This study simulates the ASC during 2008 Mar 04-05 using the WRF model with prognostic IN and detailed ice nucleation scheme. According to the model results, we propose a microphysics mechanism for phase inversion in ASC (Figure 42). In a deposition-nucleation-preferred environment, ice distribution is not only affected by temperature and humidity but also by dust number concentration. In the mixed-phase layer, ice gravitational sedimentation removes dust particles and maintains a low ice number concentration, leading to a weak WBF process. Cloud droplets thus persist for a long time, and their capturing dust particles by collision can prevent dust from nucleating efficiently by deposition nucleation. As a result, ASC has a self-regulating microphysical cycle to create a vertical inversion in effective IN which helps to maintain the phase inversion structure.
Deposition nucleation, rather than immersion freezing as suggested by previous studies (de Boer et al. 2011; Prenni et al. 2009), is the main nucleation process in this case. Furthermore, de Boer et al. (2011) indicated that the ASC formed with liquid appearing first and then following by ice, which is a prerequisite for immersion freezing.
Yet, our simulations show that ice clouds appeared first, which can occur only with effective deposition nucleation. These discrepancies imply three issues.
Firstly, because of the limitation of the single point observation, whether liquid first formed locally or was formed later but advected into the observation site from other places is not sure. Secondly, the environments of different cases may favor different nucleation processes to lead to different conclusions. If immersion freezing is dominant (when cloud temperature is somewhat higher than that of the present case (cf. Figure 1) or when WBF is too strong so that the in-cloud water vapor becomes significantly lower
than water saturation), the process of freezing itself can reduce the liquid in addition to the WBF process; therefore, a persistent liquid supplement from cloud dynamics is still required to maintain phase inversion. If the environment favors deposition nucleation, the trap of dust in cloud droplets tends to reduce the chance of ice formation and thus can maintain phase inversion by itself. The role of cloud droplets is different in these two situations, even though both involved the presence of dust in liquid.
Thirdly, deposition nucleation may be overestimated in the assumption of externally mixed dust. Whether dust particles are internally or externally mixed in MERRA-2 is unknown, and we assumed them to be externally mixed in the CTL run. Dust particles in the Arctic are mainly from the Sahara desert or eastern China (Breider et al. 2014; Tanaka and Chiba 2006) and may be coated with soluble components during the long transportation from the source to the Arctic. In the case studied in Tsai et al. (2015), 80%
in mass and 40% in number of dust particles from the Loess Plateau and Gobi Deserts are polluted during the three-day transportation from the pollution source towards the Pacific Ocean. In this studied case, according to a backward trajectory analysis (Figure 43), dust particles are likely transported to domain 1 from the dessert source in eastern China a week before Mar 04, which provides enough time for them to be internally mixed. If dust particles are dirty and internally mixed, they will have larger κ and size, the number of CN will be reduced, and the ability of deposition nucleation will further be weakened as the solute component absorbs water and forms a solution coating on dust. Such internally mixed dust particles become more efficient CN but ineffective IN, and they can immerse into droplets more easily by activation. The larger κ is tested in K0.11; however, K0.11 may not be representative of the internally mixed dust situation because the change in κ is not associated with a change in the size of dust and the number of CN in this sensitivity test. Furthermore, it is assumed that deposition nucleation is not affected by the large κ.
Therefore, the effect of internally mixed dust on ASC and phase inversion requires further research.
Apart from the chemical components of dust, other uncertainties exist in the model.
One uncertainty is from the reanalysis meteorology field from FNL. The water vapor in FNL is much less than what the in situ observation shows (Figure 6), and this may lead to an underprediction of LWP and the thickness and persistence of the liquid-dominant layer, resulting in a less significant phase inversion structure in the results. Another uncertainty is that ice particle shape variation is not considered in the model. The assumption of spherical ice particles may cause not only an underestimation of ice growth by vapor deposition (Chen and Tsai 2016) but may also underestimate the radar reflectivity (Okamoto et al. 2010). Other uncertainties may come from the lack of some microphysics processes in the model, including adsorption activation of dust or soot into cloud drops or drop freezing by contacting with IN. The adsorption can increase the probability for the largely insoluble dust to become CN without too much supersaturation (Figure 27), which in effect decreases available IN and prevents such dust particles to nucleate by deposition nucleation. Classical contact freezing needs a high number concentration of IN more than 400 cm-3 to be efficient (Cotton and Field 2002); therefore, not including this process seems to be acceptable. However, Fan et al. (2009) suggested that droplet evaporation may enhance freezing through contact freezing inside out, which could be important to ice nucleation in ASC. Not including this process may underestimate the amount of ice in the water-undersaturated environment, which could happen under strong WBF conditions. Nevertheless, the contact nucleation process is highly uncertain due to a lack of laboratory studies. These details may be worthy of future investigation.
initial and boundary conditions for dust. The aerosol data in MERRA-2 is assimilated by observation data from satellites that provide aerosol optical depth only, and the vertical structure is retrieved from the Goddard Earth Observing System version 5 (GEOS-5) that may be uncertain without the correction from observation (Randles et al. 2016). In this study, the aerosol vertical structure may be critical to ASC glaciation and the phase inversion structure. The number concentration of dust particles must be within an order of magnitude above or below those provided by MERRA-2 to produce the phase inversion structure. This may be an indication that the MERRA-2 dust data are of good quality and suitable for simulating other ASC cases.
Since the chance to observe phase inversion in ASC is slightly less than 50% (Qiu et al. 2015), one may wonder if such a probability is associated with the variation of IN species and concentration in the Arctic. When the atmosphere is lack of dust, which seems to be likely in the Arctic region due to its far distance from the major deserts, the role of soot might be more important due to the closer vicinity to the major source of black carbon emission from Russia (Breider et al. 2014). IN species seems to play an important role in ASC, and not including realistic IN species cannot ensure the correctness of model results. Lastly, this study based on one single case to conclude the importance of dust in the ASC phase inversion. Future work may extend the model simulation to other cases that have different synoptic weather patterns for further confirmation of our proposed mechanism.
REFERENCE
Bigg, E. K., 1953: The supercooling of water. Proceedings of the Physical Society. Section B, 66, 688.
Bosilovich, M., R. Lucchesi, and M. Suarez, 2015: MERRA-2: File specification.
Breider, T. J., L. J. Mickley, D. J. Jacob, Q. Wang, J. A. Fisher, R. Y. W. Chang, and B.
Alexander, 2014: Annual distributions and sources of Arctic aerosol components, aerosol optical depth, and aerosol absorption. Journal of Geophysical Research:
Atmospheres, 119, 4107-4124.
Brooks, S. D., K. Suter, and L. Olivarez, 2014: Effects of chemical aging on the ice nucleation activity of soot and polycyclic aromatic hydrocarbon aerosols. The Journal of Physical Chemistry A, 118, 10036-10047.
Chen, J.-P., 1994: Theory of deliquescence and modified Köhler curves. Journal of the atmospheric sciences, 51, 3505-3516.
Chen, J.-P., and T.-C. Tsai, 2016: Triple-moment modal parameterization for the adaptive growth habit of pristine ice crystals. Journal of the Atmospheric Sciences, 73, 2105-2122.
Chen, J.-P., A. Hazra, and Z. Levin, 2008: Parameterizing ice nucleation rates using contact angle and activation energy derived from laboratory data. Atmospheric Chemistry and Physics, 8, 7431-7449.
Cheng, C.-T., W.-C. Wang, and J.-P. Chen, 2010: Simulation of the effects of increasing cloud condensation nuclei on mixed-phase clouds and precipitation of a front system.
Atmospheric Research, 96, 461-476.
Cotton, R., and P. Field, 2002: Ice nucleation characteristics of an isolated wave cloud.
Quarterly Journal of the Royal Meteorological Society: A journal of the atmospheric sciences, applied meteorology and physical oceanography, 128, 2417-2437.
Curry, J. A., J. L. Schramm, W. B. Rossow, and D. Randall, 1996: Overview of Arctic cloud and radiation characteristics. Journal of Climate, 9, 1731-1764.
de Boer, G., T. Hashino, and G. J. Tripoli, 2010: Ice nucleation through immersion freezing in mixed-phase stratiform clouds: Theory and numerical simulations.
Atmospheric Research, 96, 315-324.
de Boer, G., H. Morrison, M. Shupe, and R. Hildner, 2011: Evidence of liquid dependent ice nucleation in high‐latitude stratiform clouds from surface remote sensors.
Geophysical Research Letters, 38.
DeMott, P. J., 1990: An exploratory study of ice nucleation by soot aerosols. Journal of applied meteorology, 29, 1072-1079.
Diehl, K., and S. Wurzler, 2004: Heterogeneous drop freezing in the immersion mode:
Model calculations considering soluble and insoluble particles in the drops. Journal of the atmospheric sciences, 61, 2063-2072.
Dusek, U., and Coauthors, 2010: Enhanced organic mass fraction and decreased hygroscopicity of cloud condensation nuclei (CCN) during new particle formation events. Geophysical Research Letters, 37.
Fan, J., M. Ovtchinnikov, J. M. Comstock, S. A. McFarlane, and A. Khain, 2009: Ice formation in Arctic mixed‐phase clouds: Insights from a 3‐D cloud‐resolving model with size‐resolved aerosol and cloud microphysics. Journal of Geophysical Research: Atmospheres, 114.
Herich, H., and Coauthors, 2009: Water uptake of clay and desert dust aerosol particles at sub-and supersaturated water vapor conditions. Physical Chemistry Chemical Physics, 11, 7804-7809.
Hoose, C., J. E. Kristjánsson, J.-P. Chen, and A. Hazra, 2010: A classical-theory-based parameterization of heterogeneous ice nucleation by mineral dust, soot, and biological particles in a global climate model. Journal of the Atmospheric Sciences, 67, 2483-2503.
Hung, H.-M., K.-C. Wang, and J.-P. Chen, 2015: Adsorption of nitrogen and water vapor by insoluble particles and the implication on cloud condensation nuclei activity.
Journal of Aerosol Science, 86, 24-31.
Konovalov, I. B., and Coauthors, 2018: Estimation of black carbon emissions from Siberian fires using satellite observations of absorption and extinction optical depths.
Atmospheric Chemistry and Physics, 18, 14889-14924.
Lin, Y. C., 2015: Numerical Simulation of Dust-Cloud Interactions Using Regional Models, Department of Atmospheric Sciences, National Taiwan University, 1-218 pp.
Liu, J., S. Fan, L. W. Horowitz, and H. Levy, 2011: Evaluation of factors controlling long‐ range transport of black carbon to the Arctic. Journal of Geophysical Research:
Atmospheres, 116.
Lupi, L., and V. Molinero, 2014: Does hydrophilicity of carbon particles improve their ice nucleation ability? The Journal of Physical Chemistry A, 118, 7330-7337.
Meyers, M. P., P. J. DeMott, and W. R. Cotton, 1992: New primary ice-nucleation parameterizations in an explicit cloud model. Journal of Applied Meteorology, 31, 708-721.
Morrison, H., and A. Gettelman, 2008: A new two-moment bulk stratiform cloud microphysics scheme in the Community Atmosphere Model, version 3 (CAM3).
Part I: Description and numerical tests. Journal of Climate, 21, 3642-3659.
development of trailing stratiform precipitation in a simulated squall line:
Comparison of one-and two-moment schemes. Monthly weather review, 137, 991-1007.
Morrison, H., G. de Boer, G. Feingold, J. Harrington, M. D. Shupe, and K. Sulia, 2012:
Resilience of persistent Arctic mixed-phase clouds. Nature Geoscience, 5, 11.
Okamoto, H., K. Sato, and Y. Hagihara, 2010: Global analysis of ice microphysics from CloudSat and CALIPSO: Incorporation of specular reflection in lidar signals.
Journal of Geophysical Research: Atmospheres, 115.
Paukert, M., and C. Hoose, 2014: Modeling immersion freezing with aerosol‐dependent prognostic ice nuclei in Arctic mixed‐phase clouds. Journal of Geophysical Research: Atmospheres, 119, 9073-9092.
Peng, J., and Coauthors, 2017: Ageing and hygroscopicity variation of black carbon particles in Beijing measured by a quasi-atmospheric aerosol evolution study (QUALITY) chamber. Atmospheric Chemistry and Physics, 17, 10333.
Petters, M., and S. Kreidenweis, 2007: A single parameter representation of hygroscopic growth and cloud condensation nucleus activity.
Prenni, A. J., P. J. Demott, D. C. Rogers, S. M. Kreidenweis, G. M. Mcfarquhar, G. Zhang, and M. R. Poellot, 2009: Ice nuclei characteristics from M-PACE and their relation to ice formation in clouds. Tellus B: Chemical and Physical Meteorology, 61, 436-448.
Prenni, A. J., and Coauthors, 2007: Can ice-nucleating aerosols affect Arctic seasonal climate? Bulletin of the American Meteorological Society, 88, 541-550.
Pruppacher, H. R., and J. D. Klett, 1980: Microphysics of clouds and precipitation. Nature, 284, 88-88.
Qiu, S., X. Dong, B. Xi, and J. L. Li, 2015: Characterizing Arctic mixed‐phase cloud
structure and its relationship with humidity and temperature inversion using ARM NSA observations. Journal of Geophysical Research: Atmospheres, 120, 7737-7746.
Randles, C., and Coauthors, 2016: Technical Report Series on Global Modeling and Data Assimilation.
Sedlar, J., M. D. Shupe, and M. Tjernström, 2012: On the relationship between thermodynamic structure and cloud top, and its climate significance in the Arctic.
Journal of Climate, 25, 2374-2393.
Shupe, M. D., and J. M. Intrieri, 2004: Cloud radiative forcing of the Arctic surface: The influence of cloud properties, surface albedo, and solar zenith angle. Journal of Climate, 17, 616-628.
Shupe, M. D., S. Y. Matrosov, and T. Uttal, 2006: Arctic mixed-phase cloud properties derived from surface-based sensors at SHEBA. Journal of the atmospheric sciences, 63, 697-711.
Shupe, M. D., and Coauthors, 2008: A focus on mixed-phase clouds: The status of ground-based observational methods. Bulletin of the American Meteorological Society, 89, 1549-1562.
Solomon, A., M. D. Shupe, P. Persson, and H. Morrison, 2011: Moisture and dynamical interactions maintaining decoupled Arctic mixed-phase stratocumulus in the presence of a humidity inversion. Atmospheric Chemistry & Physics Discussions, 11.
Sorjamaa, R., and A. Laaksonen, 2007: The effect of H 2 O adsorption on cloud-drop activation of insoluble particles: a theoretical framework.
Tanaka, T. Y., and M. Chiba, 2006: A numerical study of the contributions of dust source regions to the global dust budget. Global and Planetary Change, 52, 88-104.
investigation of the coagulation mixing between dust and hygroscopic aerosol particles and its impacts. Journal of Geophysical Research: Atmospheres, 120, 4213-4233.
Whitby, K. T., 1978: The physical characteristics of sulfur aerosols. Sulfur in the Atmosphere, Elsevier, 135-159.
TABLES
Table 1. Settings of WRF V3.3.1 real case simulation.
Time control and Domain settings
run hour (hr) 36 (2008 Mar 04 00UTC – Mar 05 12UTC)
max domain 3
interval second (s) 21600 (6hr)
eta level 51
resolution (km) 27 9 3
time step (s) 150 30 6
Physics and Dynamics options
microphysics CLR scheme [11]
longwave radiation RRTM scheme [1]
shortwave radiation Dudhia scheme [1]
surface layer Monin-Obukhov scheme [1]
land-surface Pleim-Xiu scheme [7]
boundary-layer YSU scheme [1]
cumulus X [0]
Table 2. Aerosol settings in MERRA-2 and hygroscopicity used in this study. re is the effective radius, ρ is the particle density, and κ is the hygroscopicity. κ are not included in MERRA-2, and obtained from Petters and Kreidenweis (2007) for sea salt and sulfate, Peng et al. (2017) for hydrophilic BC, Dusek et al. (2010) for hydrophilic OC, Herich et al. (2009) for dust and assumed no κ for hydrophobic particles.
bin re ρ
Table 3. Ice nucleation parameter settings adopted from Hoose et al. (2010). θ is the contact angle for calculation of the geometric factor f in equation (1) and Δg# represents the desorption energy for deposition nucleation and the activation energy for immersion freezing.
𝛉𝛉 (°) 𝚫𝚫𝚫𝚫# (10-20 J)
dust deposition nucleation 12.70 -0.621
immersion freezing 30.98 15.7
soot deposition nucleation 28.00 -20.0
immersion freezing 40.17 14.4
Table 4. Experiment design of sensitivity tests.
Experiment Type Experiment Name Note
Control dust (CTL) immersion freezing + deposition nucleation, κ=0.001, control run of this study
IN soot immersion freezing + deposition nucleation
Heterogeneous
NOCOL no collision of dust to droplets
K0.11 κ=0.1125
FIGURES
Figure 1. Ice nucleation rate for dust and soot from classical nucleation theory without considering embryo size. Solid lines are for dust, and dotted lines are for soot. The color blue indicates immersion freezing, and other colors indicate deposition nucleation.
Figure 2. The fitted MERRA-2 dust lognormal size distributions. Black boxes are the 5 bins from MERRA-2, blue line represents the fitted accumulation mode, and orange line
represents the fitted coarse mode.
Figure 3. Cloud radar reflectivity profile of (a) MMCR, the (b) dust, and (c) soot runs from 00:00 UTC Mar 04 to 12:00 UTC Mar 05, 2008.
Figure 4. Time series of depolarization ratio (shading) observed by MPL at Barrow adopted from Qiu et al. (2015).
Figure 5. Temperature profile at Barrow from ARM (purple line) and WRF in the dust run (orange line) at 12:00 UTC on Mar 04, 2008.
Figure 6. Specific humidity profile at Barrow from ARM (purple line) and WRF in the dust run (orange line) at 12:00 UTC on Mar 04, 2008.
Figure 7. MODIS image on Mar 04, 2008. Star represents Barrow observation site.
Figure 8. Domain setting for the model simulations.
Figure 9. Distributions of (a) potential temperature at 850 hPa, and (b) height at 500 hPa in domain 1 in the dust run.
Figure 10. Time series of horizontal distributions of ice (shading) and cloud liquid content (contour: 0.0001~0.01 g m-3) in domain 3 at 800 hPa on Mar 04, 2008 in the dust run.
Star represents Barrow observation site.
Figure 11. Time series of horizontal distributions of ice (shading) and cloud liquid content (contour: 0.0001~0.01 g m-3) in domain 3 at 800 hPa on Mar 04, 2008 in the soot run.
Star represents Barrow observation site.
Figure 12. (a) NE-SW and (b) NW-SE cross-sections of ice (shading), cloud liquid content (black contour: 0.0001~0.01 g m-3), and temperature (red contour) indicated by the red lines in Figure 10 at 15:00 UTC on Mar 04, 2008 in the dust run.
Figure 13. (a) NE-SW and (b) NW-SE cross-sections of ice (shading), cloud liquid content (black contour: 0.0001~0.01 g m-3), and temperature (red contour) indicated by the red lines in Figure 11 at 15:00 UTC on Mar 04, 2008 in the soot run.
Figure 14. Distributions of wind at 850 hPa in the (a) dust and (b) soot runs, and (c) difference of the soot run from the dust run in domain 1 at 12:00 UTC on Mar 04, 2008.
Shading is wind speed in (a) and (b) and wind speed difference in (c).
Figure 15. Time series of average (a) IWP and (b) LWP in domain 3 of the dust and soot runs during 12:00-17:30 UTC on Mar 04, 2008.
Figure 16. Number concentration of IN in ice (shading) and cloud liquid content (contour:
0.0001~0.01 g m-3) at the NE-SW cross-section indicated by the long red line in Figure 10 at 15:00 UTC on Mar 04, 2008 for the (a) dust and (b) soot runs.
Figure 17. Time series of average IN in ice in domain 3 of the (a) dust and (b) soot runs during 12:00-18:00 UTC on Mar 04, 2008. Black contour represents 10-8 g m-3 of liquid.
Figure 18. Time series of LWP and TWP. ARM observed LWP at Barrow.
Dust and soot model results are averaged within 5x5 grids area centered at Barrow.
Figure 19. Time series of liquid (a) mass mixing ratio fraction and (b) scattering area ratio averaged within 5x5 grids area centered at Barrow of the dust run.
Figure 20. Time series of liquid (a) mass mixing ratio fraction and (b) scattering area ratio averaged within 5x5 grids area centered at Barrow of the soot run.
Figure 21. Horizontal distributions of ice in (a) and (b), and ice number concentration in (c) and (d) in domain 3. (a) and (c) are for the dust run, and (b) and (d) are for the soot run. Black contours represent liquid in the range of 0.0001~0.01 g m-3. Time is at 00:00 UTC on Mar 05, 2008.
Figure 22. Horizontal distributions of ice (shading) and cloud liquid content (contour:
0.0001~0.01 g m-3) in domain 3 of the (a) OFF, (b) DE, and (c) IM runs at 800 hPa at 15:00 UTC on Mar 04, 2008.
Figure 23. Ice (shading) and cloud liquid content (contour: 0.0001~0.01 g m-3) at the NE-SW cross-section indicated by the long red line in Figure 10 at 15:00 UTC on Mar 04, 2008 for the (a) OFF, (b) DE, and (c) IM runs.
Figure 24. Time series of average (a) IWP and (b) LWP in domain 3 of heterogeneous
Figure 25. Number concentration of IN in liquid (shading) and cloud liquid content (contour: 0.0001~0.01 g m-3) at the NE-SW cross-section indicated by the long red line in Figure 10 at 15:00 UTC on Mar 04, 2008 for the (a) CTL, (b) DE, and (c) IM runs.
Figure 26. Time series of average the number concentration of IN in liquid in domain 3 of the (a) CTL, (b) DE, and (c) IM runs during 12:00-18:00 UTC on Mar 04, 2008. Black contour represents 10-8 g m-3 of liquid.
Figure 27. Critical saturation for the deliquescence activation of κ equaling to 0.001 (orange line) and 0.1125 (green line), and the adsorption activation of dust (blue line) at -15⁰C. Critical saturation of adsorption is calculated from the Frankel-Halsey-Hill (FHH) isotherm with A=1.25 and B=1.33 (Hung et al. 2015).
Figure 28. Number concentration of IN and cloud liquid content (contour: 0.0001~0.01 g m-3) in liquid (shading) at the NE-SW cross-section indicated by the long red line in Figure 10 at 15:00 UTC on Mar 04, 2008 for the (a) CTL, (b) NOCOL, and (c) K0.11 runs.
Figure 29. Time series of average the number concentration of IN in liquid in domain 3 of the (a) CTL, (b) NOCOL, and (c) K0.11 runs during 12:00-18:00 UTC on Mar 04, 2008. Black contour represents 10-8 g m-3 of liquid.
Figure 30. Time series of average (a) IWP and (b) LWP in domain 3 of dust immersion process sensitivity tests during 12:00-17:30 UTC on Mar 04, 2008.
Figure 31. Horizontal distribution of vertical velocity (shading) and cloud liquid content (contour: 0.0001~0.01 g m-3) at 800 hPa in the CTL run. Time is at 13:30 UTC on Mar 04, 2008, when liquid first formed.
Figure 32. Cloud liquid content (contour: 0.0001~0.01 g m-3) and (a) WBF characteristic time τ and (b) ice number concentration (shading) at the NE-SW cross-section indicated by the long red line in Figure 10 at 15:00 UTC on Mar 04, 2008 in the CTL run.
Figure 33. Probability of WBF characteristic time τ in the CTL run during 12:00-17:30
Figure 33. Probability of WBF characteristic time τ in the CTL run during 12:00-17:30