regional Assimilation and Prediction System with Chinese Unified Atmospheric Chemistry Environment (GRAPES_Meso5.1/CUACE). Real-case simulations were conducted during a winter with high pollution, comparing BC aerosols evaluated using three schemes with spherical aerosol models (external-mixing, core-shell, and volume-mixing) and the AI-NAOS scheme. The results showed that NSIH effect led to a moderate estimation of absorbing aerosol optical depth (AAOD) and obvious changes in aerosol radiative effects, short-wave heating rates, temperature profiles, and boundary layer height. The AAOD values based on three spherical schemes were 70.4 %, 125.3 %, and 129.3 % over Sichuan Basin, benchmarked to the AI-NAOS results. Compared to the external-mixing scheme, the direct radiative effect (DRE) induced by the NSIH effect reached +1.6 W/m2 at the top-of-atmosphere (TOA) and -2.9 W/m2 at surface. The NSIH effect could enhance the short-wave heating rate, reaching 20 %. Thus, the warming effect at 700 hPa and the cooling effect on the ground were strengthened by 21 % and 13 %, reaching +0.04 and –0.10 K, which led to a reduction in the height of the Planetary Boundary Layer (PBL) by –11 meters. In addition, the precipitation was inhibited by the NSIH effect, causing a 15 % further decrease. Therefore, the NSIH effects demonstrated their non-negligible impacts and highlighted the importance of incorporating them into chemical weather models.
- Preprint
(9414 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
CEC1: 'Comment on gmd-2024-51', Juan Antonio Añel, 20 Jun 2024
Dear authors,
Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our "Code and Data Policy".
https://www.geoscientific-model-development.net/policies/code_and_data_policy.html
Despite you having done good work depositing both the NAOS code and the training data in Zenodo, you still need to address the issue regarding GRAPES_Meso5.1/CUACE correctly. In the text, you state, "We cannot distribute the GRAPES_Meso5.1/CUACE code due to its license". We can not accept this. It is mandatory to publish all the code necessary to reproduce a manuscript submitted, and it must be public at the submission time. We can consider some exceptions when authors provide evidence that sharing the code is forbidden to them (for example, by law) and totally out of their control. In such cases, public sharing of the code is not requested. Still, at minimum, you must deposit it in an acceptable long-term repository and provide editors access to the code.
Therefore, please publish the GRAPES_Meso5.1/CUACE code in one of the appropriate repositories listed in our policy and reply to this comment with the relevant information (link and DOI) as soon as possible, as we can not accept manuscripts in Discussions that do not comply with our policy. Therefore, the current situation with your manuscript is irregular.
If you think you qualify for an exception, please provide us evidence (license, laws and regulations) that you are forbidden to share the code and can not influence the decision on it. In such case, it continues to be mandatory that you deposite the code in a long-term private repository and share with the editors the details to access it, and include in the text of the manuscript and reply to this comment the link and DOI of the repository. For example, Zenodo provides such an option.
Note that if you do not fix this problem, we will have to reject your manuscript for publication in our journal.Juan A. Añel
Geosci. Model Dev. Executive EditorCitation: https://doi.org/10.5194/gmd-2024-51-CEC1 -
AC1: 'Reply on CEC1', Lei Bi, 01 Jul 2024
Dear Editor,
Thank you for your feedback regarding the "Code and Data Policy" for our manuscript.We would like to clarify that the GRAPES_Meso5.1/CUACE was previously published in Wang et al., 2022, Journal of Advances in Modeling Earth Systems, with the DOI: https://doi.org/10.1029/2022MS003222. We have cited this paper in our references. Alongside the publication, the data and related code were deposited in the respiratory at the following URL:
https://datadryad.org/stash/landing/show?id=doi%3A10.5061%2Fdryad.m63xsj45f.
Please be aware that the new contribution of this manuscript is only NOAS, which is not only applicable to the GRAPES/CUACE model, but can also be extended to other weather/climate models, such as WRF/Chem.
We hope that this clarification addresses the concerns raised regarding the “Code and Data Policy”. We appreciate your time and consideration.
Best regards,
Lei Bi & Co-authorsCitation: https://doi.org/10.5194/gmd-2024-51-AC1 -
CEC2: 'Reply on AC1', Juan Antonio Añel, 04 Jul 2024
Dear authors,
Thanks for your reply. However, I have to point out that the information you have provided for the repository is incorrect. You have linked a Dryad repository that contains data. The repository for GRAPES_Meso5.1/CUACE CW V1.0 is in Zenodo, (https://zenodo.org/records/7075751), and its DOI is 10.5281/zenodo.7075750. Please, note that this is the information you have to include in the "Code and Data Availability" section of your manuscript in any new version.
The fact that the contribution here in this work is NAOS does not preclude the need to publish all the code that it is necessary to replicate your work. This is necessary to comply with the scientific method.
Regards,
Juan A. Añel
Geosci. Model Dev. Executive Editor
Citation: https://doi.org/10.5194/gmd-2024-51-CEC2 -
AC2: 'Reply on CEC2', Lei Bi, 04 Jul 2024
Dear Editor,
Thank you for your prompt reply!
We will include the Zenodo link along with its corresponding DOI in any subsequent versions of our manuscript.
Best,
Lei Bi & Co-authors
Citation: https://doi.org/10.5194/gmd-2024-51-AC2
-
AC2: 'Reply on CEC2', Lei Bi, 04 Jul 2024
-
CEC2: 'Reply on AC1', Juan Antonio Añel, 04 Jul 2024
-
AC1: 'Reply on CEC1', Lei Bi, 01 Jul 2024
-
RC1: 'Comment on gmd-2024-51', Anonymous Referee #1, 12 Jul 2024
This manuscript introduces a deep learning approach to estimate the optical properties of internally mixed aerosol particles, considering the non-sphericity of insoluble particles. The research has significant implications for atmospheric models.
Comments for Improvement:
1. Generalizability: While the discussed machine learning (ML) approach demonstrates impressive accuracy, concerns remain about its generalizability. ML-based algorithms often struggle with generalization beyond their training data (Kumar et al., 2024). The authors should address this issue explicitly.
2. Comparison with Observed AOD: Including a comparison of results with observed Aerosol Optical Depth (AOD) would enhance the manuscript. This validation step provides valuable context.
3. Time Complexity: The authors do not discuss the time complexity of the developed approach. One of the main reasons scientists are increasingly pursuing ML is due to its computational cost advantages.
4. Figure 3: Clarify whether the results in Figure 3 pertain to the test or training dataset. Additionally, provide details on pre-processing and post-processing steps, which are currently missing.
5. Tuning of Hyperparameters: While the network's results are extraordinary, the manuscript lacks information on hyperparameters and their tuning. Including this would improve transparency.
6. Methodology Reorganization: Consider reorganizing the methodology section. Present results from AI-NAOS (lines 174-190, including Figure 2) after describing the approach in detail.
7. Move lines 208-216 and Figure 3 to the results section for better flow.
Reference:
Kumar, P., Vogel, H., Bruckert, J. et al. MieAI: a neural network for calculating optical properties of internally mixed aerosol in atmospheric models. npj Clim Atmos Sci 7, 110 (2024). https://doi.org/10.1038/s41612-024-00652-yCitation: https://doi.org/10.5194/gmd-2024-51-RC1 - AC3: 'Reply on RC1', Lei Bi, 28 Oct 2024
-
RC2: 'Comment on gmd-2024-51', Anonymous Referee #2, 21 Oct 2024
The manuscript introduces AI-NAOS, an AI-based module that incorporates nonspherical and inhomogeneous aerosol particles into radiative transfer simulations. The use of deep learning and advanced optical modeling is an innovative approach in this field, significantly improving accuracy in predicting the direct radiative effects of aerosols. Real-case simulations provide substantial evidence of the AI-NAOS module's impact on atmospheric thermodynamic structures and precipitation patterns, enhancing the understanding of aerosols in weather modeling. In general, the manuscript is well-organized. Only minor revision is needed before publication:
- The model ran for 72 hours starting from January 12, 2018, with only 24 hours of spin-up time, which seems rather short. Is it possible to extend the model's spin-up time and total run time?
- Is 'deep neural network (DNN)' used in the text as a general term for a class of methods, or does it refer to a specific neural network method? Please clarify this further in the text.
- Why were different methods chosen for shortwave radiation and longwave radiation? I believe that RRTM can also be used for shortwave calculations. What considerations influenced this decision?
- Could you compare it with the latest research on non-spherical black carbon, such as the study by Chen, G., Liu, C., Wang, J., Yin, Y., & Wang, Y., JGRA, 2024, which accounts for the mixing state, nonsphericity, and heterogeneity effects of black carbon in its optical property parameterization within a climate model?
- I believe there is still considerable uncertainty regarding the impact on precipitation in this section. It's worth further consideration whether to include this in the article. Additionally, while not absolutely necessary, it might be beneficial to examine changes in the boundary layer as well.
Some minor comments:
- In Line 40, omega is used to represent SSA, while in Line 271, <SSA> is used. Please ensure consistency. The same applies to the asymmetry factor.
- Line 80: It is usually referred to as 'WRF-Chem' rather than 'WRF/chem'.
- The vertical axis of Figure 6b is labeled 'atmo. abs.', which seems less rigorous than 'DRE'.
- Section 3.4 appears twice; the precipitation part should be labeled as 3.5.
Citation: https://doi.org/10.5194/gmd-2024-51-RC2 - AC4: 'Reply on RC2', Lei Bi, 28 Oct 2024
Status: closed
-
CEC1: 'Comment on gmd-2024-51', Juan Antonio Añel, 20 Jun 2024
Dear authors,
Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our "Code and Data Policy".
https://www.geoscientific-model-development.net/policies/code_and_data_policy.html
Despite you having done good work depositing both the NAOS code and the training data in Zenodo, you still need to address the issue regarding GRAPES_Meso5.1/CUACE correctly. In the text, you state, "We cannot distribute the GRAPES_Meso5.1/CUACE code due to its license". We can not accept this. It is mandatory to publish all the code necessary to reproduce a manuscript submitted, and it must be public at the submission time. We can consider some exceptions when authors provide evidence that sharing the code is forbidden to them (for example, by law) and totally out of their control. In such cases, public sharing of the code is not requested. Still, at minimum, you must deposit it in an acceptable long-term repository and provide editors access to the code.
Therefore, please publish the GRAPES_Meso5.1/CUACE code in one of the appropriate repositories listed in our policy and reply to this comment with the relevant information (link and DOI) as soon as possible, as we can not accept manuscripts in Discussions that do not comply with our policy. Therefore, the current situation with your manuscript is irregular.
If you think you qualify for an exception, please provide us evidence (license, laws and regulations) that you are forbidden to share the code and can not influence the decision on it. In such case, it continues to be mandatory that you deposite the code in a long-term private repository and share with the editors the details to access it, and include in the text of the manuscript and reply to this comment the link and DOI of the repository. For example, Zenodo provides such an option.
Note that if you do not fix this problem, we will have to reject your manuscript for publication in our journal.Juan A. Añel
Geosci. Model Dev. Executive EditorCitation: https://doi.org/10.5194/gmd-2024-51-CEC1 -
AC1: 'Reply on CEC1', Lei Bi, 01 Jul 2024
Dear Editor,
Thank you for your feedback regarding the "Code and Data Policy" for our manuscript.We would like to clarify that the GRAPES_Meso5.1/CUACE was previously published in Wang et al., 2022, Journal of Advances in Modeling Earth Systems, with the DOI: https://doi.org/10.1029/2022MS003222. We have cited this paper in our references. Alongside the publication, the data and related code were deposited in the respiratory at the following URL:
https://datadryad.org/stash/landing/show?id=doi%3A10.5061%2Fdryad.m63xsj45f.
Please be aware that the new contribution of this manuscript is only NOAS, which is not only applicable to the GRAPES/CUACE model, but can also be extended to other weather/climate models, such as WRF/Chem.
We hope that this clarification addresses the concerns raised regarding the “Code and Data Policy”. We appreciate your time and consideration.
Best regards,
Lei Bi & Co-authorsCitation: https://doi.org/10.5194/gmd-2024-51-AC1 -
CEC2: 'Reply on AC1', Juan Antonio Añel, 04 Jul 2024
Dear authors,
Thanks for your reply. However, I have to point out that the information you have provided for the repository is incorrect. You have linked a Dryad repository that contains data. The repository for GRAPES_Meso5.1/CUACE CW V1.0 is in Zenodo, (https://zenodo.org/records/7075751), and its DOI is 10.5281/zenodo.7075750. Please, note that this is the information you have to include in the "Code and Data Availability" section of your manuscript in any new version.
The fact that the contribution here in this work is NAOS does not preclude the need to publish all the code that it is necessary to replicate your work. This is necessary to comply with the scientific method.
Regards,
Juan A. Añel
Geosci. Model Dev. Executive Editor
Citation: https://doi.org/10.5194/gmd-2024-51-CEC2 -
AC2: 'Reply on CEC2', Lei Bi, 04 Jul 2024
Dear Editor,
Thank you for your prompt reply!
We will include the Zenodo link along with its corresponding DOI in any subsequent versions of our manuscript.
Best,
Lei Bi & Co-authors
Citation: https://doi.org/10.5194/gmd-2024-51-AC2
-
AC2: 'Reply on CEC2', Lei Bi, 04 Jul 2024
-
CEC2: 'Reply on AC1', Juan Antonio Añel, 04 Jul 2024
-
AC1: 'Reply on CEC1', Lei Bi, 01 Jul 2024
-
RC1: 'Comment on gmd-2024-51', Anonymous Referee #1, 12 Jul 2024
This manuscript introduces a deep learning approach to estimate the optical properties of internally mixed aerosol particles, considering the non-sphericity of insoluble particles. The research has significant implications for atmospheric models.
Comments for Improvement:
1. Generalizability: While the discussed machine learning (ML) approach demonstrates impressive accuracy, concerns remain about its generalizability. ML-based algorithms often struggle with generalization beyond their training data (Kumar et al., 2024). The authors should address this issue explicitly.
2. Comparison with Observed AOD: Including a comparison of results with observed Aerosol Optical Depth (AOD) would enhance the manuscript. This validation step provides valuable context.
3. Time Complexity: The authors do not discuss the time complexity of the developed approach. One of the main reasons scientists are increasingly pursuing ML is due to its computational cost advantages.
4. Figure 3: Clarify whether the results in Figure 3 pertain to the test or training dataset. Additionally, provide details on pre-processing and post-processing steps, which are currently missing.
5. Tuning of Hyperparameters: While the network's results are extraordinary, the manuscript lacks information on hyperparameters and their tuning. Including this would improve transparency.
6. Methodology Reorganization: Consider reorganizing the methodology section. Present results from AI-NAOS (lines 174-190, including Figure 2) after describing the approach in detail.
7. Move lines 208-216 and Figure 3 to the results section for better flow.
Reference:
Kumar, P., Vogel, H., Bruckert, J. et al. MieAI: a neural network for calculating optical properties of internally mixed aerosol in atmospheric models. npj Clim Atmos Sci 7, 110 (2024). https://doi.org/10.1038/s41612-024-00652-yCitation: https://doi.org/10.5194/gmd-2024-51-RC1 - AC3: 'Reply on RC1', Lei Bi, 28 Oct 2024
-
RC2: 'Comment on gmd-2024-51', Anonymous Referee #2, 21 Oct 2024
The manuscript introduces AI-NAOS, an AI-based module that incorporates nonspherical and inhomogeneous aerosol particles into radiative transfer simulations. The use of deep learning and advanced optical modeling is an innovative approach in this field, significantly improving accuracy in predicting the direct radiative effects of aerosols. Real-case simulations provide substantial evidence of the AI-NAOS module's impact on atmospheric thermodynamic structures and precipitation patterns, enhancing the understanding of aerosols in weather modeling. In general, the manuscript is well-organized. Only minor revision is needed before publication:
- The model ran for 72 hours starting from January 12, 2018, with only 24 hours of spin-up time, which seems rather short. Is it possible to extend the model's spin-up time and total run time?
- Is 'deep neural network (DNN)' used in the text as a general term for a class of methods, or does it refer to a specific neural network method? Please clarify this further in the text.
- Why were different methods chosen for shortwave radiation and longwave radiation? I believe that RRTM can also be used for shortwave calculations. What considerations influenced this decision?
- Could you compare it with the latest research on non-spherical black carbon, such as the study by Chen, G., Liu, C., Wang, J., Yin, Y., & Wang, Y., JGRA, 2024, which accounts for the mixing state, nonsphericity, and heterogeneity effects of black carbon in its optical property parameterization within a climate model?
- I believe there is still considerable uncertainty regarding the impact on precipitation in this section. It's worth further consideration whether to include this in the article. Additionally, while not absolutely necessary, it might be beneficial to examine changes in the boundary layer as well.
Some minor comments:
- In Line 40, omega is used to represent SSA, while in Line 271, <SSA> is used. Please ensure consistency. The same applies to the asymmetry factor.
- Line 80: It is usually referred to as 'WRF-Chem' rather than 'WRF/chem'.
- The vertical axis of Figure 6b is labeled 'atmo. abs.', which seems less rigorous than 'DRE'.
- Section 3.4 appears twice; the precipitation part should be labeled as 3.5.
Citation: https://doi.org/10.5194/gmd-2024-51-RC2 - AC4: 'Reply on RC2', Lei Bi, 28 Oct 2024
Data sets
The data presented in this paper Xuan Wang, Lei Bi, Hong Wang, Yaqiang Wang, Wei Han, Xueshun Shen, and Xiaoye Zhang https://doi.org/10.5281/zenodo.11208003
Model code and software
The AI-NAOS aerosol optical module codes Xuan Wang, Lei Bi, Hong Wang, Yaqiang Wang, Wei Han, Xueshun Shen, and Xiaoye Zhang https://doi.org/10.5281/zenodo.11181275
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 410 | 131 | 31 | 572 | 18 | 19 |
- HTML: 410
- PDF: 131
- XML: 31
- Total: 572
- BibTeX: 18
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1