• Andreae, M. O., , D. Rosenfeld, , P. Artaxo, , A. A. Costa, , G. P. Frank, , K. M. Longo, , and M. A. F. Silva-Dias, 2004: Smoking rain clouds over the Amazon. Science, 303, 13371342.

    • Search Google Scholar
    • Export Citation
  • Baumgardner, D., , H. Jonsson, , W. Dawson, , D. O’Connor, , and R. Newton, 2001: The cloud, aerosol and precipitation spectrometer: A new instrument for cloud investigations. Atmos. Res., 59–60, 251264.

    • Search Google Scholar
    • Export Citation
  • Brenguier, J. L., , and W. Grabowski, 1993: Cumulus entrainment and cloud droplet spectra: A numerical model within a two-dimensional dynamical framework. J. Atmos. Sci., 50, 120136.

    • Search Google Scholar
    • Export Citation
  • Erlick, C., , A. Khain, , M. Pinsky, , and Y. Segal, 2005: The effect of wind velocity fluctuations on drop spectrum broadening in stratiform clouds. Atmos. Res., 75, 1545.

    • Search Google Scholar
    • Export Citation
  • Freud, E., , D. Rosenfeld, , M. O. Andreae, , A. A. Costa, , and P. Artaxo, 2008: Robust relations between CCN and the vertical evolution of cloud drop size distribution in deep convective clouds. Atmos. Chem. Phys., 8, 16611675.

    • Search Google Scholar
    • Export Citation
  • Fridlind, A., and Coauthors, 2004: Evidence for the predominance of mid-tropospheric aerosols as subtropical anvil cloud nuclei. Science, 304, 718722.

    • Search Google Scholar
    • Export Citation
  • Hallett, J., , and S. C. Mossop, 1974: Production of secondary ice crystals during the riming process. Nature, 249, 2628.

  • Heymsfield, A. J., 2007: On measurements of small ice particles in clouds. Geophys. Res. Lett., 34, L23812, doi:10.1029/2007GL030951.

  • Heymsfield, A. J., , A. Bansemer, , G. Heymsfield, , and A. O. Fierro, 2009: Microphysics of maritime tropical convective updrafts at temperatures from −20° to −60°C. J. Atmos. Sci., 66, 35303562.

    • Search Google Scholar
    • Export Citation
  • Hobbs, P.V. 1993: Aerosol-Cloud-Climate Interactions. Academic Press, 236 pp.

  • Houze, R. A., , and D. D. Churchill, 1984: Microphysical structure of winter monsoon cloud clusters. J. Atmos. Sci., 41, 34053411.

  • Jaenicke, R., 1993: Tropospheric aerosols. Aerosol–Cloud–Climate Interactions, P. Hobbs, Ed., Academic Press, 1–31.

  • Jensen, E. J., and Coauthors, 2009: On the importance of small ice crystals in tropical anvil cirrus. Atmos. Chem. Phys. Discuss., 9, 53215370.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., 2009: Notes on state-of-the-art investigations of aerosol effects on precipitation: A critical review. Environ. Res. Lett., 4, 015004, doi:10.1088/1748-9326/4/1/015004.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., , and A. Pokrovsky, 2004: Simulation of effects of atmospheric aerosols on deep turbulent convective clouds using a spectral microphysics mixed-phase cumulus cloud model. Part II: Sensitivity study. J. Atmos. Sci., 61, 29833001.

    • Search Google Scholar
    • Export Citation
  • Kogan, Y., 1991: The simulation of a convective cloud in a 3D model with explicit microphysics. Part I: Model description and sensitivity experiments. J. Atmos. Sci., 48, 11601189.

    • Search Google Scholar
    • Export Citation
  • Korolev, A. V., 1994: A study of bimodal droplet size distributions in stratiform clouds. Atmos. Res., 32, 143170.

  • Korolev, A. V., 1995: The influence of supersaturation fluctuations on droplet size spectra formation. J. Atmos. Sci., 52, 36203634.

  • Korolev, A. V., 2007: Limitations of the Wegener–Bergeron–Findeisen mechanism in the evolution of mixed-phase clouds. J. Atmos. Sci., 64, 33723375.

    • Search Google Scholar
    • Export Citation
  • Korolev, A. V., , and I. Mazin, 2003: Supersaturation of water vapor in clouds. J. Atmos. Sci., 60, 29572974.

  • Korolev, A. V., , E. F. Emery, , J. W. Strapp, , S. G. Cober, , G. A. Isaac, , M. Wasey, , and D. Marcotte, 2011: Small ice particles in tropospheric clouds: Fact or artifact? Airborne Icing Instrumentation Evaluation Experiment. Bull. Amer. Meteor. Soc., in press.

    • Search Google Scholar
    • Export Citation
  • Lasher-Trapp, S. G., , W. A. Cooper, , and A. M. Blyth, 2005: Broadening of droplet size distributions from entrainment and mixing in a cumulus cloud. Quart. J. Roy. Meteor. Soc., 131, 195220.

    • Search Google Scholar
    • Export Citation
  • Levin, Z., , and W. R. Cotton, Eds., 2009: Aerosol Pollution Impact on Precipitation: A Scientific Review. Springer, 386 pp.

  • Low, T. B., , and R. List, 1982: Collision, coalescence and breakup of raindrops, Part II: Parameterization of fragment size distributions. J. Atmos. Sci., 39, 16071619.

    • Search Google Scholar
    • Export Citation
  • Ludlam, F. H., 1980: Clouds and Storms. The Pennsylvania State University Press, 405 pp.

  • Magaritz, L., , M. Pinsky, , O. Krasnov, , and A. Khain, 2009: Investigation of droplet size distributions and drizzle formation using a new trajectory ensemble model. Part II: Lucky parcels. J. Atmos. Sci., 66, 781805.

    • Search Google Scholar
    • Export Citation
  • McFarquhar, G. M., , J. Um, , M. Freer, , D. Baumgardner, , G. L. Kok, , and G. Mace, 2007: Importance of small ice crystals to cirrus properties: Observations from the Tropical Warm Pool International Cloud Experiment (TWP-ICE). Geophys. Res. Lett., 34, L13803, doi:10.1029/2007GL029865.

    • Search Google Scholar
    • Export Citation
  • Ochs, H. T., 1978: Moment-conserving techniques for warm cloud microphysical computations. Part II: Model testing and results. J. Atmos. Sci., 35, 19591973.

    • Search Google Scholar
    • Export Citation
  • Paluch, I. R., , and C. A. Knight, 1984: Mixing and the evolution of cloud droplet size spectra in a vigorous continental cumulus. J. Atmos. Sci., 41, 18011815.

    • Search Google Scholar
    • Export Citation
  • Phillips, V. T. J., and Coauthors, 2005: Anvil glaciation in a deep cumulus updraft over Florida simulated with an explicit microphysics model. I: The impact of various nucleation processes. Quart. J. Roy. Meteor. Soc., 131, 20192046.

    • Search Google Scholar
    • Export Citation
  • Pinsky, M., , and A. P. Khain, 2002: Effects of in-cloud nucleation and turbulence on droplet spectrum formation in cumulus clouds. Quart. J. Roy. Meteor. Soc., 128, 501533.

    • Search Google Scholar
    • Export Citation
  • Politovich, M. K., 1993: A study of the broadening of droplet size distributions in cumuli. J. Atmos. Sci., 50, 22302244.

  • Politovich, M. K., , and W. A. Cooper, 1998: Variability of the supersaturation in cumulus clouds. J. Atmos. Sci., 45, 16511664.

  • Pruppacher, H. R., , and J. D. Klett, 1997: Microphysics of Clouds and Precipitation. 2nd ed. Kluwer Academic, 954 pp.

  • Reutter, P., and Coauthors, 2009: Aerosol- and updraft-limited regimes of cloud droplet formation: Influence of particle number, size and hygroscopicity on the activation of cloud condensational nuclei (CCN). Atmos. Chem. Phys., 9, 70677080.

    • Search Google Scholar
    • Export Citation
  • Rosenfeld, D., , and I. M. Lensky, 1998: Satellite-based insights into precipitation formation processes in continental and maritime convective clouds. Bull. Amer. Meteor. Soc., 79, 24572476.

    • Search Google Scholar
    • Export Citation
  • Rosenfeld, D., , U. Lohmann, , G. B. Raga, , C. D. O’Dowd, , M. Kulmala, , S. Fuzzi, , A. Reissell, , and M. O. Andreae, 2008: Flood or drought: How do aerosols affect precipitation? Science, 321, 13091313.

    • Search Google Scholar
    • Export Citation
  • Segal, Y., , M. Pinsky A. Khain, , and C. Erlick, 2003: Thermodynamic factors influencing bimodal spectrum formation in cumulus clouds. Atmos. Res., 66, 4364.

    • Search Google Scholar
    • Export Citation
  • Song, N., , and J. Marwitz, 1989: A numerical study of the warm rain process in orographic clouds. J. Atmos. Sci., 46, 34793486.

  • Squires, P., 1952: The growth of cloud drops by condensation. I. General characteristics. Aust. J. Sci. Res., 5, 5986.

  • Straub, W., , K. D. Beheng, , A. Seifert, , J. Schlottke, , and B. Weigang, 2010: Numerical investigation of collision-induced breakup of raindrops. Part II: Parameterizations of coalescence efficiencies and fragment size distributions. J. Atmos. Sci., 67, 576587.

    • Search Google Scholar
    • Export Citation
  • Su, C.-W., , S. K. Krueger, , P. A. McMurtry, , and P. H. Austin, 1998: Linear eddy modeling of droplet spectral evolution during entrainment and mixing in cumulus clouds. Atmos. Res., 47–48, 4158.

    • Search Google Scholar
    • Export Citation
  • Warner, J., 1969a: The microstructure of cumulus cloud. Part I: General features of the droplet spectrum. J. Atmos. Sci., 26, 10491059.

    • Search Google Scholar
    • Export Citation
  • Warner, J., 1969b: The microstructure of cumulus cloud. Part II: The effect of droplet size distribution of the cloud nucleus spectrum and updraft velocity. J. Atmos. Sci., 26, 12721282.

    • Search Google Scholar
    • Export Citation
  • Warner, J., 1973: The microstructure of cumulus cloud. Part IV: The effect on the droplet spectrum of mixing between cloud and environment. J. Atmos. Sci., 30, 256261.

    • Search Google Scholar
    • Export Citation
  • Xue, L., , A. Teller, , R. Rasmussen, , I. Geresdi, , and Z. Pan, 2010: Effects of aerosol solubility and regeneration on warm-phase orographic clouds and precipitation simulated by a detailed bin microphysical scheme. J. Atmos. Sci., 67, 33363354.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) CALIPSO aerosol total attenuated backscatter showing clouds and aerosols. White patches are clouds and yellow or red are aerosols. (b) Location of the measurements on Indian satellite Kalpana map; line is CALIPSO pass. (c) Location of clouds in relation to HYD; color scale is measurement height (m) above surface.

  • View in gallery

    The vertical soundings from 16 (solid) and 22 Jun (dashed). Two smooth dashed lines (labeled 16 and 22) indicate moist adiabats for the observation dates. Winds are plotted with barbs for 22 Jun.

  • View in gallery

    Vertical profiles of radiosonde (a) horizontal wind speed (bars; top x axis) and relative humidity (lines; bottom x axis), (b) PCASP aerosol concentration, and (c) CCN concentration for premonsoon (thin solid), transition (dashed), and monsoon (thick solid) case [note the logarithmic scale in (b),(c)] determined at <1% supersaturation. Error bars are standard deviation of measurements representing horizontal variations.

  • View in gallery

    CCN activity spectra measured below the cloud base for (a) premonsoon, (b) transition, and (c) monsoon cases. Each circle in the plot corresponds to a CCN measurement below the cloud base at approximately 100-m intervals.

  • View in gallery

    Vertical variation of horizontally averaged (a) cloud water content (CWC) and (b) reff in clouds observed on 16, 21, and 22 Jun. Bar diagrams in (a) represent maximum CWC on 16 (black), 21 (gray), and 22 Jun (white hatched).

  • View in gallery

    Microphysical parameters of clouds measured along airplane track on (a) 16, (b) 21, and (c) 22 Jun. Each symbol represents averaging over 1-s time increment. CWC (g m−3) is color coded; reff is given by size of symbols varying between 2 and 16 μm.

  • View in gallery

    Average DSD observed for three CAIPEEX flights during monsoon transition, on (a) 16, (b) 21, and (c) 22 Jun. Legends correspond to time (UTC; HHMMSS) and height (meters) of observations. For example, 95551.7085 indicates 0955:51 UTC at 7085 m above the surface.

  • View in gallery

    Vertical distribution of bin-averaged [in the D bins <10 μm (dashed–dotted), <20 μm (thick), 20–40 μm (thin), and 40–50 μm (dashed)] number concentration at 100-m height intervals for three CAIPEEX flights on (a) 16, (b) 21, and (c) 22 Jun. Vertical profile of the droplet spectrum dispersion (line with open circles) is also presented. All data are screened for adiabatic fraction > 0.02.

  • View in gallery

    Relationship between quasi-steady supersaturation and W for concentrations of droplets with D < 10 μm (color scale) and at different heights (size of symbols) for (a) 16, (b) 21, and (c) 22 Jun; all data screened for adiabatic fraction of 0.02.

  • View in gallery

    Relationship between mean aerosol concentrations below cloud base and bin-average number concentration of droplets (CDNC) with D < 20 μm for CAIPEEX observations. CDNC is averaged at 1–3 km and 4–6 km above cloud base from the drop size distribution sampled at 1-s intervals. Boxed symbols correspond to two premonsoon highly polluted cases.

  • View in gallery

    Plots of 1-Hz droplet size distributions in cloud core of monsoon cloud (22 Jun 2009) at (a) 2.8, (b) 3.1, and (c) 5.1 km. Legend shows time, height, cloud droplet concentration, effective radius, vertical velocity, Sqs, and adiabatic fraction. A representative CIP image is shown alongside each figure.

  • View in gallery

    Examples of bimodal and multimodal DSDs in premonsoon cloud at (a) 6.1 and (b) 5.4 km and transition cloud at (c) 5.7 and (d) 5 km where in-cloud nucleation in the strong updrafts is noticed. Legend as in Fig. 11.

  • View in gallery

    As in Fig. 11, but at (a) 4.4 and (b) 6.7 km in the presence of downdrafts.

  • View in gallery

    (a),(b) DSDs along a flight pass length from (a) 21 Jun, indicating in-cloud nucleation leading to small drops and (b) 22 Jun, indicating in-cloud nucleation contributing to broadening of DSDs. Pass length averaged DSD is bimodal as shown. (c),(d) CIP images show clear indications of mixed phase (ice, rain, and cloud droplets) on (c) 21 and (d) 22 Jun.

  • View in gallery

    The conceptual scheme of droplet spectra formation in deep convective clouds [after Pinsky and Khain (2002) with changes for monsoon clouds].

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 229 229 18
PDF Downloads 187 187 14

Microphysics of Premonsoon and Monsoon Clouds as Seen from In Situ Measurements during the Cloud Aerosol Interaction and Precipitation Enhancement Experiment (CAIPEEX)

View More View Less
  • 1 Indian Institute of Tropical Meteorology, Pune, India
  • | 2 Department of Atmospheric Sciences, The Hebrew University of Jerusalem, Jerusalem, Israel
  • | 3 Indian Institute of Tropical Meteorology, Pune, India
© Get Permissions
Full access

Abstract

Analysis of the microphysical structure of deep convective clouds using in situ measurements during the Cloud Aerosol Interaction and Precipitation Enhancement Experiment (CAIPEEX) over the Indian peninsular region is presented. It is shown that droplet size distributions (DSDs) in highly polluted premonsoon clouds are substantially narrower than DSDs in less polluted monsoon clouds. High values of DSD dispersion (0.3–0.6) and its vertical variation in the transient and monsoon clouds are related largely to the existence of small cloud droplets with diameters less than 10 μm, which were found at nearly all levels. This finding indicates the existence of a continuous generation of the smallest droplets at different heights. In some cases this generation of small droplets leads to the formation of bimodal and even multimodal DSDs. The formation of bimodal DSDs is especially pronounced in monsoon clouds. Observational evidence is presented to suggest that in-cloud nucleation at elevated layers is a fundamental mechanism for producing multimodal drop size distribution in monsoon clouds as well as in most deep convective clouds. These findings indicate that inclusion of continued nucleation away from the cloud base into numerical models should be considered to predict microphysics and precipitation of clouds in monsoons and other cloud-related phenomena.

Corresponding author address: Thara V. Prabha, Indian Institute of Tropical Meteorology, Dr. Homi Bhabha Road, Pashan, Pune 411008, India. E-mail: thara@tropmet.res.in

Abstract

Analysis of the microphysical structure of deep convective clouds using in situ measurements during the Cloud Aerosol Interaction and Precipitation Enhancement Experiment (CAIPEEX) over the Indian peninsular region is presented. It is shown that droplet size distributions (DSDs) in highly polluted premonsoon clouds are substantially narrower than DSDs in less polluted monsoon clouds. High values of DSD dispersion (0.3–0.6) and its vertical variation in the transient and monsoon clouds are related largely to the existence of small cloud droplets with diameters less than 10 μm, which were found at nearly all levels. This finding indicates the existence of a continuous generation of the smallest droplets at different heights. In some cases this generation of small droplets leads to the formation of bimodal and even multimodal DSDs. The formation of bimodal DSDs is especially pronounced in monsoon clouds. Observational evidence is presented to suggest that in-cloud nucleation at elevated layers is a fundamental mechanism for producing multimodal drop size distribution in monsoon clouds as well as in most deep convective clouds. These findings indicate that inclusion of continued nucleation away from the cloud base into numerical models should be considered to predict microphysics and precipitation of clouds in monsoons and other cloud-related phenomena.

Corresponding author address: Thara V. Prabha, Indian Institute of Tropical Meteorology, Dr. Homi Bhabha Road, Pashan, Pune 411008, India. E-mail: thara@tropmet.res.in

1. Introduction

A major experiment named the Cloud Aerosol Interaction and Precipitation Enhancement Experiment (CAIPEEX) is underway in India [conducted by the Indian Institute of Tropical Meteorology (IITM); see http://www.tropmet.res.in/~caipeex/] as an attempt to identify and understand the pathways through which aerosols may influence precipitation. Observations of convective clouds, aerosol, and cloud condensation nuclei (CCN) at several locations were carried out during this experiment with an instrumented aircraft during May–September 2009. Thus, clouds during premonsoon, transition-to-monsoon, and active monsoon situations were sampled at various locations. In situ measurements were performed in clouds at heights up to 7 km above the surface. This is significantly higher than most earlier in situ cloud measurements other than the Monsoon Experiment (MONEX) in the South China Sea (Houze and Churchill 1984) and other experiments reported over ocean pertaining to deep layer monsoon convection such as the National Aeronautic and Space Administration (NASA) African Monsoon Multidisciplinary Analysis (NAMMA) and the Aerosol and Chemical Transport in Tropical Convection Experiment (ACTIVE), (Heymsfield et al. 2009). These observations are unique over the tropical continental region. For comparison, the maximum height at which detailed measurements of drop size distribution (DSD) in the Amazon region during the Large-Scale Biosphere–Atmosphere Experiment in Amazonia (LBA) and Smoke Aerosols, Clouds, Rainfall, and Climate (SMOCC) field experiment were performed at 4.2 km above the surface.

Premonsoon clouds develop in dry conditions, and, as a result, cloud-base height exceeds 2 km. Monsoon clouds develop in a very humid atmosphere, so cloud-base height is generally well below 2 km. Premonsoon clouds develop in the extremely polluted atmosphere. In spite of the fact that concentration of aerosol particles (APs) in the atmosphere during monsoon period is quite high (see below, in section 3), their concentration is significantly lower than that in premonsoon clouds. Comparison of microphysical structure of premonsoon and monsoon clouds is of great importance for better understanding the role of both humidity and aerosols with regard to clouds and raindrop formation. In monsoon clouds, the measurements were performed at heights up to 5 km above cloud base. These deep vertical layer observations above the cloud base over continental areas had never before been available and thus the measurements in CAIPEEX are unique for the investigation of processes of DSD formation.

The purpose of this study is to analyze the microphysical structure of premonsoon, transition, and monsoon clouds developing in different thermodynamic and aerosol environments using in situ observations in CAIPEEX. The remainder of this article is organized as follows. In section 2, detailed descriptions of the cases considered in the present study are provided; cloud microphysical characteristics of premonsoon, transition-to-monsoon, and active monsoon situations are described. The microphysical structure of clouds as seen from the in situ measurements is described in section 3. Discussion of the results is presented in section 4, and section 5 summarizes the findings.

2. Case studies and equipment used

During the transition from premonsoon to monsoon, measurements were made over the Indian peninsular region with flights organized from Hyderabad (HYD; 17.45°N, 78.46°E) (Fig. 1). The Hyderabad mission registered the onset of the monsoon over that location on 21 June. Cloud DSDs were measured, along with three wind velocity components, temperature, relative humidity, CCN concentration (using a CCN counter), aerosols [size distribution, effective radius, and concentration using a Passive Cavity Aerosol Spectrometer Probe (PCASP)] and cloud parameters [droplet concentration and effective radius, cloud liquid water content, and drop size distribution using a cloud droplet probe (CDP)] onboard the CAIPEEX aircraft. The in situ data were obtained with a temporal resolution of 1 s. Parameters used in this study are listed in Table 1 along with the details of instrument make, range, and resolution.

Fig. 1.
Fig. 1.

(a) CALIPSO aerosol total attenuated backscatter showing clouds and aerosols. White patches are clouds and yellow or red are aerosols. (b) Location of the measurements on Indian satellite Kalpana map; line is CALIPSO pass. (c) Location of clouds in relation to HYD; color scale is measurement height (m) above surface.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

Table 1.

List of instruments on board the CAIPEEX aircraft with data sampling details. CIP is the cloud imaging probe, DMT is Droplet Measurement Technologies, and AIMMS is the Advanced Airborne Measurement Solutions air data probe.

Table 1.

Observations of premonsoon, transition, and monsoon convective clouds made on 16, 21, and 22 June 2009 with the CAIPEEX aircraft over Hyderabad are considered for detailed discussion. These three cloud observations were carried out centered at 17.25°N, 79°E; 19°N, 78°E; and 16.25°N, 77.3°E. Both upward and downward profiling of the cloud was done on 16 June with several in-cloud penetrations at different heights z and at 500–600 m deep inside the convective cores where droplet concentrations of 300–500 cm−3 were noted. However, concentrations of 1000 cm−3 were also noticed close to the cloud base. Observations on 21 and 22 June were carried out with downward profiling of the cloud. To isolate the convective cloud, observations carried out within a 50 × 50 km2 area with an adiabatic fraction (ratio of adiabatic liquid water and cloud liquid water) below 0.02 are screened out and excluded from the analysis presented in this study. This threshold is chosen by rigorous analysis of data of different clouds observed during CAIPEEX. Low adiabatic fraction could be used as a measure of mixing in the clouds, which also indicates pockets of isolated clouds that may not be part of the main convective cloud. This criterion filtered out elevated nonconvective clouds containing largely small droplets, and data from less diluted convective clouds were only used in the analysis. Cloud observations on 16 and 21 June characterized convective clouds at developing stages. Observations of 22 June indicated a maturing cloud.

Monsoon onset in Hyderabad was 21 June. CAIPEEX flight observations were carried out over an area of 150 × 150 km2 around Hyderabad in the northeastern, northern, and western sectors during the transition period. The flight report indicated rapid growth of clouds. Strong updrafts exceeding 3–6 m s−1 below cloud base and 10 m s−1 inside cloud were characteristic of these cases. Figure 1 shows the 22 June clouds as seen using Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) (Fig. 1a) and Indian satellite Kalpana (Fig. 1b) images. There was a CALIPSO pass at 0815 UTC on this day. The flight observations were around 0750–1015 UTC. Figure 1a shows CALIPSO total attenuated backscatter intensity, indicating clouds and aerosols (white patches are clouds and yellow or red are aerosols). This cloud developed above a polluted boundary layer, but in relatively cleaner upper layers. There are more aerosols at northern latitudes and aerosol concentrations at upper levels depended on the transport from the polluted locations. The cloud top is at around 7 km. The aircraft observations were taken when the cloud matured at 0900 UTC. Cloud observations on 16, 21, and 22 June were performed over locations indicated in Fig. 1c. Precipitation rates [from Tropical Rainfall Measuring Mission (TRMM) 3B42RT over a 100 × 100 km2 area around the cloud locations] on 16, 21, and 22 June showed rain rates of 0.008, 0.03, and 0.508 mm h−1, respectively. Corresponding accumulated rainfall estimates are 0.02, 0.09, and 1.53 mm. The precipitation from 22 June cloud was noticed 3 h after the aircraft observation.

Radiosonde observations (Fig. 2) from 16 and 22 June at Hyderabad showed signatures of a transition from premonsoon (e.g., 16 June) to monsoon (e.g., 22 June) conditions. The convective available potential energy (CAPE) calculated from the radiosonde observations were 1677.8 and 2256.6 J kg−1 and integrated precipitable water was 4.3 and 4.39 cm for 16 and 22 June, respectively. The lifting condensation level (LCL) temperatures were 11.76° and 16.64°C at pressure levels of 707.2 and 775.4 mb for 16 and 22 June, respectively. A low-level westerly jet was established at 1.5 km above the surface (Figs. 2 and 3a) and the easterly jet at the 17-km level (Fig. 2), typical of the monsoon. Average air temperature near the surface varied from about 36.2°C on 16 June to 33°C on 22 June. The cloud-top temperatures were about −15° and −20°C, respectively. Average relative humidity in the boundary layer increased from 40% in the premonsoon conditions to 70% in the monsoon case because of the low-level moisture transport by the jet.

Fig. 2.
Fig. 2.

The vertical soundings from 16 (solid) and 22 Jun (dashed). Two smooth dashed lines (labeled 16 and 22) indicate moist adiabats for the observation dates. Winds are plotted with barbs for 22 Jun.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

Fig. 3.
Fig. 3.

Vertical profiles of radiosonde (a) horizontal wind speed (bars; top x axis) and relative humidity (lines; bottom x axis), (b) PCASP aerosol concentration, and (c) CCN concentration for premonsoon (thin solid), transition (dashed), and monsoon (thick solid) case [note the logarithmic scale in (b),(c)] determined at <1% supersaturation. Error bars are standard deviation of measurements representing horizontal variations.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

Aerosol concentration profiles from the PCASP for cloud filtered data are presented in Fig. 3b. The observations of AP and CCN were carried out during the CCN cycles below the cloud base. Aircraft ascended from HYD, which is in the middle of three observation points (Fig. 1c), on 16 June (in the southeastern sector), 21 June (in the northwestern sector) and 22 June (in the southwestern sector). Figure 3b showed a maximum concentration of 1000–2000 cm−3 at 500 m above the surface that reduced to 50 cm−3 at 6.5 km. The premonsoon case was characterized by AP concentrations exceeding 600 cm−3 throughout the lower 4.5-km layer. However, monsoon case is comparatively clean and high aerosol concentrations are noticed primarily below 2 km. It is to be noted that AP concentration decreases drastically above the boundary layer as the monsoon progresses. Note that the AP concentration shown in Fig. 3b does not include APs smaller than 0.1 μm. This explains the fact that CCN concentrations (evaluated at 0.4% of supersaturation) are much higher than AP concentrations (Fig. 3c).

CCN measurements at different supersaturations—0.2%, 0.4%, and 0.55% or 0.6%, corrected for pressure changes [the inlet pressure of CCN counter was kept at 500 hPa, and the SS drops at 0.07% (100 hPa)−1]—were done below the cloud base to estimate the CCN activity spectra from Twomey equation NCCN= NoSk, where No is the CCN concentration at 1% supersaturation and k is the slope parameter. The spectral estimates were carried out for all updrafts with temperature changes in the CCN chamber less than 0.15°C s−1. Estimated spectra from observations are presented in Fig. 4 for the premonsoon, transition, and monsoon cloud samples. Average CCN concentrations (estimated for 1% of supersaturation) from measured CCN spectra below the cloud base were 7268.5 cm−3 on 16 June, 4761 cm−3 on 21 June, and 1655.4 cm−3 on 22 June, again indicating highly polluted to less polluted conditions after the monsoon onset on 21 June. These CCN measurements are made at 1-Hz intervals before entering the cloud base. Estimated slope parameters for premonsoon (16 June), transition (21 June), and monsoon (22 June) clouds were larger than 1 and gradually decreased with a reduction in the pollution levels. Such values of slopes exceed those typically reported in measurements in both maritime and continental air masses (Pruppacher and Klett 1997). One can suppose that the values of the slope parameters decrease beyond the range of the supersaturation settings used in the experiment. Note that high values of the slope parameters indicate the existence of a significant amount of small CCN that can be activated at higher supersaturations. This point is important for the analysis presented below.

Fig. 4.
Fig. 4.

CCN activity spectra measured below the cloud base for (a) premonsoon, (b) transition, and (c) monsoon cases. Each circle in the plot corresponds to a CCN measurement below the cloud base at approximately 100-m intervals.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

3. Microphysical structure of premonsoon and monsoon clouds

a. Vertical variation of cloud microphysical parameters

Deep convective clouds profiled in the flight observations are considered for the analysis. Values of the horizontally averaged cloud water content (CWC; for droplet diameters less than 40 μm) measured in these flights are shown in Fig. 5a. Each symbol represents a CAIPEEX flight observation of a deep convective cloud. Increase in the air humidity led to a decrease in the cloud-base height from 2710 m on 16 June to 2450 m on 21 June to 1950 m on 22 June. At heights below 4.5 km, the averaged CWC in monsoon clouds reaches 1.5 g m−3 and exceeds those in premonsoon clouds and in clouds developed during the transition period. Above this level, CWC in the monsoon clouds decreases, while in premonsoon clouds CWC continues to increase with height. Above the 6-km level, CWC is larger in the premonsoon cloud than in the monsoon cloud. The CWC in premonsoon cloud reaches its maximum near 7 km, where the averaged value of CWC is approximately 2 g m−3 and the maximum CWC exceeded 3 g m−3. The cause of the difference in the CWC behavior in premonsoon and monsoon clouds can be derived from Fig. 5b, which shows the vertical variation of horizontally averaged values of effective radius 〈reff〉 in these cases. The value of 〈reff〉 in the monsoon clouds increases with height, much faster than in premonsoon clouds, and reaches 14 μm near 4.5 km. According to in situ measurements (e.g., Freud et al. 2008) and numerical simulations (e.g., Pinsky and Khain 2002; N. Benmoshe et al. 2011, unpublished manuscript), in clouds developing in a relatively clean atmosphere, efficient raindrop formation begins when the value of effective radius exceeds 13–14 μm and in clouds developing in very polluted air when 〈reff〉 reaches 10–11 μm. Accordingly, the decrease in the CWC above 4.5 km in monsoon clouds can be attributed to efficient warm rain above this level (observed in situ measurements). In premonsoon clouds 〈reff〉 reaches 10–11 μm at heights of 6.5–7 km, indicating significant delay in the first raindrop formation as compared to monsoon clouds. As noted by N. Benmoshe et al. (2011, unpublished manuscript), the lower threshold value of 〈reff〉 in polluted clouds can be explained by the fact that the value of 〈reff〉 is determined largely by the diffusion droplet growth, which is slow in these clouds because of high droplet concentration (and low supersaturation). Droplet collisions lead to formation of a low concentration tail of large cloud droplets, while most cloud droplets remained small. Accordingly, droplets ascend in cloud updrafts, so that the CWC increases in the premonsoon clouds up to heights of 7 km. The appearance of raindrop formation in monsoon clouds at height of 4.5 km is related to higher air humidity (i.e., lower cloud base level), largely, because of lower droplet concentration as compared to the premonsoon clouds.

Fig. 5.
Fig. 5.

Vertical variation of horizontally averaged (a) cloud water content (CWC) and (b) reff in clouds observed on 16, 21, and 22 Jun. Bar diagrams in (a) represent maximum CWC on 16 (black), 21 (gray), and 22 Jun (white hatched).

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

Figure 6 shows the detailed microphysical structure of the clouds during all three flights. In contrast to results shown in Fig. 5, each symbol in Fig. 6 shows 1-s averaged microphysical characteristics (corresponding to about 100-m intervals as aircraft flew at average speed of 100 m s−1). Only in-cloud observations with a CDP concentration greater than 20 cm−3 were used in the analysis. As mentioned earlier, only droplets with adiabatic fraction exceeding 0.02 within a 50 × 50 km2 area are considered to isolate the deep convective cloud observations. This additional screening is carried out to exclude any secondary cloud that may contaminate the analysis. Each symbol represents the following microphysical parameters: (a) vertical distribution of cloud droplet concentration (horizontal axis scale), (b) cloud liquid water content (g m−3; for droplet diameters D < 50 μm; color coded), and (c) effective radius reff (indicated by the size of symbols, varying between 2 and 16 μm. Several symbols of different sizes and colors at a specific height indicate that the aircraft was flying horizontally at same height, where different effective radius and cloud water content are measured.

Fig. 6.
Fig. 6.

Microphysical parameters of clouds measured along airplane track on (a) 16, (b) 21, and (c) 22 Jun. Each symbol represents averaging over 1-s time increment. CWC (g m−3) is color coded; reff is given by size of symbols varying between 2 and 16 μm.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

On 16 June, droplet spectra are characterized by higher droplet concentration (1000 cm−3) than on 22 June (200–300 cm−3). This is a direct indication of transition from highly polluted premonsoon continental to less polluted maritime monsoon conditions. Lower droplet concentration in monsoon clouds is associated with lower number of CCN available at the cloud base during the monsoon case compared to the premonsoon case.

One can see significant variability of CWC at each height level. The maximum values of CWC are of about 3 g m−3 in all three cases (shown in blue in Fig. 6). In monsoon clouds these high values of CWC are located in large droplets (reff calculated over the 100-m flight segments is about 16 μm), with concentration up to 200 cm−3 (Fig. 6c). In premonsoon clouds the large CWC is concentrated in intermediate size droplets (reff ≈ 12 μm, calculated over the 100-m flight segments) with concentration of about 900 cm−3 (Fig. 6a). It is interesting that in spite of the high variability of CWC, variability of local (i.e., 1-s averaged) reff at horizontal traverses is comparatively low (the size of the symbols in Fig. 6 at each height level does not change significantly). The low horizontal variability of reff in the presence of high variability of CWC was also reported by Freud et al. (2008) in the case of the LBA-SMOCC campaign in the Amazon region. The reason for such an effect requires special investigation.

At the same time, Fig. 6 shows the existence of the 100-m intervals in clouds where reff is below approximately 8 μm (i.e., significantly lower than the horizontally averaged values of reff). In these intervals the values of CWC are typically low. Most of the time, concentration of these smallest droplets in these spatial intervals is below 100 cm−3. These small droplets exist in premonsoon clouds at all height levels (see Fig. 6a). In the transient and monsoon clouds, the 100-m spatial intervals with very low reff were found at certain height levels (see Fig. 6). Note that even when reff is large, small droplets can exist in the spectra. But if reff is small, the existence of small droplets is obvious. Below we will show that very small droplets exist at most of height levels in the transient and monsoon clouds in spite of comparatively high values of horizontally averaged reff.

b. Vertical variation of drop size distribution functions

The most informative microphysical characteristics of clouds are the number concentration spectra (DSDs) and their evolution with height. Figures 7a–c show the combined DSD from a CDP and a cloud imaging probe (CIP) measured at different elevations for the 3 days (16, 21, and 22 June). CIP data were corrected for shattering of hydrometeors using interarrival time (Baumgardner et al. 2001) and for nonspherical particles based on algorithms provided by the manufacturers. Each DSD is obtained by averaging over flight pass lengths (we will show subsequently that the flight pass length averages conceal some very important aspects of DSD) ranging from 300 to 1000 m. As monsoon conditions progress, a gradual shift in the DSD toward higher droplet diameters is noticeable, indicating bigger droplets and higher cloud liquid water content in the droplets that grew beyond the limit of warm rain onset (D > 40–50 μm). The shift of the DSD maximum toward larger sizes on 22 June is clearly noted. DSDs in the monsoon clouds are wider and contain more large cloud droplets at lower levels as compared to the premonsoon clouds. The peak of the DSD in the monsoon clouds is located at diameters of about 50 μm as compared to 30 μm in premonsoon clouds (indicated by vertical lines in Fig. 7). The process of raindrop formation is much more efficient in the monsoon clouds due to collisions and coalescence. While in the premonsoon clouds, nonnegligible concentration of large droplets with D > 50 μm (obtained from CIP) appear at elevations of 7 km, raindrops appear at levels of 4–5 km in the monsoon case.

Fig. 7.
Fig. 7.

Average DSD observed for three CAIPEEX flights during monsoon transition, on (a) 16, (b) 21, and (c) 22 Jun. Legends correspond to time (UTC; HHMMSS) and height (meters) of observations. For example, 95551.7085 indicates 0955:51 UTC at 7085 m above the surface.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

A specific feature of all DSDs is the existence of small droplets with diameters lower than 15 μm at higher levels. In some cases these small droplets form a second mode in the DSDs; that is, bimodal (or even multimodal) DSDs are formed. Note that horizontal averaging can mask some fine features (say, the existence of bimodal DSD) because second peaks in the DSDs in flight pass lengths are centered at different drop sizes.

To investigate formation of DSD in more detail, 1-s DSDs were used for further analysis. These DSDs were used to integrate number concentration in the respective droplet diameter bins. The average number concentration is calculated in four drop diameter ranges (D < 10 μm, D < 20 μm, 20 < D < 40 μm, 40 < D < 50 μm). These ranges were chosen to follow the growth of droplets by diffusion and to reveal formation of large drops by collisions. Only CDP data were used in this analysis and CIP data (D > 50 μm) were not used. Figure 8 shows vertical profiles of DSD dispersion (where σ is the width of the spectrum and is the droplet mean radius), as well as number concentration of droplets within the chosen size ranges. One can see that DSD dispersions do not show any tendency to decrease with height (as should follow from the equation for the diffusion growth). Such behavior of DSD dispersion was reported in several studies (e.g., Politovich 1993), but in much weaker cumulus clouds with significantly lower cloud-top height. The DSD dispersion is about 0.32 in the premonsoon clouds. Dispersion of DSD in the monsoon cloud is highly variable (0.25–0.6) with height. At 6.5 km the DSD dispersion reaches 0.5–0.6 in monsoon (22 June) clouds. Since the mean droplet radius typically increases with height, as such the DSD dispersion behavior indicates DSD broadening with height in all cases.

Fig. 8.
Fig. 8.

Vertical distribution of bin-averaged [in the D bins <10 μm (dashed–dotted), <20 μm (thick), 20–40 μm (thin), and 40–50 μm (dashed)] number concentration at 100-m height intervals for three CAIPEEX flights on (a) 16, (b) 21, and (c) 22 Jun. Vertical profile of the droplet spectrum dispersion (line with open circles) is also presented. All data are screened for adiabatic fraction > 0.02.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

The DSD broadening is closely related to the formation of large drops and of the smallest droplets. The changes of dispersion are well correlated with changes of the concentration of small cloud droplets: the increase in the concentration of the smallest droplets leads to an increase in the DSD dispersion (e.g., Pinsky and Khain 2002). While the concentration of the smallest droplets indicates the tendency to decrease with height in premonsoon clouds, it is nearly constant above 3 km in monsoon clouds. Note that at some heights the concentration of the smallest droplets increases with height. For instance, a significant increase was observed at z > 6 km in monsoon clouds (Fig. 8c).

Observations also show that concentration of cloud droplets with diameters 20 < D < 40 μm and large droplets with 40 < D < 50 μm increases with height. Such an increase is caused by diffusional growth of smaller droplets and drop–drop collisions. These factors act as the sink of smallest droplets. The presence of the small droplets in the DSD at most height levels indicates the existence of the source of the small droplets inside of clouds. This is especially pronounced in monsoon clouds in which concentration of droplets in the three lowest size ranges remains nearly constant above 3 km, despite the continuous generation of the largest drops and small raindrops.

The question then arises: what is the mechanism responsible for appearance of small droplets at elevated layers?

In this study we present evidence of two mechanisms responsible for small droplets production and DSD broadening: in-cloud nucleation in cloud updrafts and cloud droplet evaporation in cloud downdrafts.

c. In-cloud nucleation in cloud updrafts

Several physical mechanisms were proposed to explain the formation of small droplets in clouds well above the cloud base, including mixing of cloudy volumes with dry environment near the lateral cloud boundaries (Warner 1973); partial evaporation of cloud droplets within undersaturation zones due to entrainment and mixing of relatively dry environmental air (Paluch and Knight 1984; Brenguier and Grabowski 1993; Su et al. 1998; Lasher-Trapp et al. 2005); and nucleation of new droplets on CCN penetrating a cloud through lateral cloud boundaries (e.g., Phillips et al. 2005; Fridlind et al. 2004).

Another mechanism of appearance of small droplets and formation of bimodal and multimodal DSD is in-cloud nucleation of droplets on CCN “penetrating a cloud through the cloud base and ascending together with cloud droplets formed at low levels” (Pinsky and Khain 2002, p. 502). Such in-cloud nucleation takes place when supersaturation within an ascending cloud parcel exceeds its local maximum near the cloud base (Ludlam 1980; Ochs 1978; Song and Marwitz 1989; Korolev 1994; Pinsky and Khain 2002; Segal et al. 2003; Phillips et al. 2005; Magaritz et al. 2009).

It is known that supersaturation is not measured in situ observations. However, in warm and mixed-phase clouds, in zones with significant liquid water content, supersaturation rapidly tends to the quasi-state value Sqs (Squires 1952; Korolev and Mazin 2003):
e1
where νm is the coefficient of molecular diffusion of water vapor, N is the droplet concentration, and W is the vertical velocity. The coefficient , where T is temperature, Ra and Rυ are the gas constants for air and water vapor, Lw is the latent heat of condensation, and g is gravitational acceleration. Thus, in these zones Sqs can be used for estimation of real supersaturation.

The supersaturation maximum near cloud base is typically below 1%. If W is high enough in cloud updraft, the supersaturation above the cloud base increases and may become higher than the supersaturation maximum near the cloud base leading to in-cloud nucleation and formation of the second mode in DSD consisting of small cloud droplets. In a cloud layer in which supersaturation continuously increases with height, in-cloud nucleation will also take place continuously. Equation (1) also indicates the possibility of another mechanism of the supersaturation increase: in the regions where droplet collisions are efficient, the concentration of droplets decreases. Besides fallout of raindrops through unloading, an increase in the vertical velocity also leads to an increase in supersaturation. These mechanisms may lead to in-cloud nucleation (Pinsky and Khain 2002; A. Khain et al. 2010, unpublished manuscript, hereafter KPPB).

1) Observational indications of in-cloud nucleation

Below we present observed data that can be interpreted as support for the existence of the in-cloud droplet nucleation on small CCN penetrating through cloud base. Note first that the concentration of aerosols surrounding the cloud rapidly decreases with height in the monsoon cases (Figs. 3b,c). Thus, it is unlikely that the lateral CCN penetration of AP is the major cause of the in-cloud nucleation above the 5-km level. Second, estimates show that the distance that aerosols can penetrate into the cloud through the lateral cloud boundary due to turbulent diffusion during the time of an air parcel ascent from cloud base to cloud top does not exceed several hundred meters (e.g., Khain and Pokrovsky 2004). At the same time, the increase in the concentration of small droplets was often observed within clouds at distances exceeding 500 m from cloud edges, quite close to the cloud core.

One necessary condition for in-cloud nucleation is comparatively high values of supersaturation. Figure 9 shows the relationship between Sqs [Eq. (1)] and W for concentrations of droplets with diameters below 10 μm (represented by color scale) and at different heights (represented by size of symbols). We concentrate first on the relationship between the cloud updrafts and positive Sqs in the figure. Some important features could be noted: there is considerable variation in Sqs along a flight track and at different elevations, with peaks up to 5%–8% in cloud updrafts. Such values significantly exceed the maximum (~0.5%) values of Sqs in comparatively small cumulus clouds measured during the Cooperative Convective Precipitation Experiment (CCOPE) (Politovich and Cooper 1998). These values are also larger than the maximum of 3% calculated for intermediate-size (5-km top height) cumulus clouds simulated using the cloud model with spectral bin microphysics (Kogan 1991). High peaks of Sqs were observed at different height levels, including the 6-km level. The values of Sqs calculated at levels close to cloud base were typically lower than 2%, but in some cases the peaks reached a few percent more. Note that the values of Sqs cannot be used for estimation of real supersaturation near cloud base (Korolev and Mazin 2003): in the vicinity of cloud base the mean drop radius is very small, and Sqs calculated using Eq. (1) is significantly higher than the real supersaturation (e.g., Politovich and Cooper 1998). Taking into account that the maximum values of droplet concentration measured in premonsoon and monsoon clouds were about 1000 and 400 cm−3, respectively (Fig. 6), and with the CCN activity spectra shown in Fig. 4, cloud-base supersaturations of these clouds can be evaluated as about 0.3% in premonsoon cloud and 0.4% in monsoon clouds, respectively. It means that peaks of supersaturation aloft exceeded the values of supersaturation at cloud base. It is apparent that the supersaturation peaks are associated with high values of W or there is decrease in the number concentration of droplets (high supersaturation is also noticed at locations where concentration is low, due to unloading of large drops).

Fig. 9.
Fig. 9.

Relationship between quasi-steady supersaturation and W for concentrations of droplets with D < 10 μm (color scale) and at different heights (size of symbols) for (a) 16, (b) 21, and (c) 22 Jun; all data screened for adiabatic fraction of 0.02.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

Supersaturations of 5%–8% can lead to activation of the smallest aerosols with diameters of soluble part of about 0.005 μm. According to the evaluations, cloud-base supersaturation did not exceed about 0.4%, so that aerosols with diameters of soluble fraction below about 0.005 μm remained unactivated at the level of cloud-base supersaturation maxima. Significant concentrations of such aerosols in both maritime and continental air masses were reported in many studies (e.g., Hobbs 1993; Jaenicke 1993; Pruppacher and Klett 1997).

Another necessary condition for in-cloud nucleation is the existence of a significant reservoir of small CCN. Figure 10 presents a scatter diagram showing the relationship between the cloud averaged concentration of droplets with D < 20 μm in the 1–3-km and 4–6-km layers above the cloud base and the aerosol concentration in the layer below the cloud base (excluding the entrainment layer of 200-m depth just below cloud base) for several CAIPEEX flight records of deep convective clouds. Each symbol represents a CAIPEEX flight observation of a deep convective cloud. The diagram also includes observations from premonsoon highly polluted cases (indicated with rectangular boxes) near the Himalayan region. One can see a positive correlation between concentration of small droplets and concentration of CCN below cloud base. We interpret this relationship as evidence for the fact that small APs in the boundary layer indeed cause the in-cloud nucleation (in zones of high supersaturation aloft) and the DSD broadening at the elevated layers.

Fig. 10.
Fig. 10.

Relationship between mean aerosol concentrations below cloud base and bin-average number concentration of droplets (CDNC) with D < 20 μm for CAIPEEX observations. CDNC is averaged at 1–3 km and 4–6 km above cloud base from the drop size distribution sampled at 1-s intervals. Boxed symbols correspond to two premonsoon highly polluted cases.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

2) Formation of bimodal and multimodal DSDs in the monsoon cloud

Examples of DSDs measured in the cloud core of monsoon cloud (22 June 2009) at 2.8-, 3.1-, and 5.1-km heights are presented in Fig. 11. Each DSD in the figures is plotted for every 1 s (which means approximately at 100-m length interval). Formation of a bimodal DSD is clearly seen at z = 2.8 km (Fig. 11a) and z = 3.1 km (Fig. 11b), with two maxima in DSD spectra at 20 and near 10 μm. CIP images showed cloud droplets and no precipitation-sized particles. Note the existence of high variability of DSD along aircraft legs at the same height levels. Some DSDs contain a significant amount of small droplets and highly pronounced secondary peaks in the DSD, while other DSDs do not reveal such secondary peaks. As mentioned above, averaging the DSD along the airplane track may mask the in-cloud nucleation. The variation of concentration, LWC, effective radius, Sqs, and adiabatic fraction also could be registered along with DSD. The bimodal DSDs are especially noticed at locations where adiabatic fraction exceeds 0.5 in the strong updrafts. The width of DSDs increases with height. The main peak is shifted at the 5.1-km height from 20 to about 30 μm. The DSDs at 5.1 km also indicate a significant concentration of droplets smaller than 25 μm forming the specific shape of distributions with low slopes within drop diameters from about 10 to 25 μm. This part of DSD forms, supposedly, as a result of several acts of in-cloud nucleation with successive diffusional growth of nucleated droplets. Since diameters of smaller droplets increase faster in cloud updrafts than those of larger droplets, the gaps between secondary peaks decrease (or fully disappear), resulting in formation of plateaus with low slope. CIP images in this case showed the existence of a few rain drops.

Fig. 11.
Fig. 11.

Plots of 1-Hz droplet size distributions in cloud core of monsoon cloud (22 Jun 2009) at (a) 2.8, (b) 3.1, and (c) 5.1 km. Legend shows time, height, cloud droplet concentration, effective radius, vertical velocity, Sqs, and adiabatic fraction. A representative CIP image is shown alongside each figure.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

3) Formation of bimodal DSD in the premonsoon cloud

DSDs in highly polluted premonsoon clouds are typically unimodal. Since DSDs are narrower in these clouds, the diameters of the first (larger) mode in these clouds are smaller and the gaps between modes are less pronounced or absent. This finding in the premonsoon clouds is presented in Fig. 12a. The increase in the number concentration of smaller droplets (<10 μm) results in the broadening of DSD, without a well-defined second mode in the DSD. Unimodal DSD with broadening is rather persistent and a DSD without such effects is rare. It is remarkable; in these clouds at some 100-m length intervals bimodal and multimodal DSD were observed in cloud cores as well (Fig. 12b). Bimodal DSD were found in transition period clouds with updrafts exceeding 4 m s−1 and an adiabatic fraction greater than 0.5 (Fig. 12c). The prevalence of multimodal DSD increased for the transition cloud (Fig. 12c). This type of situation was more often noticed in this cloud at different levels. Another interesting situation is presented in Fig. 12d at 0939:14 and at 0939:15 UTC that a significant increase in the droplet concentration is noted in strong cloud updrafts and with a relatively small adiabatic fraction (0.05 and 0.08), but at the cloud core 600–700 m away from the cloud edge. This is a good example where in-cloud nucleation is noticed in regions where liquid water content has decreased because of the unloading of large droplets. The subsequent increase in vertical velocity leads to supersaturations of 0.92% and 1.17%.

Fig. 12.
Fig. 12.

Examples of bimodal and multimodal DSDs in premonsoon cloud at (a) 6.1 and (b) 5.4 km and transition cloud at (c) 5.7 and (d) 5 km where in-cloud nucleation in the strong updrafts is noticed. Legend as in Fig. 11.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

More often the formation of bimodal and multimodal DSDs in clean maritime clouds is attributed to higher values of supersaturation in maritime clouds. In this context, it is remarkable that observations (Figs. 12b,c) indicate the existence of bimodal DSD in highly polluted premonsoon and transition period clouds as well.

d. Observations of small droplets in cloud downdrafts

In this section we consider the existence of small droplets in the cloud downdrafts. Figure 9 shows that in deep cumulus clouds strong downdrafts exist with velocities up to 12 m s−1. Such downdrafts are especially pronounced in monsoon clouds at the mature stage. Typically the absolute values of Sqs in downdrafts are lower than 1% (Fig. 9); that is, undersaturation is quite low. The comparatively low absolute values of Sqs in downdrafts are related to the fact that downdrafts in cloud cores typically arise from upper levels where liquid water content [and integral radii length in Eq. (1)] is high.

Figure 13 shows 1-Hz DSDs in cloud core of monsoon cloud (22 June 2009) at 4.4- and 6.7-km height levels in the presence of downdrafts. Analysis shows the existence of small droplets in zones of cloud downdrafts. We suppose that the smallest droplets observed in downdrafts are formed by partial evaporation of cloud droplets, some of which formed during in-cloud nucleation in previous updrafts and originate from higher levels in the cloud. This evaporation can lead to the formation of the minimum in the DSD at drop diameters of 18–20 μm (seen in Fig. 13a at z = 4.4 km). Evaporation of larger cloud droplets and small raindrops cannot be the only reason for the formation of the 10-μm-diameter droplets because the peaks of large droplets centered at 25–35 μm remain in downdrafts. Another profound example of bimodal DSD in the premonsoon cloud is presented in Fig. 13c at 6.2 km. A few raindrops were found in this section of the cloud pass. Two modes are clearly divisible in the weak undersaturation. It is remarkable that adiabatic fraction is approximately 0.4.

Fig. 13.
Fig. 13.

As in Fig. 11, but at (a) 4.4 and (b) 6.7 km in the presence of downdrafts.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

One can argue that contribution of raindrop breakup to the formation of smallest droplets is likely negligible. Breakup of raindrops is efficient for raindrops with diameters exceeding a few millimeters (Low and List 1982; Pruppacher and Klett 1997; Straub et al. 2010). The concentration of such raindrops is several orders of magnitude lower than the concentration of the smallest cloud droplets. According to recent results (e.g., Straub et al. 2010), the number of fragments (i.e., smaller drops) formed during breakup of a raindrop varies within the range of 2–5. It means that the number of drops formed by breakup remains much lower than the concentration of the smallest droplets observed in downdrafts. Moreover, the size of drops formed as a result of breakup typically exceeds 100–200 μm (Straub et al. 2010) (i.e., is much larger than the size of smallest droplets).

The partial (and possibly total) evaporation of small droplets in downdrafts suggests the possibility of the mechanism of successive nucleation/denucleation (evaporation) during recirculation in cumulus clouds. Typically this mechanism is considered in relation to DSD formation in stratocumulus clouds where updrafts and downdrafts are on the same order of magnitude (e.g., Erlick et al. 2005; Magaritz 2009). The mechanism by which aerosols formed by evaporation of droplets in one cloud give rise to the formation of another cloud was discussed by Xue et al. (2010), who analyzed linear warm clouds over two successive mountains. The process of droplet evaporation, as well as recirculation and nucleation/denucleation mechanisms, leads to the DSD broadening in stratocumulus clouds (Korolev 1995; Erlick et al. 2005; Magaritz et al. 2009). The role of such processes in mature cumulus clouds that already contain raindrops requires further investigation.

e. Multimode DSD distribution in mixed-phase clouds

The number of detailed measurements of DSDs in mixed-phase clouds at heights of 7 km is quite limited. Our emphasis in current section is to analyze the DSD shape at elevated layers where cloud droplets, rain drops, and graupel coexist. Figure 14 shows DSDs for the approximately 6.7-km level (4.2 km from cloud base) from 21 June (Fig. 14a) and for the approximately 7.1-km level (4.3 km from cloud base) from 22 June (Fig. 14b). Flight passage length is about 600 m. DSDs on 22 June are significantly wider than those on 21 June. One can see high variability of DSDs along the flight track. The concentration of small droplets varies considerably (by more than an order of magnitude). Some of the DSDs are bimodal and contain a second peak at small drop sizes. This bimodal nature of the DSD may at times get obscured by averaging along flight pass length as illustrated here for average DSD. It is to be emphasized that similar DSDs are seen for several flight passes and the bimodal distribution is statistically grounded. On 22 June the second mode centered at D ≈ 10 μm is seen at each pass, so the two modes do not disappear by spatial averaging. One can see that the second mode (which we attribute to in-cloud nucleation) takes place in the presence of supercooled raindrops and graupel (see CIP images in Fig. 14d for the same time sections, indicating the presence of cloud water).

Fig. 14.
Fig. 14.

(a),(b) DSDs along a flight pass length from (a) 21 Jun, indicating in-cloud nucleation leading to small drops and (b) 22 Jun, indicating in-cloud nucleation contributing to broadening of DSDs. Pass length averaged DSD is bimodal as shown. (c),(d) CIP images show clear indications of mixed phase (ice, rain, and cloud droplets) on (c) 21 and (d) 22 Jun.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

We interpret the existence of small droplets in the presence of ice as the evidence that supersaturation in cloud updrafts in mixed-phase clouds can be both over the water and ice. KPPB showed that formation of ice may decrease droplet concentration by riming, which in turn can increase supersaturation with respect to water and foster new in-cloud nucleation. Recently, Heymsfield et al. (2009) reported the formation (new nucleation) of cloud droplets above 6 km in maritime clouds. It seems that the DSD structure shown in Fig. 15 indicates the efficiency of such mechanism in continental clouds as well.

Fig. 15.
Fig. 15.

The conceptual scheme of droplet spectra formation in deep convective clouds [after Pinsky and Khain (2002) with changes for monsoon clouds].

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/2011JAS3707.1

The appearance of small crystals in measurements due to breakup of large aggregates at the tips of the probe (the effect of shattering in ice clouds) is now widely discussed (McFarquhar et al. 2007; Heymsfield 2007; Jensen et al. 2009; Korolev et al. 2011). The probability of artificial appearance of small droplets in the measured DSDs is much less for many reasons. Comparison of DSDs measured in clouds using an old probe and a modified probe (free from this instrumental deficiency) does not reveal any difference (A. Korolev 2010, personal communication) for water clouds. The shattering effect is a widespread problem in the case of ice clouds containing a significant amount of large aggregates (Korolev et al. 2011). In the mixed-phase cloud observations presented here, the number concentration of large aggregates is very low (see Figs. 14c,d). But it is apparent that shattering effects could not contribute to such large concentrations of the smallest particles. So the small particles are water droplets, observed primarily in the cloud updrafts. The contribution of shattering of aggregates, if any exists, will be investigated in a future study for clouds containing such aggregates.

4. Discussion

In situ measurements provided unique information concerning mechanisms of DSD formation in clouds developing under different thermodynamic and aerosol environments. The formation of small droplets and DSDs with significant values of dispersion, as well as formation of bimodal and multimodal 1-Hz DSDs, likely can be attributed to the process of in-cloud nucleation of small CCN ascending from the cloud base. This hypothesis is supported by high values of quasi-stationary supersaturation that can reach about 5%–8% at significant distances above cloud base, as well as by high correlation of the concentration of smallest droplets at high levels (e.g., at 6–7 km) with the AP concentration in the boundary layer below cloud base.

These findings correspond to the scheme of the DSD formation described by Pinsky and Khain (2002) and KPPB and adapted for specific conditions of monsoon clouds (Fig. 15). At the cloud base, the size distribution of aerosols (nonactivated CCN) is wide and includes particles with dry radii ranging from about 0.002 to about 2 μm (Hobbs 1993; Pruppacher and Klett 1997). The existence of small CCN follows from the in situ observations indicating very high values of slope parameter. According to our results, supersaturation at cloud base did not exceed 0.3%–0.4%, so that CCN with dry radii below about 0.05 μm remained nonactivated near cloud base. Droplets activated near the cloud base give rise to the formation of the first mode in DSD, which plays a dominant role in warm rain formation. If the supersaturation in updraft exceeds the maximum supersaturation values reached below, a new CCN activation (i.e., in-cloud nucleation of droplets) takes place, giving rise to formation of the second cloud mode of the DSD. In zones of efficient collisions between drops and rapid fallout of rain, the droplet concentration decreases and vertical velocity increases (possibly because of partial unloading of the large drops). As a result, the supersaturation may substantially increase and a new portion of the smallest CCN is activated, producing small droplets at heights 4–6 km above the cloud base.

The continuous presence of small droplets at higher levels in monsoon clouds indicates continuous in-cloud nucleation caused by an increase in supersaturation with height in accelerating cloud updrafts. Note that small droplets after their in-cloud nucleation start growing faster than larger droplets, which leads to a decrease or even disappearance of gaps between the modes, leading in turn to a unimodal DSD having a dispersion of 0.2–0.3 and containing very small droplets at high levels.

It is typically assumed that in-cloud nucleation takes place largely in maritime clouds where cloud-base supersaturation is low and rapidly grows with height because of low droplet concentration and an increase in vertical velocity (Warner 1969a,b; Pinsky and Khain 2002; Segal et al. 2003). The results of the in situ observations presented in the present study show that in-cloud nucleation is a much more widespread phenomenon than was widely assumed. This finding agrees well with that of Heymsfield et al. (2009), who reported an especially high concentration of small droplets at elevated heights in polluted maritime clouds observed in the African monsoon study NAMMA.

Another finding is the existence of strong downdrafts in convective clouds, especially pronounced at the mature stage of monsoon clouds. Analysis shows that small droplets form in downdrafts, likely because of the evaporation of larger cloud droplets, most of which were probably nucleated in cloud updrafts. Formation of small droplets in cloud downdrafts also fosters DSD broadening. Evaporation (partial or total) of droplets in cloud downdrafts suggests that the mechanism of successive droplet in-cloud nucleation/denucleation (evaporation) can take place not only in stratocumulus clouds, but also in deep cumulus clouds. The role of such mechanisms should be investigated in more detail in numerical simulations.

The features of DSD evolution in polluted premonsoon and monsoon clouds found in the in situ measurements are of great importance in many aspects. Note that the drop–drop collision rate is determined by the shape of local DSDs (calculated at scales of 100 m and even smaller), and not by DSDs averaged over large distance. As shown by Pinsky and Khain (2002), the existence of the second mode may accelerate the collision rate by about 30%. The possible effects of new droplets on the rate of riming, as well as on vertical velocity related to additional latent heat release (KPPB) are also of importance.

Formation of new droplets by in-cloud nucleation may help to explain many cloud microphysical phenomena such as a high optical depth of cloud anvils in maritime deep convective clouds within the intertropical convergence zone, high concentration of small ice crystals in cloud anvils above the level of homogeneous freezing, lightning in the eyewalls of tropical clouds (charge separation requires a significant supercooled water at temperatures colder than −20°C), and so on. Since concentration of small ice crystals affect cloud radiative properties, in-cloud nucleation leading to formation of small droplets and small crystals by homogeneous freezing becomes an important mechanism influencing cloud radiative properties.

The existence of small droplets in mixed-phase clouds indicates the existence of supersaturation in cloud updrafts both over the ice and water. Moreover, the results suggest that formation of ice can decrease droplet concentration and increase supersaturation with respect to water, leading to new in-cloud nucleation (as simulated by KPPB). We believe that second mode seen in Fig. 14 consists of liquid droplets, also seen in the CIP images, but not of the smallest ice crystals formed by the Hallet–Mossop ice-multiplication mechanism (hereafter the H-M mechanism; Hallet and Mossop 1974), since the concentration of the smallest droplets is much higher than that of crystals that can be produced by the H-M mechanism. This finding indicates that the Wegener–Bergeron–Findeisen (WBF) process, by which ice in mixed-phase clouds grows at the expense of evaporating droplets, turns out to be inefficient. This conclusion agrees well with those reached from theoretical considerations of Korolev (2007) and numerical simulations of mixed-phase cumulus clouds (e.g., KPPB). Note that the theoretical background of glaciogenic cloud seeding with silver iodide (AgI-aimed rain enhancement) is based on the WBF mechanism. It means that the concept of glaciogenic seeding, at least in relation to convective clouds, should possibly be reconsidered.

Unique in situ measurements during CAIPEEX indicated dramatic differences in microphysics (and seemingly in accumulated surface precipitation) between premonsoon and monsoon clouds considered in this study. The difference in CCN concentrations seems to be the main factor, determining the difference in cloud microphysics. Using a cloud parcel model, Reutter et al. (2009) showed that the concentration of droplets depends on the ratio between vertical velocity (determining supersaturation) and the CCN concentration, which is able to activate at supersaturations forming in clouds at particular vertical velocities. In their simulations, Reutter et al. (2009) calculated droplet concentrations for a wide range of updraft velocities from 0.25 to 20 m s−1. In the CAIPEEX clouds under consideration here, the differences in the vertical velocities in the cases considered were not as crucial as in the simulations by Reutter et al. (2009). Under this situation, the difference in concentration of CCN that can be activated at supersaturations below 0.3% is, as we believe, the major factor responsible for difference in the heights of first raindrop formation. The crucial effect of CCN concentration on the height of raindrop formation is stressed in many observational (e.g., Rosenfeld and Lensky 1998; Rosenfeld et al. 2008; Freud et al. 2008; Andreae et al. 2004) and numerical studies [see review by Khain (2009)]. Therefore, any numerical model aimed at prediction of precipitation during the monsoon period should take aerosol effects into account. The implications for observations is also highlighted that small aerosols (CCN) are often not measured by standard CCN probes, which indicates the necessity of increasing the range of supersaturation to determine the proper CCN size spectrum.

5. Summary of main findings

Seminal observations of continental clouds up to 7 km above ground are performed during CAIPEEX. In situ measurements of dynamics and microphysics of premonsoon and monsoon clouds indicate dramatic differences. The analysis reveals several new findings as follows:

  • DSDs broaden faster in continental monsoon clouds developing in relatively clean air as compared to premonsoon clouds developing in highly polluted air.
  • Cloud droplets grow faster in monsoon clouds and reach a size that can trigger efficient collisions earlier and at lower levels than in premonsoon clouds. This leads to different heights of onset of raindrop formation above the cloud base: about 4.5 km in premonsoon clouds and about 2.7 km in monsoon clouds.
  • Intense raindrop formation in monsoon clouds begins when effective radius reaches about 14 μm, which is in agreement with other studies (e.g., Freud et al. 2008; Pinsky and Khain 2002). In the case of very polluted premonsoon clouds, the effective radius remains below 14 μm, similar to results reported by Freud et al. (2008) for extremely polluted clouds measured in situ in the LBA-SMOCC campaign in the Amazon region. Thus, for very polluted cases the critical size of effective radius can be considered to be about 10–12 μm.

Detailed analysis of DSD evolution with height, as well as of the variability of DSDs and other microphysical parameters along flight trajectories at the same altitude, was performed. The following main features of DSDs were found:

  • In spite of significant variations of CWC along horizontal flight trajectories, variations of effective radius at these levels are comparatively small. This result agrees with that reported by Freud et al. (2008) and requires further investigation.
  • Droplet spectrum dispersion in premonsoon cloud is about 0.3, and ranges from 0.3 to 0.6 in monsoon clouds. No tendency of the dispersion to decrease with height up to height levels of 7 km was found, which is similar to studies of much weaker cumulus clouds by Politovich (1993). These values of dispersion are much higher than those expected in adiabatic cloud updraft due to droplet growth through diffusion.
  • Small droplets with diameters below about 10–20 μm exist nearly at each level, including the levels located at the distances of several kilometers above cloud base. These small droplets were found within clouds at significant distances from cloud edges in zones of high supersaturations in cloud updrafts. Small droplets were found also in downdrafts in undersaturated regions but were more frequently observed for monsoon cloud.
  • The concentration of these small droplets at some height levels a few kilometers above cloud base can experience dramatic jumps. In the monsoon cloud, the rapid increase in the concentration of small droplets was found just above the level of intense raindrop formation. Within the layer of appearance of small droplets, the concentration of large droplets decreased, indicating unloading due to raindrop settling.
  • High correlation was found between concentration of smallest droplets at high levels (e.g., at 6 km) and aerosol concentration below cloud base.
  • Many individual (1 Hz) DSDs are bimodal or multimodal (especially in monsoon clouds), which is attributed to the presence of small droplets. The horizontal averaging can mask the multimodality of individual spectra because of different locations of secondary modes in the individual DSDs.

In general, the results clearly indicate a dramatic effect of aerosols on cloud microphysics and precipitation formation in agreement with earlier studies (e.g., Andreae et al. 2004; Freud et al. 2008; Levin and Cotton 2009; Khain 2009) and show essentially that in-cloud nucleation and evaporation and resulting bimodal DSDs are fundamental processes to be accounted for in numerical models, including large-scale and climate models to predict monsoon rainfall.

Acknowledgments

The CAIPEEX project and IITM are fully funded by Ministry of Earth Sciences, Government of India, and New Delhi. The authors express their gratitude to Prof. D. Rosenfeld, E. Freud, and D. Axisa for stimulating discussions and for their contribution on the data and quality checks. Dr. Jim Dudhia, Dr. Wojciech W. Grabowski, Dr. Roelof Bruintjes, and Dr. Lulin Xue are also thanked for scientific discussions. Dr. A. Karipot is thanked for several suggestions and edits of the manuscript. Authors acknowledge with gratitude that the team effort and dedication of several scientists at IITM made CAIPEEX a grand success. AK is partially supported through the project HAMP (Project FY2008-06-16) supported by the Department of Homeland Security of the United States, and the Israel Science Foundation (Grant 140/07). Three anonymous reviewers are thanked for their critical and detailed comments on the first version of the manuscript leading to considerable improvement of the manuscript. We are also grateful to the editor, Prof. Robert Houze, for valuable suggestions for synthesizing the reviews, thus enhancing quality of the paper. The first author would also like to acknowledge partial support from the Mesoscale Microscale Meteorology Division, National Center for Atmospheric Research (NCAR), United States, through a faculty visitors program.

REFERENCES

  • Andreae, M. O., , D. Rosenfeld, , P. Artaxo, , A. A. Costa, , G. P. Frank, , K. M. Longo, , and M. A. F. Silva-Dias, 2004: Smoking rain clouds over the Amazon. Science, 303, 13371342.

    • Search Google Scholar
    • Export Citation
  • Baumgardner, D., , H. Jonsson, , W. Dawson, , D. O’Connor, , and R. Newton, 2001: The cloud, aerosol and precipitation spectrometer: A new instrument for cloud investigations. Atmos. Res., 59–60, 251264.

    • Search Google Scholar
    • Export Citation
  • Brenguier, J. L., , and W. Grabowski, 1993: Cumulus entrainment and cloud droplet spectra: A numerical model within a two-dimensional dynamical framework. J. Atmos. Sci., 50, 120136.

    • Search Google Scholar
    • Export Citation
  • Erlick, C., , A. Khain, , M. Pinsky, , and Y. Segal, 2005: The effect of wind velocity fluctuations on drop spectrum broadening in stratiform clouds. Atmos. Res., 75, 1545.

    • Search Google Scholar
    • Export Citation
  • Freud, E., , D. Rosenfeld, , M. O. Andreae, , A. A. Costa, , and P. Artaxo, 2008: Robust relations between CCN and the vertical evolution of cloud drop size distribution in deep convective clouds. Atmos. Chem. Phys., 8, 16611675.

    • Search Google Scholar
    • Export Citation
  • Fridlind, A., and Coauthors, 2004: Evidence for the predominance of mid-tropospheric aerosols as subtropical anvil cloud nuclei. Science, 304, 718722.

    • Search Google Scholar
    • Export Citation
  • Hallett, J., , and S. C. Mossop, 1974: Production of secondary ice crystals during the riming process. Nature, 249, 2628.

  • Heymsfield, A. J., 2007: On measurements of small ice particles in clouds. Geophys. Res. Lett., 34, L23812, doi:10.1029/2007GL030951.

  • Heymsfield, A. J., , A. Bansemer, , G. Heymsfield, , and A. O. Fierro, 2009: Microphysics of maritime tropical convective updrafts at temperatures from −20° to −60°C. J. Atmos. Sci., 66, 35303562.

    • Search Google Scholar
    • Export Citation
  • Hobbs, P.V. 1993: Aerosol-Cloud-Climate Interactions. Academic Press, 236 pp.

  • Houze, R. A., , and D. D. Churchill, 1984: Microphysical structure of winter monsoon cloud clusters. J. Atmos. Sci., 41, 34053411.

  • Jaenicke, R., 1993: Tropospheric aerosols. Aerosol–Cloud–Climate Interactions, P. Hobbs, Ed., Academic Press, 1–31.

  • Jensen, E. J., and Coauthors, 2009: On the importance of small ice crystals in tropical anvil cirrus. Atmos. Chem. Phys. Discuss., 9, 53215370.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., 2009: Notes on state-of-the-art investigations of aerosol effects on precipitation: A critical review. Environ. Res. Lett., 4, 015004, doi:10.1088/1748-9326/4/1/015004.

    • Search Google Scholar
    • Export Citation
  • Khain, A. P., , and A. Pokrovsky, 2004: Simulation of effects of atmospheric aerosols on deep turbulent convective clouds using a spectral microphysics mixed-phase cumulus cloud model. Part II: Sensitivity study. J. Atmos. Sci., 61, 29833001.

    • Search Google Scholar
    • Export Citation
  • Kogan, Y., 1991: The simulation of a convective cloud in a 3D model with explicit microphysics. Part I: Model description and sensitivity experiments. J. Atmos. Sci., 48, 11601189.

    • Search Google Scholar
    • Export Citation
  • Korolev, A. V., 1994: A study of bimodal droplet size distributions in stratiform clouds. Atmos. Res., 32, 143170.

  • Korolev, A. V., 1995: The influence of supersaturation fluctuations on droplet size spectra formation. J. Atmos. Sci., 52, 36203634.

  • Korolev, A. V., 2007: Limitations of the Wegener–Bergeron–Findeisen mechanism in the evolution of mixed-phase clouds. J. Atmos. Sci., 64, 33723375.

    • Search Google Scholar
    • Export Citation
  • Korolev, A. V., , and I. Mazin, 2003: Supersaturation of water vapor in clouds. J. Atmos. Sci., 60, 29572974.

  • Korolev, A. V., , E. F. Emery, , J. W. Strapp, , S. G. Cober, , G. A. Isaac, , M. Wasey, , and D. Marcotte, 2011: Small ice particles in tropospheric clouds: Fact or artifact? Airborne Icing Instrumentation Evaluation Experiment. Bull. Amer. Meteor. Soc., in press.

    • Search Google Scholar
    • Export Citation
  • Lasher-Trapp, S. G., , W. A. Cooper, , and A. M. Blyth, 2005: Broadening of droplet size distributions from entrainment and mixing in a cumulus cloud. Quart. J. Roy. Meteor. Soc., 131, 195220.

    • Search Google Scholar
    • Export Citation
  • Levin, Z., , and W. R. Cotton, Eds., 2009: Aerosol Pollution Impact on Precipitation: A Scientific Review. Springer, 386 pp.

  • Low, T. B., , and R. List, 1982: Collision, coalescence and breakup of raindrops, Part II: Parameterization of fragment size distributions. J. Atmos. Sci., 39, 16071619.

    • Search Google Scholar
    • Export Citation
  • Ludlam, F. H., 1980: Clouds and Storms. The Pennsylvania State University Press, 405 pp.

  • Magaritz, L., , M. Pinsky, , O. Krasnov, , and A. Khain, 2009: Investigation of droplet size distributions and drizzle formation using a new trajectory ensemble model. Part II: Lucky parcels. J. Atmos. Sci., 66, 781805.

    • Search Google Scholar
    • Export Citation
  • McFarquhar, G. M., , J. Um, , M. Freer, , D. Baumgardner, , G. L. Kok, , and G. Mace, 2007: Importance of small ice crystals to cirrus properties: Observations from the Tropical Warm Pool International Cloud Experiment (TWP-ICE). Geophys. Res. Lett., 34, L13803, doi:10.1029/2007GL029865.

    • Search Google Scholar
    • Export Citation
  • Ochs, H. T., 1978: Moment-conserving techniques for warm cloud microphysical computations. Part II: Model testing and results. J. Atmos. Sci., 35, 19591973.

    • Search Google Scholar
    • Export Citation
  • Paluch, I. R., , and C. A. Knight, 1984: Mixing and the evolution of cloud droplet size spectra in a vigorous continental cumulus. J. Atmos. Sci., 41, 18011815.

    • Search Google Scholar
    • Export Citation
  • Phillips, V. T. J., and Coauthors, 2005: Anvil glaciation in a deep cumulus updraft over Florida simulated with an explicit microphysics model. I: The impact of various nucleation processes. Quart. J. Roy. Meteor. Soc., 131, 20192046.

    • Search Google Scholar
    • Export Citation
  • Pinsky, M., , and A. P. Khain, 2002: Effects of in-cloud nucleation and turbulence on droplet spectrum formation in cumulus clouds. Quart. J. Roy. Meteor. Soc., 128, 501533.

    • Search Google Scholar
    • Export Citation
  • Politovich, M. K., 1993: A study of the broadening of droplet size distributions in cumuli. J. Atmos. Sci., 50, 22302244.

  • Politovich, M. K., , and W. A. Cooper, 1998: Variability of the supersaturation in cumulus clouds. J. Atmos. Sci., 45, 16511664.

  • Pruppacher, H. R., , and J. D. Klett, 1997: Microphysics of Clouds and Precipitation. 2nd ed. Kluwer Academic, 954 pp.

  • Reutter, P., and Coauthors, 2009: Aerosol- and updraft-limited regimes of cloud droplet formation: Influence of particle number, size and hygroscopicity on the activation of cloud condensational nuclei (CCN). Atmos. Chem. Phys., 9, 70677080.

    • Search Google Scholar
    • Export Citation
  • Rosenfeld, D., , and I. M. Lensky, 1998: Satellite-based insights into precipitation formation processes in continental and maritime convective clouds. Bull. Amer. Meteor. Soc., 79, 24572476.

    • Search Google Scholar
    • Export Citation
  • Rosenfeld, D., , U. Lohmann, , G. B. Raga, , C. D. O’Dowd, , M. Kulmala, , S. Fuzzi, , A. Reissell, , and M. O. Andreae, 2008: Flood or drought: How do aerosols affect precipitation? Science, 321, 13091313.

    • Search Google Scholar
    • Export Citation
  • Segal, Y., , M. Pinsky A. Khain, , and C. Erlick, 2003: Thermodynamic factors influencing bimodal spectrum formation in cumulus clouds. Atmos. Res., 66, 4364.

    • Search Google Scholar
    • Export Citation
  • Song, N., , and J. Marwitz, 1989: A numerical study of the warm rain process in orographic clouds. J. Atmos. Sci., 46, 34793486.

  • Squires, P., 1952: The growth of cloud drops by condensation. I. General characteristics. Aust. J. Sci. Res., 5, 5986.

  • Straub, W., , K. D. Beheng, , A. Seifert, , J. Schlottke, , and B. Weigang, 2010: Numerical investigation of collision-induced breakup of raindrops. Part II: Parameterizations of coalescence efficiencies and fragment size distributions. J. Atmos. Sci., 67, 576587.

    • Search Google Scholar
    • Export Citation
  • Su, C.-W., , S. K. Krueger, , P. A. McMurtry, , and P. H. Austin, 1998: Linear eddy modeling of droplet spectral evolution during entrainment and mixing in cumulus clouds. Atmos. Res., 47–48, 4158.

    • Search Google Scholar
    • Export Citation
  • Warner, J., 1969a: The microstructure of cumulus cloud. Part I: General features of the droplet spectrum. J. Atmos. Sci., 26, 10491059.

    • Search Google Scholar
    • Export Citation
  • Warner, J., 1969b: The microstructure of cumulus cloud. Part II: The effect of droplet size distribution of the cloud nucleus spectrum and updraft velocity. J. Atmos. Sci., 26, 12721282.

    • Search Google Scholar
    • Export Citation
  • Warner, J., 1973: The microstructure of cumulus cloud. Part IV: The effect on the droplet spectrum of mixing between cloud and environment. J. Atmos. Sci., 30, 256261.

    • Search Google Scholar
    • Export Citation
  • Xue, L., , A. Teller, , R. Rasmussen, , I. Geresdi, , and Z. Pan, 2010: Effects of aerosol solubility and regeneration on warm-phase orographic clouds and precipitation simulated by a detailed bin microphysical scheme. J. Atmos. Sci., 67, 33363354.

    • Search Google Scholar
    • Export Citation
Save