A new electrodynamic balance design for low temperature studies

• Users may freely distribute the URL that is used to identify this publication. • Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private study or non-commercial research. • User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?) • Users may not further distribute the material nor use it for the purposes of commercial gain.

qualitatively discussed but no measured data (ideally at various set temperatures) are given.
Response: We performed additional measurements and determined the temperature profile around the null point of the trap (this is now line 5-21 of page 9 and table 1 on page 27).We now provide more information about the measured temperature gradients within the cell (see Table 1 within the revised paper).
Referee comments 4: The main drawback of the chosen design is the use of an integrated liquid nitrogen cold trap for the precooling of the gas ï ň Ćow, which inevitably results in an extremely dry gas ï ň Ćow around the droplet.This excludes experiments at environmentally relevant conditions and leads to a rapid evaporation of the droplet even at low temperatures.This effect is aggravated by the fact that the droplet evaporates not into a stagnant atmosphere, but into a gas ï ň Ćow.The authors mention this fact brieï ň Ćy but do not discuss its implications or possible remedies.
Response: The first version of the instrument presented in this publication was only tested for dry conditions using liquid nitrogen as coolant.We mention now the possibility to perform experiments at higher RH in future studies (p. 9, line 27-32).However, it should be noted that one benefit of using the liquid N2 and the resultant low RH is that it removes an experimental uncertainty that is often present in ice nucleation studies what is the environmental RH.Within the CEDB using liquid nitrogen it is clear that the RH is very dry.
Referee comments 5: A rapidly evaporating droplet will assume a temperature that is lower than its surrounding.This effect is neither discussed in the manuscript nor is it considered in the following experiments on evaporation and immersion freezing.
Response: We now address the effect of evaporation on the temperature of the droplets.(26-30 lines of page 10, 1-6 line of page 11).
Referee comments 6: Even though I understand that the experiments described are C4235 to illustrate the capabilities of the EDB, I am missing important information, which the authors should provide: Evaporation rates of water droplets: How do the measured evaporation curves compare to theoretical expectations?
Response: To model the evaporation rates one needs a value for the accommodation coefficient (alpha) of water vapour into liquid water.This value is highly controversial within the literature with alpha values reported within the last decade varying between ∼0.1-1 (Eames et al., 1997, Li et al., 2001, Winkler et al., 2004, Davies et al., 2014).This factor of ten presents a large uncertainty for the modelling of cloud water.Whilst it is possible for us to use the theoretical model of Davies et al. (2014) to calculate alpha from our measurements (and we have) we intend to perform more experiments over a wider range of conditions before we publish this result, hence the mention of the forthcoming paper.We believe that publication of this value before these further experiments have been performed would be premature and could further muddy the water in this controversial arena.We further justify this approach by pointing out that we choose AMT for this initial paper because it was primarily to highlight our new CEDB approach not to provide new data.However, we are mindful that the paper will benefit from more useful data that the community can utilize.Towards this aim we have much extended the amount of data presented on the freezing of droplets.Referee comments 7: Why is the time for evaporation proportional to the radius?

Response:
According to the Maxwell equation the radius change rate of a SWD is proportional to the radius reciprocal of it.Under the same environment, larger SWD will have smaller radius decay rate.In other words, the life time of larger droplet will be longer.This is now explicitly mentioned on p. 11, line 25-27.

Referee comments 8: What does this tell us about the ï ň Ćow characteristics?
Response: As explained in response to the above comment the evaporation time depends on the initial particle size.Flow conditions were constant over all experiments reported here as described in section 2.1 Referee comments 9: How do the measured evaporation rates compare to droplet evaporation into a stagnant atmosphere of zero RH? Response: We did not investigate the evaporation into a stagnant atmosphere as described on p.7, line 2-10.Referee comments 10: What are the expectations based on the literature on evaporation into an laminar gas ï ň Ćow?
Response: The corrections due to Stefan flow are negligible, we now state this in the manuscript.
Referee comments 11: Immersion freezing of birch pollen washing water: How do the evaporative cooling and the continuous volume reduction affect the freezing rate?
Response: The reviewer correctly points out that evaporation will affect the droplet temperature.However, in PWW droplets where freezing was observed, this happened within 0.3 seconds after injecting the PWW droplets into the trap.In this short time the droplets do not evaporate significantly, see Figure 4a.In a very small number of PWW droplets (< 5%) freezing occurred after 0.3 seconds but these droplets were not considered in the data analysis for Figure 7 and 8 (this is now explicitly mentioned on p. 14. lines of 9-15).Referee comments 12: How were the ice fractions in Fig. 7 determined?Response: 50-158 droplets were analysed at each temperature.The number ratio between frozen to total (frozen + unfrozen) PWW droplets was calculated as ice fraction.At very high and low temperatures, less PWW droplets were analysed.(more details are given now at 2-7 lines of page 14).

C4237
Referee comments 13: Why are there no vertical error bars?
Response: The fraction of particle freezing at a particular temperature (i.e. the y-axis in Figure 7) was calculated as mention above (equation3, p. 13).Thus it is not possible to determine an error bar for ice fraction.