Microwave and submillimeter wave scattering of oriented ice particles

Brath, Manfred; Ekelund, Robin; Eriksson, Patrick; Lemke, Oliver; Buehler, Stefan A.

doi:https://doi.org/10.5194/amt-13-2309-2020

Articles | Volume 13, issue 5

https://doi.org/10.5194/amt-13-2309-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/amt-13-2309-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 13, issue 5

Research article

| Highlight paper

|

13 May 2020

Research article | Highlight paper |

| 13 May 2020

Microwave and submillimeter wave scattering of oriented ice particles

Manfred Brath, Robin Ekelund, Patrick Eriksson, Oliver Lemke, and Stefan A. Buehler

Download

Final revised paper (published on 13 May 2020)
Preprint (discussion started on 14 Oct 2019)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Brath et al. review', Anonymous Referee #1, 03 Dec 2019
- AC1: 'Answers to Anonymous Referee #1', Manfred Brath, 31 Jan 2020
SC1: 'Comments on Microwave and submillimeter wave scattering of oriented ice particles', Davide Ori, 03 Dec 2019
- AC3: 'Answers to short comment #1 (Davide Ori)', Manfred Brath, 31 Jan 2020
RC2: 'Review of Microwave and submillimeter wave scattering of oriented ice particles', Anonymous Referee #2, 09 Dec 2019
- AC2: 'Answers to Anonymous Referee #2', Manfred Brath, 31 Jan 2020

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Manfred Brath on behalf of the Authors (17 Feb 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (21 Feb 2020) by Stefan Kneifel

RR by Anonymous Referee #2 (27 Feb 2020)

RR by Anonymous Referee #3 (03 Mar 2020)

Suggestions for revision or reasons for rejection

The quality of the paper improved substantially with respect to the first submission.
I appreciated how the authors have incorporated the suggestion of the reviewers and increased the scientific value of the paper.
I want to specially mention the new block between lines 300 and 314. Here the authors make a very nice explanation of why the approach of sampling the scattering features on an icosahedral (refinable) grid is superior to other approaches. This approach is ideal for the scope of the paper and it is superior even to other equally spaced grids like the Fibonacci set which has been found good for averaging the scattering properties of particles (Zhang 2018) but does not provide the same refining advantage.

My recommendation is to publish after the authors take the chance to consider a few last points and in particular one regarding the accuracy of the calculations.

My line reference will point to the line numbering of the edited version made available to the reviewers.

MAJOR points:

THE ACCURACY OF THE DDA CALCULATIONS:
At the moment, I am not convinced about the various statements made on the accuracy of the DDA calculations.
By looking at the answers given to reviewer 2, the answer to the public comments and the text edited it is implied that the scattering properties of the database are accurate within a few percent errors.
I want to stress the fact that this level of accuracy is not proven in the paper.

The accuracy of the DDA calculations for irregular particles are difficult to assess.
The general mkd rule is to be taken as a "rule of thumb" for the validity of the calculations. It is rather a necessary condition than a sufficient one to claim that the DDA calculations are reliable. To have a general idea of this, one can just look at the various limits that have been suggested for the mkd maximum value depending on the required application. The highest one to my knowledge is 42 dipoles per wavelength (roughly double the resolution included in this database) for having a good backscattering cross-section (Petty and Huang 2010).

One of the reasons to reduce the internal stopping criterion to 0.01 might be the difficulty of the DDA algorithm to converge in case of very large size parameters (see fig 1 in Yurkin et al 2007). If the DDA algorithm struggles to reach a convergence point how can the final result really be trusted.
The internal stopping criterion only measures the convergence of the internal electric fields and not the accuracy of the far-field scattering quantities. Moreover, the convergence is measured in terms of the relative norm which means that locally the internal electric field might be much more biased. For a study on how internal electric fields structure influence the far-field scattering I recommend McCusker et al. (2019)
A similar point can be made for the spherical harmonics interpolation which only ensures the RMSE of the M matrix to be within 0.5% overall and not for specific directions.

Very few published studies assess the accuracy of DDA for specific applications (e.g.Yurkin et al (2006) and more recently Ori and Kneifel (2018)).
The problem boils down to two distinct issues: how accurate the DDA shapefiles can represent the real shape that you want to simulate and how much the DDA assumptions to be valid. These are referred as shape and discretization error in the aforementioned references.
To assess the accuracy of a scattering calculation one has to compare his/her own calculation with something more accurate both in terms of shape and discretization.

Considering the stated purpose of the database (simulating MW and subMM polarized satellite radiances) one should also consider how much the available range of shapes and sizes can be considered representative of the natural variability of snowflakes affecting the radiation coming to the satellite sensor. To me, it is very unlikely that few sizes of hexagonal plates and plate aggregates can be representative of that. Another particle shape with similar size and mass is likely to give a very different scattering feature.

I want to be extremely clear here. This database is an incredible step forward in the understanding of ice cloud radiative properties in the microwave. Moreover, I do not think that having a database accurate up to a few percent is even necessary. The uncertainties in the scattering calculations are likely to be vastly overcome by the many other factors affecting all-sky radiative transfer calculations. Even a much less accurate database would be useful to extract some information from the polarized measurements of ice clouds.
I just think that, despite the vast effort shown in the description of the database construction, there is not enough evidence to make any claim about the accuracy of the scattering calculations and the authors should just present the procedure that they applied as it is.

THE SHAPES USED:
Line 200-213 It is stated that the shapes used for the pristine plates (type 1) are different from those used in Eriksson (2018). However, I see differences also in the aggregates. By looking at the content of the database particles with Dveq<555 um seems to have a corresponding counterpart in Eriksson (2018). More massive particles have a corresponding shape in Eriksson (2018) in terms of mass, but not in terms of Dmax; they are all slightly different by approximately 1%
What changed here?
If the shapefiles are different from those included in Eriksson (2018) I suggest editing the database including the shapefiles as well.

MINOR points:

Line 104 - I wouldn't mention the real orientation as a result of the "balance" of gravitational and aerodynamical drag forces (plus others).
Due to the complex shape of snowflakes and the turbulent nature of air motion it is unlikely that these forces will ever be found in balance. Snowflakes are observed to continuously move, change direction, spin and tumble. This is the result of unbalanced forces. I would rephrase say just that the exact orientation is the result of the effect of several forces.

Line 138 (equation 8) I think on the lefthand side it is the cosine of Theta, not Theta itself.
There is also a typo in the cosine of the difference of the azimuth angles
The formula is very easy to understand in connection to Figure D1 in the appendix, why not refer to it?

Line 221 (Table2) I did not understand why the frequencies are organized in channel sets and how these sets are used.
Is it possible that the channel sets are designed as limits to interpolate within? It was not clearly stated anywhere else

Line 533 - You changed the frequency from 164.1 to 165.1 GHz. I didn't understand why. In table 2 only 164.1 is included and I have checked also the database and there is only 164.1 there.
If the scattering at 165.1 is obtained interpolating in channel set 6 (see my previous comment) isn't it the same calculating 166 by interpolation and doing the average between 164.1 and 166.9?

Line 554 - mirrorsymmetric -> mirror symmetric

Line 666 - Didn't you said that the original DDA data has been erased? If you can only provide the single orientation results expanded on the spherical harmonics that is what you should mention in the data availability section and the text.

References:

Zhang (2018) On numerical orientation averaging with spherical Fibonacci point sets and compressive scheme
https://doi.org/10.1016/j.jqsrt.2017.10.026

McCusker et al (2019) Analysis of the internal electric fields of pristine ice crystals and aggregate snowflakes, and their effect on scattering
https://doi.org/10.1016/j.jqsrt.2019.04.019

Petty and Huang (2010) Microwave Backscatter and Extinction by Soft Ice Spheres and Complex Snow Aggregates
https://doi.org/10.1175/2009JAS3146.1

Yurkin et al (2006) Convergence of the discrete dipole approximation. II. An extrapolation technique to increase the accuracy
DOI: 10.1364/JOSAA.23.002592

Yurkin M.A., Maltsev V.P., and Hoekstra A.G. (2007) The discrete dipole approximation for simulation of light scattering by particles much larger than the wavelength. https://doi.org/10.1016/j.jqsrt.2007.01.033

Ori and Kneifel (2018) Assessing the uncertainties of the discrete dipole approximation in case of melting ice particles
https://doi.org/10.1016/j.jqsrt.2018.06.017

Hide

RR by Anonymous Referee #1 (13 Mar 2020)

ED: Publish subject to minor revisions (review by editor) (17 Mar 2020) by Stefan Kneifel

AR by Manfred Brath on behalf of the Authors (27 Mar 2020) Author's response Manuscript

ED: Publish as is (31 Mar 2020) by Stefan Kneifel

AR by Manfred Brath on behalf of the Authors (07 Apr 2020)

Short summary

Microwave dual-polarization observations consistently show that larger atmospheric ice particles tend to have a preferred orientation. We provide a publicly available database of microwave and submillimeter wave scattering properties of oriented ice particles based on discrete dipole approximation scattering calculations. Detailed radiative transfer simulations, recreating observed polarization patterns, are additionally presented in this study.