Extremely strong polarization of an active asteroid (3200) Phaethon

The near-Earth asteroid (3200) Phaethon is the parent body of the Geminid meteor stream. Phaethon is also an active asteroid with a very blue spectrum. We conducted polarimetric observations of this asteroid over a wide range of solar phase angles α during its close approach to the Earth in autumn 2016. Our observation revealed that Phaethon exhibits extremely large linear polarization: P = 50.0 ± 1.1% at α = 106.5°, and its maximum is even larger. The strong polarization implies that Phaethon’s geometric albedo is lower than the current estimate obtained through radiometric observation. This possibility stems from the potential uncertainty in Phaethon’s absolute magnitude. An alternative possibility is that relatively large grains (~300 μm in diameter, presumably due to extensive heating near its perihelion) dominate this asteroid’s surface. In addition, the asteroid’s surface porosity, if it is substantially large, can also be an effective cause of this polarization.

Polarization measurements of solid bodies in the solar system have a long history, not all of it positive. The classic problem is that the interpretation of the polarization is ambiguous, because it is a function of very complicated optics in the porous regolith material that coats most solar system bodies.
Having said that, I found the Ito result to be quite unexpected and the paper to be very interesting. They have a measurement of polarization in Phaethon that is so extreme that the interpretation is relatively clear -big particles coat the surface. Even if the interpretation were suspect, Figure 2 shows that Phaethon is an extraordinary body by the standards of other solar system objects, which is itself interesting.
As an astronomer, I am interested in Phaethon because of its special role as the source of Geminids. I am not so sure of the broader interest in the community outside of astronomy: polarization is a hard-sell. This issue of the suitability for Nature, in my view, is the only weakness of the paper. If the paper is considered for Nature Astronomy, then I think it is a good match and I would support publication after re-writing in the Nature style.
Reviewer #2 (Remarks to the Author): I read through the paper by Ito et al on the particle size distribution of grain size from polarimetry observations. While I do not have major concerns with the science presented, I am not sure if this a result that scientifically interesting or ground-breaking that warrants its publication in Nature. We have know that spectral slope and particle size are related for many years now, so what is presented in the paper is not necessarily new. I will leave it to the editor to decide if this paper should be accepted but below are some specific comments to improve the manuscript. It would also be good to repeat the observations when Phaethon makes a close flyby next month and cover the lower phase angle range if possible.
Specific Comments: -Line 49: There are several asteroids that have blue spectra, so saying Phaethon is unique because of its blue spectrum is not accurate. Unique means one of a kind and in this case there is an entire class of asteroids that have this feature.
-Please don't use random acronyms like SSSBs. Just call them small bodies.
-Line 100: Please explain what kind of corrections you are doing to account for observations at 10 deg elevation? Also please list the airmass of the observations on Table 1. -Line 107: Please include the observation section as appendix or supplementary material.
-The altitude of the objects when the first three measurements were made is extremely low. Typically one does not observe below 2 airmass and in the case of the first three observations it was between 3 and 5 airmass. I have serious concerns about the lack of detailed quantification of the uncertainties plotted in Figure 1.
- Figure 1: Based on Figure 1 and if we trust the error bars for the highest phase angle data points, Phaethon has similar polarimetric properties as comets 2P and 209P. Given the general acceptance that Phaethon is an active asteroid/dead comet nuclei this result should not be too surprising as the authors claim.
- Figure 2: I don't think the data on this plot are valid for the discussion. Phaethon is a dark primitive comet-type object. The data shown in this plot are all silicate-rich material and not necessarily valid for comparing with Phaethon. Ideally this should have comets and primitive asteroids rather than Icarus (which is a Q-type), Mercury, and terrestrial rocks and lunar dust.
-The general discussion towards the end of the paper is just speculative and can be trimmed down. Same goes for the last paragraph which would be obsolete by the time this paper is published in another journal.
Reviewer #3 (Remarks to the Author): The paper presents unique measurements of the linear polarization degree of an asteroid of a rare spectral type. The observations covered a wide range of solar phase angles up to the angles above 100 deg, where the maximum of the polarization phase curve typically occurs. The authors have shown for the first time that the polarization degree of a Solar system body can reach extremely large values of 50%. It is substantially larger than the values obtained for any other small solar system bodies. This result shows that the surface properties of Solar system bodies are more diverse than considered before. The authors give important constraints on the surface characteristics of this particular asteroid Phaethon, which has been chosen as a target of future JAXA space mission.
I have two main comments to the paper: 1) there is no Table with Fornasier et al. paper in the reference list). The previous measurement was close to the inversion angle and revealed that the inversion angle of Phaethon is typical for main-belt asteroids. According to laboratory measurements (the authors cited these papers) not only polarization maxima but also inversion angles depend on grain sizes. Why Phaethon has usual inversion angle and extremely large polarization maximum? It is important to discuss in the paper all possible explanations of extremely large polarization and their shortcomings.
The paper is of great interest in a wider field than planetary science providing new knowledge to our understanding of light scattering processes. The paper is worth to be published after moderate revision.
• The frequency of multiple light scattering largely depends on albedo and grain size of 150 surface material of the target object. In case of Phaethon, its surface albedo has been 151 independently determined by mid-infrared thermal observation (e.g. Hanuš et al., 2016).

152
Therefore, it is possible to make a unique estimate of the effective grain size (through 153 Umow's law) from the value of P max obtained from polarimetric observation. Validity and 154 accuracy of this type of estimate has been established through laboratory experiments 155 using terrestrial samples (e.g. Dollfus and Geake, 1977; Geake and Dollfus, 1986).

156
• Based on the reasons described above, polarimetric observation to obtain P max has a 157 complementarity to spectroscopic observation in terms of estimating surface texture of 158 small solar system bodies.

159
• It is needless to say that polarimetric observation has been employed not only for studies 160 of the small solar system bodies, but also in a wide field of astronomy and physics. 161 Therefore we believe that our observation result presented in the present paper will have 162 a ripple effect on broader field of science.

163
Based on the background stated above, let us describe the significance of our own study 164 presented in this paper again:

165
• Investigation into the correlation between the maximum polarization degree of (P max ) of 166 the actual small solar system bodies (asteroids, comets, · · ·) and the surface grain size constraints about P max of asteroids or comets. The strongest point of our paper lies here.

171
• Combining our polarimetric observational result on Phaethon with the fact that this as-172 teroid greatly approaches the Sun on a regular basis, we surmise that the grain coarsening 173 5 effect due to the extensive heating (sintering) happens near at its perihelion, which grows 174 the grain size to a great deal. 175 • In addition, very strong solar radiation pressure that works on this asteroid near its 176 perihelion can also be relevant, as we mentioned in the paper.

177
• The large effective grain size inferred from our polarimetric observation is consistent 178 with the conventional knowledge known from the spectroscopic studies of this asteroid-179 Phaethon's spectrum is very blue.

180
• As Umow's law says, and as laboratory experiments have shown, solar system objects can 181 have large P max when their albedo is low. However, Phaethon's albedo is not low. This 182 is another statement that we raise in this paper.

183
Making an inference on the surface grain size of small bodies through polarimetric measure-184 ment has another advantage. We can think of a micro physical process that connects the strong 185 polarization and the coarse surface grains, even though qualitatively. As we mentioned in lines 186 118-127 in p. 8 of the manuscript, it is about the frequency difference of multiple scattering 187 of incident light at the unit optical depth. This fact (= that a picture/theory exists on micro 188 physical processes that account for the relation between large P max and coarse surface grains) 189 has an advantage over the circumstance about the correlation between the blue spectral slope 190 and the coarse surface grain size of small bodies: The correlation between blue spectrum and 191 coarse grains is still mostly empirical, and its micro physical background is not well understood. 192 Up to now we have stated our reasoning why we believe our polarimetric observation result is 193 important and useful in the context of small solar system studies in various ways. Let us also add 194 that Phaethon has many other interesting aspects as an active, near-Sun asteroid as well as the 195 parent object of a meteor stream. Therefore, we think that our detailed polarimetric observation 196 of Phaethon provides fundamental and underlying information on the surface environment of 197 small solar system bodies in a broad aspect.

198
Let us mention a point to note here. Different from P max , so-called the polarimetric neg-199 ative branch parameters (the minimum of polarization (P min ) and the solar phase angle that 200 marks the border between the negative branch and positive branch (α inv )) seem to have weaker 201 dependence on the surface texture of objects than P max . The negative branch parameters 202 seems to depend more strongly on small body's taxonomy (e.g. Belskaya et al., 2017). It is 203 likely that underlying physical processes that dominate the polarimetric positive branch and 204 negative branch are different. But at the same time, we can say that this difference indicates 205 that polarimetric measurement of small bodies serves as an effective, complementary tool to 206 spectroscopy, particularly when it is carried out both at small and large solar phase angles.   We also tried to obtain polarimetric data in the range of α < 30 • . But unfortunately, 220 bad weather wiped out our observation opportunities, ending up having no data in this range.

221
Therefore we could not obtain polarimetric data that can determine the inversion angle α inv or 222 the negative branch minimum P min .

223
As for Phaethon's polarization degree at its smaller solar phase angle, we consider that a    Thank you very much for the comment. We found two "unique" in the Introduction section 246 in the previous submission. We replaced the first one for "intriguing," and the second one for 247 "curious," in the revised version (p. 2). We hope this rephrase protects reviewers and readers 248 from getting any confusions.  Thank you very much for the suggestion. Following it, in the revised manuscript we spelled 251 it out and just use the term "small solar system bodies" or "small bodies" as suggested.  Table 1. 254 For the revised manuscript we made a new section named Methods, particularly a subsection through relative photometry). Even when it does, the influence is very little as we see in Figure   262 3. Please also see our response to the review comment #2-8.

263
Along with the request from the reviewer, we also added the airmass values when we ob-264 served the asteroid in Table 1 of the revised manuscript.   This is a large difference from ordinary photometric or spectroscopic observation. As is well known, Phaethon is an object that has characteristics both of an asteroid and 289 of a comet. Being recognized as the parent body of a major meteor stream, we may want to 290 say that it has a typical characteristics as a comet. As far as we are aware, however, in recent 291 years it is often emphasized that Phaethon is more like an asteroid, rather than a comet. Let 292 us itemize some studies that imply Phaethon is rather close to the asteroidal end-members: Jupiter that surround the Pallas family region in the main belt. From this dynamical 303 viewpoint, we had better say that Phaethon has an asteroidal origin.

305
Phaethon is known to be the parent body of the Geminid meteor stream. This fact 306 is usually associated with a conjecture that the parent body is a comet. However in 307 the case of Phaethon, there is a report that the typical density of the Geminid meteors  The relatively high albedo that Phaethon possesses produces a remarkable difference of 319 this object from others, as we showed in our Figure 2. Geometric albedo of the comets 320 2P and 209P is as low as ∼0.05 (see our response to the review comment #2-10).

321
Considering the above peculiarities and characteristics that Phaethon has, we believe this 322 object is not just a simple dead/extinct comet.  Figure 2 of the revised manuscript. This is for comparing the 333 (A, P max ) values of these comets with Phaethon, and better show their differences.

334
As for the above comment, let us mention two points. 335 First, as we explained in the response to the review comment #2-9, we think that com- object, we would like to say that it is rather closer to the asteroidal end-member based on the 339 consideration that we stated in our response to #2-9.

340
Here is the second point of ours. Hinted by the above comment (#2-10), we test-revised our 341 Figure 2 including five more objects whose P r data are plotted in Figure 1 < 50µm. Ra-Shalom, a C-type asteroid with a large albedo, is rather closer to Icarus. In any 352 case Phaethon's polarization degree is higher than any of the five objects, and it does not seem 353 to belong to any groups that other five objects belong to.

354
Based on the above discussion, we have updated Figure 2 in the revised manuscript including 355 the data points of for 2P and 209P whose P r are supposed to be close to their P max .

356
Incidentally, we altered the description of (1566) Icarus from as being an S-type asteroid to 357 being a Q-type asteroid (p. 4, line 59) along with the above comment from the reviewer. We 358 appreciate the reviewer for pointing it out, and leading us to a more precise description.  According to this suggestion, we removed/shrunk several descriptions from this section.  better now, and its presentation is more persuasive for readers.  About P min and α inv Now, let us move on to a discussion on the negative branch of po- between F-type and B-type asteroids. What distinguishes F-type asteroids from the B-type 507 asteroids is an absence of UV absorption (e.g. Belskaya et al., 2005). But observations to 508 confirm this feature are not easy to carry out.

509
Another feature that can be unique to F-type asteroids is the non-existence of absorption 510 in the 3µm wavelength. Infrared observation in this wavelength is not easy either, but let 511 us bring up an experimental study (Hiroi et al., 1996) as an example. We transcribed their  Figure 3. If Phaethon has some affinity characteristics to F-type as we conjecture, this 519 asteroid can be located near the F-type group in this diagram: Its inversion angle α inv would 520 be smaller than that of typical B-type asteroids. It could also mean that this asteroid has 521 a surface texture rather close to "dust-free" (or, covered by coarse grains) rather than fine grains. This would be consistent with our conclusion from our own polarimetric observation 523 of Phaethon that the effective grains on its surface are coarse, or the surface is practically free 524 from fine grains (except for a short period right after its perihelion passage and temporary dust 525 emission). 526 We thus infer that the result of our polarimetric measurement presented in this paper is 527 not inconsistent to, rather in accordance with, the polarimetric measurement in the negative 528 branch presented in past studies. As we have mentioned several times, we are aware that our 529 statement in the current response is still speculative, and needs a lot of quantitative confir-  In what follows we itemize three possible explanations that can cause strong polarization on 544 the surface of the small solar system bodies, as well as their shortcomings: Low albedo, large 545 porosity, and coarse grains.

547
When the albedo (A) of a solar system object is low, it usually shows strong polarization 548 (P max ). This is nothing but Umow's law. As we mentioned in the main manuscript as 549 19 well as in this document, multiple scattering of light happens more often on the surface 550 with high albedo than on the surface with low albedo. As a result, the polarization degree 551 (particularly P max ) of the surface with higher albedo gets weaker, and that with lower 552 albedo gets stronger. The advantage to adopt this mechanism is that, Umow's law is well 553 established and theoretically explained, particularly about P max .

554
However, we cannot ascribe the cause of Phaethon's strong polarization (large P max ) to 555 its albedo-its albedo is moderately high shown in Figure 2 of our manuscript. This is 556 the shortcoming in this case. In addition, let us note that the dependence of P min on 557 albedo is not as evident as P max (Belskaya et al., 2017) in general.

559
It has been numerically confirmed that large porosity increases polarization of mate- The shortcoming of this potential cause is that, we do not have precise information on 575 how porous/imporous Phaethon's (and most asteroids') surface is. It is rather possible 576 that Phaethon's surface is not quite porous, deducing from the very high bulk density of 577 the Geminid meteors (Borovička et al., 2010), but its direct measurement is yet to come.

578
Understanding the relation between Phaethon's surface porosity and its polarimetric char-579 acteristics (including both the positive and negative branches) is an important future task.

580
It can have an influence on broader studies of surface texture of the small solar system 581 bodies.

583
As we have mentioned in this document, laboratory experiments have confirmed that the 584 dominance of coarse grains on material surface causes strong polarization, particularly 585 P max . We still believe this mechanism is the likely cause of Phaethon's large P max . This 586 mechanism is consistent to other circumstances that this asteroid has, such as its very 587 blue spectrum, its temporary dust emission for a short period only after the perihelion 588 passages, and its very severe thermal environment.

589
As a shortcoming, we can say that the dependence of P min and α inv on grain size (or on 590 P max is harder to be measured than P min (such as the main belt asteroids), additional 592 information should be combined with polarimetric observation.

#3-5
The paper is of great interest in a wider field than planetary science providing new knowl-594 edge to our understanding of light scattering processes. The paper is worth to be published after 595 moderate revision. 596 We again appreciate that Reviewer #3 highly evaluates the scientific value of our study. In 597 this revision we made as much effort as we could for improving and clarifying the presentation of 598 the manuscript. We hope the revised manuscript is now better ready for a broader community 599 of astronomical sciences.  we obtain a pretty low albedo value of p = 0.081. In Figure 2 of the revised manuscript, we 90 placed two more symbols for Phaethon using these albedo estimates (0.081 and 0.10. Note that 91 the albedo values have been converted into A in this figure).

92
As for the uncertainty of absolute magnitude determination using of the H-G magnitude 93 system, it is known that it can reach 0.1 magnitude even for asteroids whose phase curve is 94 measured down to very small solar phase angle (Belskaya and Shevchenko, 2000 with the above two possibilities, we also placed a description on the potentially large surface 109 porosity of Phaethon as the third possibility that can enhance its polarization degree.

110
The above listed three causes (lower albedo, prevalence of larger grains, and/or large surface should be further investigated in future observational studies.

138
Avoiding the use of "coarse"/"fine" in the manuscript 139 Let us say that, we suspect that our use of the term "coarse" has perhaps invoked uninten-140 tional confusions among reviewers. On the path of making this revision, we came into attention 141 that the term "coarse" is often used in various different meanings, depending on who uses it.

142
For example, we found that the grain size of 100 µm is regarded both as being "coarse" (e.g. in different communities. In some cases, use of the term "coarse" automatically indicates that 145 the grain size is in the order of millimeter or even larger (e.g. Gundlach and Blum, 2013). This is different from what we intend. In order to suppress this kind of confusion, in this revision we 147 decided to avoid the use of the terms "coarse" and "fine" as much as possible except in com-148 monly used phrases such as "grain coarsening". Instead, we use more general terms ("large" 149 or "small") for describing grain size. Also, we have tried to explicitly state that we are dealing 150 with the surface grain size of Phaethon that can be "larger" than the well-studied lunar and 151 terrestrial samples. deg. 162 We are afraid that here is a little bit of misunderstanding. By the cited sentence "In this strongly as Pmax". It is not true. The negative branch parameters strongly depend on surface depends on asteroid's taxonomy means that surface texture are very similar for asteroids of the 181 same taxonomic class.

182
Thank you very much for pointing out. As we mentioned in the reply to the comment II 183 #3-2, we correctly understand this point now, and the description is entirely rewritten. Please 184 see the 2nd paragraph in p. 10 in the revised manuscript about the modified description.

186
Finally, let us thank again all the reviewers (including Reviewers #1 and #2) and the editors 187 for their great patience and the length of time that they spent for reviewing our manuscript.

III #3-1
The authors took into account all my comments. I do not have any other comments and recommend the paper to publication.
We appreciate very much that Reviewer #3 approved the significance of our work.
Let us thank once again all the three reviewers and the responsible editor for their great patience and a large amount of time that they spent for reviewing our manuscript. Their detailed and constructive comments suggested directions that significantly improved the quality of this paper. Thank you very much!