the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Submitted as: model evaluation paper 08 Feb 2021
Submitted as: model evaluation paper | 08 Feb 2021
Simulation study of a Squall line hailstorm using High-Resolution GRAPES-Meso with a modified Double-Moment Microphysics scheme
- 1National Meteorological Center, Beijing, 100081, China
- 2Numerical Weather Prediction Center of China Meteorological Administration, Beijing, 100081, China
- 3Chongqing Meteorological Service, Chongqing, 401147, China
- 1National Meteorological Center, Beijing, 100081, China
- 2Numerical Weather Prediction Center of China Meteorological Administration, Beijing, 100081, China
- 3Chongqing Meteorological Service, Chongqing, 401147, China
Abstract. This study uses the high-resolution GRAPES_Meso (the mesoscale version of the Global/Regional Assimilation and Prediction System) to simulate a severe squall line hailstorm in Shandong province. The accumulated precipitation, radar reflectivity, and cloud hydrometeor properties simulated using a modified double-moment microphysics scheme are compared with observation. Results show that simulations captured the basic character of this squall line hailstorm. The simulated accumulation precipitation and radar reflectivity are comparable with the observation. The cross-section of the dynamic, microphysical, and radar reflectivity structures of the simulated hailstorm was analyzed. The simulated hailstorm has shown a reasonable result in both macrostructure and micro hail production rates. The development of the simulated hailstorm is consistent with the conceptual model of hailstorm evolution. Results imply the ability of high-resolution GRAPES_Meso on forecasting hailstorm.
-
Withdrawal notice
This preprint has been withdrawn.
-
Preprint
(1763 KB)
Zhe Li et al.
Interactive discussion
Status: closed
-
RC1: 'Comment on gmd-2020-439', Anonymous Referee #1, 10 Mar 2021
SUMMARY:
This manuscript summarizes a mesoscale modeling study of a hailstorm using the GRAPES-Meso NWP model with a 2-moment bulk microphysics scheme (BMS). The authors run GRAPES to simulate a hail-producing squall line for a 24-h simulation at a 3-km horizontal grid spacing. Comparisons are made between model fields and observations for accumulated precipitation and radar reflectivity. Some other model fields are also shown and discussed briefly, including vertical motion, hydrometeor mixing ratios and hail microphysical process rates. The authors claim that the simulation appears to be reasonable and conclude that GRAPES-Meso with the 2-moment BMS is therefore capable of simulating hailstorm (and presumably of predicting hail).
To be frank, there is very little scientific value in this manuscript for the meteorological, NWP, or modeling community. There is very little depth in the analysis. There is no way of telling what the effects of the 2-moment BMS are since no comparisons to other schemes are made, nor any sensitivity tests conducted. The comparisons to observations are very superficial and show little more than that the model happened to produce a reasonable simulation for this single case. The authors do not even show accumulated hail from the model for comparison to the surface observations (Fig. 1). There is not really any knowledge demonstrated about hail or hail modeling in the manuscript. So, in my opinion there is no publishable material here. Any necessary revisions needed would be too great to turn this into a publishable paper.
SOME SPECIFIC COMMENTS:
The background description of natural hail growth is quite weak. For example, “conversion” of graupel to hail is discussed as though this were a natural process (it is not; it is a modeling concept). No mention of frozen raindrops as hail embryos is made.
The description of the 2-moment scheme is strange. For example, most of the equations given are “final” equations, without the original “base” equations (though references are given) – but what is the purpose of this? The reader is not going to try to code a BMS based on these equations. The functional form of the hydrometeor size distributions is not stated. There are also a few strange aspects to this scheme that I see – e.g., what is the basis for a fixed collection efficiency of 0.8 (line 98)? What is the physical basis and meaning of the conversion parameter A (line 103)? Is there no distinction between wet and dry growth of hail?
The comparison to observations is quite week. For the observed precipitation (Fig. 4a), is this radar-based or gauge-based? Why are not plots of model hail precipitation (similar to Fig. 4b) shown (for comparison to Fig. 1)? [I recognize that hail mixing ratios at the surface are shown in Fig. 8.] For the radar reflectivity comparisons, this is tempting and common thing to do, but there are subtleties in model reflectivity that must be understood (and should be discussed in this paper). For example, uncontrolled size sorting in 2-moment bulk schemes can lead to an artificially broad size distribution, which inflates the calculation of the 6th moment (reflectivity).
There is no discussion about the impact of the specific model configuration. It is well recognized that grid spacing of 3 km is quite coarse and insufficient to resolve the updrafts in severe convection. But updrafts are strongly linked to hail (in nature and models). So what does this imply as far as this study is concerned? Discussion is needed.
What should be take away from the hail production rates shown in Fig. 10? There is a brief description of this figure in the text, but no discussion of what the reader should learn from this regarding the utility of this model to simulate hail. Is this better than a 1-moment BMS? What strikes me as the strangest is that most of the hail growth comes from “autoconversion” of graupel to hail, rather than accretion of liquid water (this does not seem realistic).
-
CEC1: 'Comment on gmd-2020-439', Juan Antonio Añel, 03 Apr 2021
Dear authors,
We have some concerns about the availability of code and data for your paper. First, we need to have a better explanation of what is the "confidential requirement" that you mention. Indeed, I would like to read it. Is it a contract? Maybe a license?. Ideally, you should grant access to the GRAPES-Meso code.
Also, the fact that you can not distribute the GRAPES-Meso model does not imply that you can not share the improved microphysics scheme you developed. Therefore, please, provide the code for the new scheme that you have developed. If you can not do it, then we would like to see an explanation of the reasons for not doing it.
Finally, it would be better to move the data included with the paper to a long-term and more trustful repository, such as Zenodo. Moreover, this would avoid the problems of realising on a web page in Chinese that it is necessary to introduce a download code and then look for it in the manuscript.
Juan A. Añel
Geosc. Mod. Dev. Executive Editor -
RC2: 'Comment on gmd-2020-439', Anonymous Referee #2, 16 Apr 2021
General comments
This manuscript described a GRAPES-meso simulation of a hail-producing storm using a double-moment microphysics scheme. The results are compared with radar and surface observations. Some routine analysis/comparison were done and it was concluded that the model was able to simulate the case reasonably well. Overall, the scientific merit of the study is pretty poor since it does not provide new knowledge or understanding in terms of either the scheme or storm evolution/characteristics. There is no quantitative comparison at all. Actually, the position of simulated storm is displaced north by 200 km and weaker (Fig. 4). Probably due to this displacement, simulated hail occurred mostly over the water. but there is no discussion about this at all.
For the scheme description part, those equations are simply presented without any justification and references. What are the physical reasoning behind to use those empirical parameters? What are the size distribution function assumed? How does the scheme differ from other schemes?
Without comprehensive/quantitative comparison with available observations (e.g., hail size, accumulated rainfall, among others), it is hard to be convinced the simulation well captured the storm. The study would benefit if another simulation using a different microphysics scheme can be conducted and contrasted. Finally, a detailed observational analysis and diagnosis of the storm itself, including its evolution, might be worth further investigation.
Specific comments
References to the observational and modeling effort of hail storm in the introduction is rather limited and some of them might be outdated.
Do the radar observations indicate any hints of the dry/wet growth of hail?
Note that the model calculated radar reflectivity might be significantly biased without proper consideration of hail properties.
For the cross section analysis, it would be nice to choose one perpendicular to the squall line.
Interactive discussion
Status: closed
-
RC1: 'Comment on gmd-2020-439', Anonymous Referee #1, 10 Mar 2021
SUMMARY:
This manuscript summarizes a mesoscale modeling study of a hailstorm using the GRAPES-Meso NWP model with a 2-moment bulk microphysics scheme (BMS). The authors run GRAPES to simulate a hail-producing squall line for a 24-h simulation at a 3-km horizontal grid spacing. Comparisons are made between model fields and observations for accumulated precipitation and radar reflectivity. Some other model fields are also shown and discussed briefly, including vertical motion, hydrometeor mixing ratios and hail microphysical process rates. The authors claim that the simulation appears to be reasonable and conclude that GRAPES-Meso with the 2-moment BMS is therefore capable of simulating hailstorm (and presumably of predicting hail).
To be frank, there is very little scientific value in this manuscript for the meteorological, NWP, or modeling community. There is very little depth in the analysis. There is no way of telling what the effects of the 2-moment BMS are since no comparisons to other schemes are made, nor any sensitivity tests conducted. The comparisons to observations are very superficial and show little more than that the model happened to produce a reasonable simulation for this single case. The authors do not even show accumulated hail from the model for comparison to the surface observations (Fig. 1). There is not really any knowledge demonstrated about hail or hail modeling in the manuscript. So, in my opinion there is no publishable material here. Any necessary revisions needed would be too great to turn this into a publishable paper.
SOME SPECIFIC COMMENTS:
The background description of natural hail growth is quite weak. For example, “conversion” of graupel to hail is discussed as though this were a natural process (it is not; it is a modeling concept). No mention of frozen raindrops as hail embryos is made.
The description of the 2-moment scheme is strange. For example, most of the equations given are “final” equations, without the original “base” equations (though references are given) – but what is the purpose of this? The reader is not going to try to code a BMS based on these equations. The functional form of the hydrometeor size distributions is not stated. There are also a few strange aspects to this scheme that I see – e.g., what is the basis for a fixed collection efficiency of 0.8 (line 98)? What is the physical basis and meaning of the conversion parameter A (line 103)? Is there no distinction between wet and dry growth of hail?
The comparison to observations is quite week. For the observed precipitation (Fig. 4a), is this radar-based or gauge-based? Why are not plots of model hail precipitation (similar to Fig. 4b) shown (for comparison to Fig. 1)? [I recognize that hail mixing ratios at the surface are shown in Fig. 8.] For the radar reflectivity comparisons, this is tempting and common thing to do, but there are subtleties in model reflectivity that must be understood (and should be discussed in this paper). For example, uncontrolled size sorting in 2-moment bulk schemes can lead to an artificially broad size distribution, which inflates the calculation of the 6th moment (reflectivity).
There is no discussion about the impact of the specific model configuration. It is well recognized that grid spacing of 3 km is quite coarse and insufficient to resolve the updrafts in severe convection. But updrafts are strongly linked to hail (in nature and models). So what does this imply as far as this study is concerned? Discussion is needed.
What should be take away from the hail production rates shown in Fig. 10? There is a brief description of this figure in the text, but no discussion of what the reader should learn from this regarding the utility of this model to simulate hail. Is this better than a 1-moment BMS? What strikes me as the strangest is that most of the hail growth comes from “autoconversion” of graupel to hail, rather than accretion of liquid water (this does not seem realistic).
-
CEC1: 'Comment on gmd-2020-439', Juan Antonio Añel, 03 Apr 2021
Dear authors,
We have some concerns about the availability of code and data for your paper. First, we need to have a better explanation of what is the "confidential requirement" that you mention. Indeed, I would like to read it. Is it a contract? Maybe a license?. Ideally, you should grant access to the GRAPES-Meso code.
Also, the fact that you can not distribute the GRAPES-Meso model does not imply that you can not share the improved microphysics scheme you developed. Therefore, please, provide the code for the new scheme that you have developed. If you can not do it, then we would like to see an explanation of the reasons for not doing it.
Finally, it would be better to move the data included with the paper to a long-term and more trustful repository, such as Zenodo. Moreover, this would avoid the problems of realising on a web page in Chinese that it is necessary to introduce a download code and then look for it in the manuscript.
Juan A. Añel
Geosc. Mod. Dev. Executive Editor -
RC2: 'Comment on gmd-2020-439', Anonymous Referee #2, 16 Apr 2021
General comments
This manuscript described a GRAPES-meso simulation of a hail-producing storm using a double-moment microphysics scheme. The results are compared with radar and surface observations. Some routine analysis/comparison were done and it was concluded that the model was able to simulate the case reasonably well. Overall, the scientific merit of the study is pretty poor since it does not provide new knowledge or understanding in terms of either the scheme or storm evolution/characteristics. There is no quantitative comparison at all. Actually, the position of simulated storm is displaced north by 200 km and weaker (Fig. 4). Probably due to this displacement, simulated hail occurred mostly over the water. but there is no discussion about this at all.
For the scheme description part, those equations are simply presented without any justification and references. What are the physical reasoning behind to use those empirical parameters? What are the size distribution function assumed? How does the scheme differ from other schemes?
Without comprehensive/quantitative comparison with available observations (e.g., hail size, accumulated rainfall, among others), it is hard to be convinced the simulation well captured the storm. The study would benefit if another simulation using a different microphysics scheme can be conducted and contrasted. Finally, a detailed observational analysis and diagnosis of the storm itself, including its evolution, might be worth further investigation.
Specific comments
References to the observational and modeling effort of hail storm in the introduction is rather limited and some of them might be outdated.
Do the radar observations indicate any hints of the dry/wet growth of hail?
Note that the model calculated radar reflectivity might be significantly biased without proper consideration of hail properties.
For the cross section analysis, it would be nice to choose one perpendicular to the squall line.
Zhe Li et al.
Data sets
Simulation data from GRAPES_Meso for hail event Z. Li https://pan.baidu.com/share/init?surl=fxfk4f8OgMy9MQKRVZ6wrA
Zhe Li et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 242 | 69 | 13 | 324 | 4 | 6 |
- HTML: 242
- PDF: 69
- XML: 13
- Total: 324
- BibTeX: 4
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1