The authors presented results of pollen hygroscopic behavior obtained from an acoustic levitation technique. From the two pollen grains study, using either deposited method or levitation method, area and diameter changes were presented, and compared with results from previous studies that were mainly done with mass measurements. The authors concluded that although the direct imaging methods might suffer from stability of the levitated pollen grains, they offer direct observation on the shape dynamics of the pollen grains during their hygroscopic growth. The technique itself is quite interesting in measuring hygroscopic behaviors of large particles such as pollen grains. The manuscript is also well written. I do, however, have a major concern on how to present the observed results, and would recommend Major Revision before publication

Main:

• The authors claimed that the main advantage of the acoustic levitation technique is that it can provide direct imaging on the shape change of the pollen grain during hygroscopic growth or shrinkage. Yet, there is little information on such an advantage, i.e., either images at different RH for the same pollen grain, or a parameter to show the shape “factor”.
• Related to the point above, in addition to the “area ratio” and “diameter ratio”, would the aspect ratio be a good parameter to show that the pollen grains are becoming more and more spherical after taking up more water, as stated in the text?
• It seems from the surface-fixed results that front lit results were clearer (Table 2 and Figure 3). How come it was not used in the acoustic levitation method later on?

Minor:

• L220-230: Fig. 2 should be Fig. 3?
• Could the authors comment on how potential lateral motion during acoustic levitation would affect the determination of area and diameter ratios?
• Figures 3-6 need further modification to font size bigger.
• Is there a way to obtain some estimates of the uncertainties for the area ratios and diameter ratios, such that readers can appreciate what changes can be understood as significant?
• How was the diameter ratio defined (i.e., normalized to what)? Why is it always less than unity?

Review of “Acoustic levitation of pollen and visualisation of hygroscopic behaviour” by Sophie Mills et al., submitted to AMT

Mills et al. present an acoustic levitation technique and the results of hygroscopic water uptake experiments of two types of pollen grains, from Oriental Lily and from Cottonwood (roughly 30 to 100 microns in diameter). The commercial acoustic levitator is coupled to a humidity-controlled air flow, and the RH used ranged from dry to 95%. While the technique and its application to primary biological particles are interesting to the atmospheric science community.

However, there are some issues with the framing of the study. More precision and substantiation are needed in the introductory sketch of the state of the science. Further, the technique is not able to measure water uptake by pollen grains with enough precision and enough repeatability for the results to be conclusive. The technique does not represent an improvement to existing methods, and therefore the publication of this technique is in doubt.

Comments on the introductory text

The claim that discussion of pollen as CCN in the literature is sustained or increasing should be more carefully supported. The importance of pollen in the atmosphere is not limited to impacts on CCN, perhaps other impacts should be emphasized. (Line 8-9, Abstract (“Pollen are hygroscopic and so have the potential to act as cloud condensation nuclei (CCN) in the atmosphere. This could have yet uncertain implications for cloud processes and climate.”), line 61-62, Introduction (“While there have been increasing discussions in the literature postulating the significance pollen may have for atmospheric cloud processes and climate, …”)).

The impact of giant pollen CCN on cloud droplet number or supersaturation should be considered in more detail by the authors before being suggested. (Lines 55-59, Introduction)

The claim that acoustic levitation has progressed significantly in recent years, and that this is a good way to study aerosol, should be substantiated by citation to recent papers (Lines 114-116).

The assumption that pollen grains can restructure when humidified should be substantiated and discussed here in the context of the findings presented, as the implications of pollen restructuring are broad (Line 250)

Comments on the technique

The accuracy of the measurements is low, as noted by the authors and as evident in the spread in the area and diameter ratios displayed in Figure 6 (see, e.g., line 247-249 (“While these results show some evidence of the hygroscopic size increase we had expected with increasing RH, they also suggest apparent inaccuracy and lack of consistency which must be considered. This variability may be due to limitations of the method in terms of imaging capabilities.”), Line 271 (“yet the change may not be considered conclusive.”); line 275 (“it would be necessary to conduct more experiment instances before making conclusive assertions from these results”); line 284 (“The results for visual size changes of the Lilium orientalis pollen while levitated are somewhat inconclusive.”); line 287 (“generally also did not show a conclusive trend.”); line 293 (“These results suggest that the measurement accuracy is hindered by the fact that the grains are being levitated freely.”); Line 305 (“It should be noted that the images for this pollen grain even among the same RH increment displayed considerable variability, implying that the observed orientation was not fixed.”); lines 306-308 (“Due to the instability of the levitated pollen grain, it was difficult to capture consistent snapshots of the grain even at constant RH. This can also be observed as the data points themselves for area and diameter ratio display considerable variance indicating a large error margin.”)).

It is assumed in the calculations that the grains are always the same distance from the camera, even though migration toward and away from the camera has been noted. The uncertainty in size due to positioning should be quantified. (Line 293-295 (“The significant but unexpected changes in observed area may be largely affected by their orientation while suspended, as there is no way to guarantee we are always looking at the same angle of the pollen grain.”)).

Comments on results

Perhaps more data should be collected. (Line 299 (“it was very difficult to collect a complete set of image data across all previous humidity increments”)).

The hygroscopicity derived from the reported results falls within the wide range for pollen reported by the literature. However, hygroscopicity is an intensive property of fully dissolved molecules. The correct term to apply in pollen studies, assuming some insoluble fraction, is the “apparent hygroscopicity.” The wet pollen grain is a mixture of soluble molecules and an insoluble part.

The figure text should be roughly the same size as the caption text after the figures are resized to their final publication-ready dimensions. As submitted, the figure fonts are about 50% as large as the figure caption font, meaning that the font size should be doubled and the plot area reduced accordingly. Some guidelines also suggest that data symbols appear similar in size to the fonts.

Mills et al. reported an interesting study that echos Robert Brown's experiment 200 years ago. Yet, in the present case, the instability of small pollen particles is a technical issue to be addressed. The scope of the present study fits with that of Atmos. Meas. Tech. The manuscript is well written and easy to follow. I recommend the publication of this article after the following technical concerns are satisfactorily addressed.

Technical:

To what extent does the acoustic field affect the thermodynamic properties of the air surrounding the levitated particles? When acoustic standing wave is formed inside the levitation chamber, I would assume that the pressure and density of the air near the levitated particles differ from that of ambient air. Yet the relative humidity (RH) was measured outside the acoustic field. This gives rise to two technique questions: First, does the measured ambient RH accurately reflect the true RH near the surface of the levitated particles in the acoustic field? Second, more importantly, how does the acoustic field affect the mass transport of airborne water molecules, such as their flux from the ambience to the surface of levitated pollen particles.

It is indeed difficult to measure the pollens' size accurately when they are vibrating. I encourage the authors to perform statistically analysis on the pollens' image in a more systematic manner. For example, one may calculate the pixel values (i.e., brightness or darkness of a pixel, hereafter, P) and then plot the pixel value P as a function of pixels' distance to the geometric center of the pollen (hereafter r). This P(r) distribution function may comprise the neccesary information to quantify (or, better, filter out) the blurriness owing to the vibration of pollen particles. Next, one may fit the P(r) function to invert a length scale parameter (hereafter L), a length which can be related to the known pollen size (hereafter, dp). For example, one may establish this L versus dp relationship at dry condition. This relationship can then serve as a calibration curve to invert the dp from L during the vary RH experiments. Try it out and see whether the uncertainty could be better constrained.

Presentation:

The author may consider placing scale bars next to the pollens' image.

Minor:

Line 29-39: "PBAPs have been found to consitute almost 25% of insoluble aerosol particles..." Is the 25% a mass fraction? 

Line 38-47: This paragraph discusses the size range and the aerodynamic properties of pollens. I am curious about the size parameters mentioned here. Are they aerodynamic diameters or phyiscal size? Please specify.

Line 48: "altitude of 3000m".  3km is better

Line 48-49: "considerable lengths of time with favorable meteorology" is vague. How long extactly do pollens reside in the troposphere? How is their lifetime compared with the characteristic time of cloud processes?

Line 55 "relatively small in number concentration" is again not concrete enough. What is the typical number concentration?