|Review by Minghui Diao|
This manuscript uses a combined data set from MOZAIC and IAGOS programs to analyze a series of atmospheric conditions in the upper troposphere and lower stratosphere (UT/LS). These conditions include temperature, relative humidity with respect to ice (RHice), water vapor, and ozone. In addition, the relationship of RHi, water vapor volume mixing ratio (H2O VMR) and the occurrence of ice supersaturated regions (ISSR) are analyzed in relation to several tropopause height, i.e., thermal tropopause, dynamic tropopause and chemical tropopause. Long-term time series are used to analyze the trend of ISSR occurrence frequencies. The analysis is focused on three regions – Eastern North America, North Atlantic, and Europe. Overall, the results shown in this study are consistent with previous MOZAIC data, satellite observations of ice cloud occurrence frequency, and research aircraft observations from other field campaigns. The long-term measurements from MOZAIC and IAGOS provide a valuable data set that can be used to compare with reanalysis data, such as ERA-interim.
Overall, the manuscript is well written, especially after addressing the comments from the other three reviewers. The reviewer has some comments below. The manuscript can be considered for publication in ACP after addressing these comments.
Line numbers refer to the clean manuscript without tracked changes, called “acp-2019-735-manuscript-version3”.
1. For the analysis of RHice versus temperature in Figure 4, the figure cropped off some of the higher RHi. Please include the entire RH data set unless there is a reason against that. In addition, there are some measurements of RHice above liquid saturation line. Can the authors comment on the percentage of the RHice data that is over the liquid saturation line? In Section 2.2 RH and O3 instrumentation, the RH measurements have been discussed for their uncertainties, which are 4% and 5% for the middle troposphere and tropopause, respectively. However, Figure 4 shows much higher RHice data than liquid saturation line, such as 40%. Can the authors comment on whether these very high RHi data above liquid saturation are artifacts that are not real? And if so, what is the implication to the interpretation of the RHi accuracy? For example, does this suggest that even with the rigorous method applied to process RH data by comparing with the limit of detection plus 5 ppmv (as stated in Section 2.3), there still are some RHi data that have large biases? Any suggestion on the atmospheric conditions that these large biases usually occur?
In line 342, there is a statement of “Obviously, RHice observations remain inside the physical boundaries led by the water saturation line and the line for homogeneous ice nucleation.” This is incorrect. Suggest revising to: [X% and Y% of] RHice observations remain inside the physical boundaries led by the water saturation line and the line for homogeneous ice nucleation[, respectively.]
2. A related suggestion about Figure 4 is to add another sub-panel to Figure 4, which plots RHice versus temperature in scatter plot, and color coded by H2O VMR (ppmv). Please refer to Diao et al. (2014, ACP) figure 5. The observations of RHice in Diao et al. (2014) showed that the H2O VMR of most ISS observations (99 %) are above 20 ppmv. This new figure will help to illustrate the magnitude of H2O VMR for the high RHice seen in the current Figure 4.
3. It is interesting that Figure 13 shows similar frequency of ISSR and satellite-based ice cloud fraction. Some of these satellite retrievals are at much coarser horizontal scales than 1 km. Can the author comment on why the different spatial scales do not seem to affect the similarity between the aircraft and satellite data?
4. It would be helpful to add a sub-panel figure to Figure 1 for number of samples of various temperatures (such as binned by 1 K) for the MOZAIC and IAGOS data used in this study. This new figure will complement the current Figure 1, which only showed the horizontal distribution of samples. Can the authors add precision of accuracy of temperature measurements into Section 2.2? Is vertical velocity measurement available from MOZAIC and IAGOS programs, such as from the aircraft avionic system? They would be helpful for analysis of updrafts and their impacts on the ISSR occurrences.
Several places in the introduction and result sections could benefit from more comparisons with previous studies. The reviewer suggests adding a few references that are highly relevant to analysis of ice supersaturated regions. Brackets show where new text is added.
Line 94, “As a result, these air parcels are both colder and of higher relative humidity than the embedded sub-saturated atmosphere.” This is not always the case that ISSRs are colder than their horizontally adjacent sub-saturated air. Suggest revising this sentence to:
“As a result, these air parcels are [generally] both colder and of higher relative humidity than the embedded sub-saturated atmosphere. [Using research aircraft observations from 87N to 67S, Diao et al. (2014) (their Figure 7) showed that 73% of the ISSRs have both lower temperature and higher H2O VMR than their horizontally adjacent sub-saturated air, while 27% of the ISSRs show higher temperature and higher H2O VMR than their surroundings.]”
Diao, M., M.A. Zondlo, A.J. Heymsfield, L.M. Avallone, M.E. Paige, S.P. Beaton, T. Campos and D.C. Rogers. “Cloud-scale ice supersaturated regions spatially correlate with high water vapor heterogeneities”, Atmospheric Chemistry and Physics, 14, 2639-2656, 2014.
Line 97, “In the northern mid-latitudes, ISSR occurrence coincides strongly with the storm tracks over the North Atlantic (…), [on the anticyclonic side of the polar jet stream (Diao et al., 2015), and inside the anvil cirrus clouds (D’Alessandro et al. 2017).]”
Diao, M., J.B. Jensen, L.L. Pan, C.R. Homeyer, S. Honomichl, J.F. Bresch and A. Bansemer. “Distributions of ice supersaturation and ice crystals from airborne observations in relation to upper tropospheric dynamical boundaries”, Journal of Geophysical Research: Atmosphere, 120, 5101–5121. doi: 10.1002/2015JD023139, 2015.
D'Alessandro, J. J., M. Diao, C. Wu, X. Liu, M. Chen, H. Morrison, T. Eidhammer, J.B. Jensen, A. Bansemer, M.A. Zondlo, J.P. DiGangi. Dynamical conditions of ice supersaturation and ice nucleation in convective systems: a comparative analysis between in-situ aircraft observations and WRF simulations, Journal of Geophysical Research: Atmosphere, 122, doi:10.1002/2016JD025994, 2017.
Line 110, “occurs in most cases below the thermal tropopause … [In addition, research aircraft observations over North America shows that most of the clear-sky ISSRs are located within +/- 500 m from the thermal tropopause (Diao et al., 2015 their figure 10).] This is the same reference as above (doi: 10.1002/2015JD023139).
Line 135, “However, the vertical resolution provided by space-borne instruments in the Ex-UTLS is very limited… [In addition, satellite observations such as NASA AIRS data contain biases in temperature and water vapor retrievals compared with aircraft observations by 1 – 2 Kelvin and 30%-40% of H2O VRM, respectively (Diao et al., 2013)].” Adding this sentence helps to show that it is not only the resolution problem but also the accuracy of satellite data that can limit the usage of retrievals of RHice.
Diao, M., L. Jumbam, J. Sheffield, E. Wood and M.A. Zondlo. “Validation of AIRS/AMSU-A water vapor and temperature data with in situ aircraft observations from surface to UT/LS at 87°N–67°S”, Journal of Geophysical Research: Atmospheres, 118 (12), 6816–6836, 2013.
Line 471, “Thus, these air parcels are known as both colder and more humid than the embedding sub-saturated air masses.” As mentioned above, this is not always the case. Suggest revising to: Thus, these air parcels are [generally] both colder and more humid than the [surrounding] sub-saturated air masses (Gierens et al., 1999; Spichtinger et al., 2003b)[, although 27% of them were found to be warmer than the surroundings (Diao et al., 2014). Same reference as above for Diao et al. (2014): doi.org/10.5194/acp-14-2639-2014
Line 486, at the end of this paragraph, add “Previously, using the CO-O3 tracer correlation, the majority (69%) of clear-sky ISSRs were found within the transition layer of the extratropical tropopause, while the rest was located below the transition layer (Diao et al., 2015 their figure 14).” This helps to corporates the finding in this paper that most ISSRs were found to locate at the chemical tropopause defined by O3 = 120 ppbv.
Line 626, “… approx. 80% of the observed ice-supersaturation events are associated with in-cloud conditions. [An analysis of 11 hours of ISSR observations from flight campaigns that did not specifically target cirrus clouds showed that 25% ISSRs were associated with in-cloud conditions (Diao et al., 2015).]” There may be more cirrus clouds being targeted in the campaigns used in Kramer et al. (2016), and therefore citing a different study that did not target on cirrus clouds can be helpful.
Line 813, “… both in number and strength of supersaturation. [Accurately representing the magnitude of ISSR as well as its coexistence with ice crystals are crucial for quantifying radiative forcing, since mistakenly representing ISSR as ice crystals can lead to an average decrease of 2.7 W/m2 in surface radiation (Tan et al., 2016).” Adding this sentence helps to reinforce the importance of representing ISSR accurately in reanalysis or model data.
Tan, X., Y. Huang, M. Diao, A. Bansemer, M. A. Zondlo, J. P. DiGangi, R. Volkamer, and Y. Hu. An assessment of the radiative effects of ice supersaturation based on in situ observations, Geophysical Research Letter, 43, 11,039–11,047, doi:10.1002/2016GL071144, 2016.
Other minor comments:
Line 151, a few places still use the terminology of “UTH” instead of RHice, such as line 428 and 430. Please change all UTH to RHice to be consistent.
Line 202, ERA-I (0:75 degree * 0:75 degree), do you mean 0.75 instead of 0:75?
Line 230, “The conversion to RHice uses the equations by Sonntag (1994).” Many other studies use the equation of saturation pressure with respect to ice (es_ice) from Murphy and Koop (2005). Please provide the differences in RHice calculations between Sonntag (1994) and Murphy and Koop (2005) for the temperature range analyzed in this study.
Murphy, D. M. and Koop, T.: Review of the vapour pressures of ice and supercooled water for atmospheric applications, Q. J. Roy. Meteorol. Soc., 131(608), 1539–1565, doi:10.1256/qj.04.94, 2005.
Line 311, “The difference between… determines the sensor offset voltage…” Can the authors elaborate on whether a constant voltage offset is applied to these 15 consecutive flights, or the voltage offset is a function sensor temperature? Figure 2 seems to show that the voltage offset is a function of temperature.
Line 358, MOAZIC should be MOZAIC.
Line 427, Focussing should be Focusing.
Line 491. “…, indicated by O3 VMR = 120 ppbv at delta_p_TPH = 0 hPa [for ISSRs]”. Add “for ISSRs” because the non-ISSRs show O3 VMR = 80 ppbv at thermal tropopause.
Line 680-681, The terminologies of TOVS and CALIPSO have not been defined.
Figure 3b legend still says w/t IFC, should be “w/o” IFC.