• Barrett, J. C., and Clement C. F. , 1988: Growth rates for liquid drops. J. Aerosol Sci, 19 , 223242.

  • Bilde, M., and Svenningsson B. , 2003: CCN activation of slightly soluble organics: Importance of small amounts of inorganic salt and particle phase. Tellus, 56B , 93182.

    • Search Google Scholar
    • Export Citation
  • Bird, R., Stewart W. , and Lightfoot E. , 1960: Transport Phenomena. John Wiley, 780 pp.

  • Bohren, C. F., and Huffman D. R. , 1983: Absorption and Scattering of Light by Small Particles. John Wiley, 530 pp.

  • Brenguier, J. L., Bourrianne T. , Coelho A. A. , Isbert J. , Peytavi R. , Trevarin D. , and Wechsler P. , 1998: Improvements of droplet size distribution measurements with the fast-FSSP. J. Atmos. Oceanic Technol, 15 , 10771090.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bunz, H., Möhler O. , Neumann K-H. , Saathoff H. , Schöck W. , and Schurath U. , 1996: The novel aerosol chamber facility AIDA: Status and first results. Proceedings of the Seventh Symposium on Physicochemical Behavior of Atmospheric Pollutants, B. Larsen, B. Versino, and G. Angeletti, Eds., EUR 17482, 673–677.

    • Search Google Scholar
    • Export Citation
  • Charlson, R. J., Seinfeld J. S. , Nenes A. , Kulmala M. , Laaksonen A. , and Facchini M. C. , 2001: Reshaping the theory of cloud formation. Science, 292 , 20252026.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chuang, P. Y., Charlson R. J. , and Seinfeld J. H. , 1997: Kinetic limitations on droplet formation in clouds. Nature, 390 , 594596.

  • Chuang, P. Y., Nenes A. , Smith J. N. , Flagan R. C. , and Seinfeld J. H. , 2000: Design of a CCN instrument for airborne measurements. J. Atmos. Oceanic Technol, 17 , 10051019.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cziczo, D. J., and Abbatt J. P. D. , 1999: Deliquescence, efflorescence, and supercooling of ammonium sulfate aerosols at low temperature: Implications for cirrus cloud formation and aerosol phase in the atmosphere. J. Geophys. Res, 104 , 1378113790.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cziczo, D. J., Nowak J. B. , Hu J. H. , and Abbatt J. P. D. , 1997: Infrared spectroscopy of model tropospheric aerosols as a function of relative humidity: Observation of deliqescence and crystallization. J. Geophys. Res, 102 , 1884318850.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DeMott, P. J., and Rogers D. C. , 1990: Freezing nucleation rates of dilute solution droplets measured between −30° and −40°C in laboratory simulations of natural clouds. J. Atmos. Sci, 47 , 10561064.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dye, J. E., and Baumgardner D. , 1984: Evaluation of the forward scattering spectrometer probe. Part I: Electronic and optical studies. J. Atmos. Oceanic Technol, 1 , 329344.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Facchini, M. C., Mircea M. , and Charlson R. J. , 1999: Cloud albedo enhancement by surface-active organic solutes in growing droplets. Nature, 401 , 257259.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fluent, 2001: FLUENT 6 user's guide. Fluent Inc.

  • Grant, L. O., and Steele R. L. , 1966: The calibration of silver iodide generators. Bull. Amer. Meteor. Soc, 47 , 713717.

  • Hagen, D. E., Schmitt J. , Trueblood M. , Carstens J. , White D. R. , and Alofs D. J. , 1989: Condensation coefficient measurement for water in the UMR cloud simulation chamber. J. Atmos. Sci, 46 , 803816.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hall, W. D., and Pruppacher H. R. , 1976: The survival of ice particles falling from cirrus clouds in subsaturated air. J. Atmos. Sci, 33 , 19952006.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hindman, E. E., 1989: Formation and optical properties of a warm cloud in a slow-expansion cloud chamber. J. Atmos. Sci, 46 , 36533663.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hudson, J. G., 1989: An instantaneous CCN spectrometer. J. Atmos. Oceanic Technol, 6 , 10551065.

  • Hudson, J. G., and Squires P. , 1973: Evaluation of a recording continuous cloud nucleus counter. J. Appl. Meteor, 12 , 175183.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kim, Y. J., and Boatman J. F. , 1990: Corrections for the effects of particle trajectory and beam intensity profile on the size spectra of atmospheric aerosols measured with a forward scattering spectrometer probe. J. Atmos. Oceanic Technol, 7 , 673680.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Knutson, E. O., and Whitby K. T. , 1975: Aerosol classification by electric mobility: Apparatus, theory, and applications. J. Aerosol Sci, 6 , 443451.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kulmala, M., Laaksonen A. , Korhonen P. , Vesala T. , Ahonen T. , and Barrett J. C. , 1993: The effect of atmospheric nitric acid vapor on cloud condensation nucleus activation. J. Geophys. Res, 98 , 2294922958.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Leaitch, R., and Megaw W. J. , 1982: The diffusion tube: A cloud condensation nucleus counter for use below 0.3% supersaturation. J. Aerosol Sci, 13 , 297319.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mahata, P. C., Alofs D. J. , and Sinnarwalla A. M. , 1973: Supersaturation development in a vertical flow thermal diffusion chamber. J. Appl. Meteor, 12 , 13791383.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Möhler, O., and Coauthors, 2002: Experimental investigation of homogeneous freezing of sulphuric acid particles in the aerosol chamber AIDA. Atmos. Chem. Phys. Discuss, 2 , 14291465.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pruppacher, H. R., and Klett J. D. , 1997: Microphysics of Clouds and Precipitation. Kluwer Academic, 945 pp.

  • Rogers, D. C., 1988: Development of a continuous flow thermal gradient diffusion chamber for ice nucleation studies. Atmos. Res, 22 , 149181.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Saxena, V. K., and Carstens J. C. , 1971: On the operation of cylindrical thermal diffusion cloud chambers. J. Rech. Atmos., 11– 23.

  • Saxena, V. K., Burford J. N. , and Kassner J. L. Jr., 1970: Operation of a thermal diffusion chamber for measurements on cloud condensation nuclei. J. Atmos. Sci, 27 , 7380.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shulman, M. L., Jacobson M. C. , Charlson R. J. , Synovec R. E. , and Young T. E. , 1996: Dissolution behavior and surface tension effects of organic compounds in nucleating cloud droplets. Geophys. Res. Lett, 23 , 277280.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Twomey, S., 1974: Pollution and the planetary albedo. Atmos. Environ, 8 , 12511256.

  • Vietti, M. A., and Fastook J. L. , 1975: Water droplet growth in an expansion cloud chamber operating at small supersaturations. J. Rech. Atmos, 4 , 181195.

    • Search Google Scholar
    • Export Citation
  • Wendisch, M., Keil A. , and Korolev A. V. , 1996: FSSP characterization with monodisperse water droplets. J. Atmos. Oceanic Technol, 13 , 11521165.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wilck, M., Stratmann F. , and Whitby E. R. , 2002: A fine particle model for FLUENT: Description and application. Proc. Sixth Int. Aerosol Conf., Taipei, Taiwan, Chinese Association for Aerosol Research in Taiwan/International Aerosol Research Assembly, 1269–1270.

    • Search Google Scholar
    • Export Citation
  • View in gallery
    Fig. 1.

    (a) Temperature, (b) saturation, and (c) particle–droplet number concentration profiles inside the flow tube (flow direction top to bottom)

  • View in gallery
    Fig. 2.

    Saturation as a function of axial position z along the tube centerline (r = 0) for different inlet saturations Se

  • View in gallery
    Fig. 3.

    Particle/droplet diameter as a function of axial position z along the tube centerline (r = 0) for different inlet saturations Se

  • View in gallery
    Fig. 4.

    Liquid water mass fraction as a function of radial position r at the tube outlet for different inlet saturations Se

  • View in gallery
    Fig. 5.

    Experimental setup of LACIS: CPC, condensation particle counter; DMA, differential mobility analyzer; and MFC, mass flow controller

  • View in gallery
    Fig. 6.

    LACIS inlet

  • View in gallery
    Fig. 7.

    Comparison of measured (symbols) and ideal (solid line) normalized axial velocities at the tube outlet as a function of the radial position

  • View in gallery
    Fig. 8.

    Side view of the fast-FSSP location at the outlet of the LACIS tube: 1, LACIS tube with cooling jacket and extension pipe; 2, particle stream; 3, He–Ne laser beam; 4, light scattered into forward directions and redirected by the prism, 5, into the collecting optics; the laser beam is dumped by the black stop on the prism

  • View in gallery
    Fig. 9.

    Normalized number of counts (number of counts in pulse height channel i divided by total number of counts in all channels) distributed over the first 45 pulse height channels. Error bars indicate the variability within three sets of independent measurements

  • View in gallery
    Fig. 10.

    Actual response curve of the fast-FSSP calculated via Mie theory for spherical water droplets (solid line). The dashed line represents the smoothed curve used in the standard size retrieval procedure. Inset shows the size range relevant to the case discussed

  • View in gallery
    Fig. 11.

    Calculated (lines) and measured (symbols) mean droplet diameters dp as a function of saturator temperature Ts and for two different wall temperatures Tw

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 444 129 15
PDF Downloads 201 79 3

Laboratory Studies and Numerical Simulations of Cloud Droplet Formation under Realistic Supersaturation Conditions

F. StratmannInstitute for Tropospheric Research, Leipzig, Germany

Search for other papers by F. Stratmann in
Current site
Google Scholar
PubMed
Close
,
A. KiselevInstitute for Tropospheric Research, Leipzig, Germany

Search for other papers by A. Kiselev in
Current site
Google Scholar
PubMed
Close
,
S. WurzlerInstitute for Tropospheric Research, Leipzig, Germany

Search for other papers by S. Wurzler in
Current site
Google Scholar
PubMed
Close
,
M. WendischInstitute for Tropospheric Research, Leipzig, Germany

Search for other papers by M. Wendisch in
Current site
Google Scholar
PubMed
Close
,
J. HeintzenbergInstitute for Tropospheric Research, Leipzig, Germany

Search for other papers by J. Heintzenberg in
Current site
Google Scholar
PubMed
Close
,
R. J. CharlsonDepartment of Atmospheric Sciences, University of Washington, Seattle, Washington

Search for other papers by R. J. Charlson in
Current site
Google Scholar
PubMed
Close
,
K. DiehlInstitute for Tropospheric Research, Leipzig, Germany

Search for other papers by K. Diehl in
Current site
Google Scholar
PubMed
Close
,
H. WexInstitute for Tropospheric Research, Leipzig, Germany

Search for other papers by H. Wex in
Current site
Google Scholar
PubMed
Close
, and
S. SchmidtInstitute for Tropospheric Research, Leipzig, Germany

Search for other papers by S. Schmidt in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

In this paper, a new device is introduced to study the formation and growth of cloud droplets under near-atmospheric supersaturations. The new device, called the Leipzig Aerosol Cloud Interaction Simulator (LACIS), is based on a laminar flow tube. It has been designed to reproduce the thermodynamic conditions of atmospheric clouds as realistically as possible.

A series of experiments have been conducted that prove the definition and stability of the flow field inside the LACIS as well as the stability and reproducibility of the generated droplet size distributions as a function of the applied thermodynamic conditions. Measured droplet size distributions are in good agreement with those determined by a newly developed Eulerian particle–droplet dynamical model.

Further investigations will focus on the influences of latent heat release during vapor condensation on the tube walls and the development of a more suitable optical particle counter for droplet size determination.

Corresponding author address: Dr. F. Stratmann, Institute for Tropospheric Research, Permoserstr. 15, 04318 Leipzig, Germany. Email: straddi@tropos.de

Abstract

In this paper, a new device is introduced to study the formation and growth of cloud droplets under near-atmospheric supersaturations. The new device, called the Leipzig Aerosol Cloud Interaction Simulator (LACIS), is based on a laminar flow tube. It has been designed to reproduce the thermodynamic conditions of atmospheric clouds as realistically as possible.

A series of experiments have been conducted that prove the definition and stability of the flow field inside the LACIS as well as the stability and reproducibility of the generated droplet size distributions as a function of the applied thermodynamic conditions. Measured droplet size distributions are in good agreement with those determined by a newly developed Eulerian particle–droplet dynamical model.

Further investigations will focus on the influences of latent heat release during vapor condensation on the tube walls and the development of a more suitable optical particle counter for droplet size determination.

Corresponding author address: Dr. F. Stratmann, Institute for Tropospheric Research, Permoserstr. 15, 04318 Leipzig, Germany. Email: straddi@tropos.de

1. Introduction

Atmospheric clouds are complicated systems, exerting globally important controls on precipitation and planetary albedo, which are major climate factors. Cloud droplet size is one of the key factors influencing both precipitation initiation and albedo. However, cloud droplet size depends on a large number of physical, chemical, and thermodynamic parameters and processes. Many of them are still not sufficiently understood, for example, conversion of aerosol particles to droplets, turbulence, and the interactions of soluble gases and organic compounds with numerous cloud processes. The initial size and composition of aerosol particles and/or of droplets present during cloud formation influence the physical and chemical behavior of the cloud (e.g., optical properties, gas scavenging, and the formation of precipitation). However, it is very difficult to study real clouds in the atmosphere because of their complexity, inconvenient locations, and sporadic occurrence. Therefore, laboratory investigations in controlled and reproducible environments similar to real atmospheric conditions are of great importance. These experiments should be accompanied by numerical simulations that consider the crucial parameters and processes and that describe droplet formation (starting from the aerosol particles) and growth.

Traditionally, the equilibrium size of aerosol particles for different relative humidities is determined by Köhler theory (e.g., Pruppacher and Klett 1997). The growth of the aerosol particles and droplets by diffusion of water vapor is determined by the usual growth equation (Pruppacher and Klett 1997). In both cases, the effects of the water vapor concentration, the surface curvature (Kelvin effect), and the amount and type of soluble material in the particles–droplets (Raoult effect) are considered. Recent papers suggest extensions of this theory to include kinetic growth limitations (Chuang et al. 1997), soluble gases (Kulmala et al. 1993), slightly soluble substances (Shulman et al. 1996), and surface-active organics (Facchini et al. 1999). Recently, Bilde and Svenningsson (2003, manuscript submitted to Tellus) presented laboratory experiments on the effects of slightly soluble organic compounds on the equilibrium size of moist aerosol particles. The growth behavior of insoluble particles (crustal material, biomass burning, and biological particles), which is affected by, for example, capillary effects, is not described in the usual growth equation either. The influence of these different factors on the physical and chemical properties of clouds is estimated to be of global significance (Charlson et al. 2001). Importantly, the suggested extensions concerning the theoretical description of cloud droplet activation and growth are based mainly on numerical model calculations that still have to be verified experimentally.

There are a number of specific scientific tasks that need to be addressed by suitable laboratory studies:

  • comparison of the measured equilibrium size of pure salt particles and mixtures of salts with insoluble (or slightly soluble) material of various original sizes and compositions as a function of the water vapor saturation with Köhler theory;

  • comparison of the measured growth of aerosol particles and droplets by water vapor diffusion with known growth equations;

  • experimental investigation of the influence of atmospheric trace gases (e.g., HCl, HNO3) on the growth of aerosol particles (soluble, insoluble, and mixed);

  • experimental determination of the competition effect when aerosol particles of different sizes grow simultaneously as compared to the growth of the individual sizes; and

  • verification of parameterizations regarding the relationship between particle number–mass concentration and droplet number concentration as a function of the chemical and microphysical environment of the cloud. These are often used to estimate aerosol indirect forcing (Twomey 1974).

It is a challenge to study or even produce realistic clouds in a laboratory. Whenever attempting to develop a new device to study cloud processes in the laboratory, a dual perspective has to be maintained. First, the thermodynamic conditions and their evolution in time need to be considered in order to simulate the formation and evolution of an atmospheric cloud as realistically as possible. In tropospheric clouds, during the initial cloud droplet growth, relatively small supersaturations (≤2%) occur over periods of some minutes and distances of some 10–100 m. Second, the technical challenges in achieving, maintaining, monitoring, and controlling certain thermodynamic conditions near water vapor saturation have to be mastered.

Numerous attempts to reproduce water supersaturation conditions in laboratory devices are known. Cloud chambers have been developed in which the growth of an ensemble of aerosol particles or hydrometeors in a certain relative humidity range is studied (e.g., Grant and Steele 1966; Hudson and Squires 1973; Hudson 1989; Bunz et al. 1996; Möhler et al. 2002). Cloud chambers are classified with respect to the mechanism (mostly expansion or heat–vapor diffusion) used to achieve supersaturation. In expansion chambers (e.g., Vietti and Fastook 1975; Hindman 1989; Hagen et al. 1989; DeMott and Rogers 1990), supersaturations from atmospherically relevant ranges of up to several hundred percent are achieved by either geometrically expanding the chamber or by evacuating the system. Regarding diffusion chambers, basically two types exist: static diffusion chambers (e.g., Saxena et al. 1970; Saxena and Carstens 1971), which classically consist of two parallel horizontal metal plates with different temperatures and wetted surfaces, and flow diffusion chambers (e.g., Mahata et al. 1973; Leaitch and Megaw 1982; Rogers 1988; Cziczo et al. 1997; Cziczo and Abbatt 1999), which can be continuous-flow parallel-plate or flow tube diffusion chambers. Various technical solutions have been tried, including streamwise saturation gradients (Chuang et al. 2000).

The general conclusions from reviewing the known literature is that for the design of a device to be feasible for performing such investigations, a flow diffusion cloud chamber consisting of a laminar flow tube seems to be most suitable—even though,

  • the temperature and saturation fields inside such chambers are strongly inhomogeneous;

  • the exact conditions (temperatures, saturations, etc.) under which the particles–droplets grow cannot be measured;

  • due to the strong coupling between flow, heat–vapor transport and particle–droplet dynamics, numerical models have to be employed to determine these conditions; and

  • model calculations show that one of the major challenges in designing flow diffusion cloud chambers is the temperature (uncertainties in the temperatures should be <0.1 K) and consequently relative humidity control.

In this paper, a new laminar tube flow diffusion cloud chamber together with a new numerical model are introduced. This new device is called the Leipzig Aerosol Cloud Interaction Simulator (LACIS).

The basic LACIS design is somewhat similar to the diffusion tube based cloud condensation nuclei (CCN) counter of Leaitch and Megaw (1982) and to the streamwise gradient CCN instrument introduced by Chuang et al. (2000). Differences between existing devices and the one introduced here include how water vapor and particles are supplied to the flow tube, the direction of the heat flow, and how the actual thermodynamic conditions are controlled. The LACIS principle of operation is similar to that of the TSI condensation nucleus counters (e.g., TSI 3025, TSI Inc., St. Paul, Minnesota). In addition, the design suggested here is scalable. It allows upscaling for the construction of a larger cloud chamber, which makes it possible to study the initial stages of particle–droplet activation and growth close to both atmospheric supersaturations and time scales. All other similar devices reported in the literature were designed for downscaling in order to make the instrument applicable for aircraft measurements. LACIS will have a process control system, which allows a precise control of the thermodynamic profiles inside the tube based on the optical measurement of the droplet temporal evolution inside the tube.

Theory and a numerical model to describe the new device are introduced in section 2. The design and the detailed modeling of the prototype flow tube are discussed in section 3. Section 4 presents the experimental characterization of the device and the first results of measurements of the growth of synthetic particles of known size and composition and their respective comparisons with the model output.

2. Theory and numerical model

To describe the activation and growth of seed particles–droplets by vapor deposition inside the new laminar flow tube, the spatial air velocity, temperature, and vapor mass fraction distributions have to be known. These distributions are the result of the coupled mass– heat transfer, flow, and the phase transition processes. These processes are mathematically described by (a) the vapor mass transport, (b) the energy, and (c) momentum equations of the vapor–carrier-gas mixture, as well as (d) the equations for particle–droplet dynamics. The indices υ and g stand for vapor and gas, respectively.

Accounting for molecular and thermal diffusion, the mass transport equation for the vapor phase takes the form (Bird et al. 1960)
i1520-0426-21-6-876-e1
where jυ is the mass flux of vapor relative to the mass average velocity u; ρ is the density and T is the absolute temperature of the vapor–carrier-gas mixture; ωυ, Dυ, and αυ,g are the vapor mass fraction, binary vapor diffusion coefficient, and thermal diffusion factor, respectively; and Sυ stands for the vapor sink due to condensation onto droplets.
Considering the heat transport due to conduction and vapor transport and accounting for the Dufour term, the energy equation for a binary mixture (here water vapor in air) is given by
i1520-0426-21-6-876-e2
where q is the heat flux; h is the specific enthalpy; α = k/(ρcp) is the thermal diffusivity; k is the heat conductivity; cp is the specific heat capacity; M is the molar weight of the vapor–gas mixture; hυ and hg are the specific enthalpies; Mυ and Mg are the molar weights of the vapor and the carrier gas, respectively; and Sh = LυSυ represents the heat source, that is, the latent heat (Lυ is the vapor heat of vaporization) released by the vapor condensing onto particles/droplets.
Assuming axial rotational symmetry, a vertical orientation of the flow tube, and accounting for natural convection, the momentum equations take the form
i1520-0426-21-6-876-e3
where μ is the dynamic viscosity of the vapor–gas mixture, Vz and Vr are additional viscosity terms not included in ∇·(μu) and ∇·(μυ), and gz is the gravitational acceleration.
To describe particle/droplet transport due to convection, diffusion, external forces (here thermophoresis and sedimentation), and particle/droplet growth due to vapor condensation, a moving monodisperse model was used that solves for the total particle/droplet number concentration Np,
ρuupNpρDpNp
and particle/droplet mass concentrations Mp,i,
i1520-0426-21-6-876-e5
(i denotes the different substances in the particle/droplet phase). Here up is the particle–droplet external velocity, that is, the sum of the sedimentation and thermophoretic velocities, and ∂mp,i/∂t is the single particle growth law according to Barrett and Clement (1988):
i1520-0426-21-6-876-e6
where S and Sp are the vapor saturations in the gas phase and at the particle–droplet surface, respectively; dp is the particle–droplet diameter; Rυ is the vapor specific gas constant; pυ,e is the equilibrium vapor pressure; kg is the carrier gas heat conductivity; and Dυ is the vapor diffusion coefficient. For the air–water vapor mixture considered here, kg was determined with the Mason– Saxena equation (Pruppacher and Klett 1997), and Dυ was calculated following Hall and Pruppacher (1976). Here, fmass and fheat are the mass and heat transfer transition functions, respectively. Mass and heat accommodation coefficients are set equal to one. For the calculation of the saturation at the particle/droplet surface Sp, Kelvin and Raoult effects are accounted for.

In this work, Eqs. (1)–(5) are solved using the Computational Fluid Dynamics (CFD) code FLUENT 6 (Fluent 2001) together with the Fine Particle Model (Wilck et al. 2002), a newly developed Eulerian particle dynamical add-on module to FLUENT 6. This combined CFD–particle dynamics model allows the coupled solution of Eqs. (1)–(5) and consequently the determination of the flow velocity, temperature, vapor mass fraction and saturation, particle number, and particle/ droplet mass concentration fields inside the flow. The following important and potentially important processes are accounted for: vapor depletion due to vapor condensation on particle/droplets and tube walls, heat production due to latent heat release during vapor condensation, multicomponent and thermal vapor diffusion, and particle transport due to diffusion, thermophoresis, and sedimentation.

Presently, the model does not account for latent heat release due to vapor condensation on the tube walls.

3. Theoretical design of the flow tube

To determine a feasible design for the new flow tube, calculations were carried out using the model described in section 2. System parameters such as geometrical dimensions and initial and boundary conditions were varied such that near-atmospheric supersaturation levels can be achieved inside the flow tube. The resulting set of parameters is given in Table 1. A top to bottom flow direction was chosen to minimize possible technical problems due to gravitational settling of the droplets. The corresponding results are presented below.

Figure 1 shows the calculated temperature (panel a), saturation (panel b), and particle–droplet number concentration (panel c) as a function of radial (r) and axial (z) coordinates. It is obvious from Fig. 1a that the vapor– gas mixture is cooled by the tube walls; that is, the temperature first decreases near the walls. Temperatures near the tube center (r = 0) start to decrease significantly just behind the flow tube inlet (z about 0.1 m). Toward the outlet of the tube (z = 1.5 m), the vapor–gas mixture reaches thermal equilibrium with the tube walls. The saturation profile in Fig. 1b reflects the particularities of the water–air system [i.e., the vapor diffusion coefficient, Dυ, in Eq. (1) is approximately equal to the thermal diffusivity of the water vapor air mixture, α, in Eq. (2), i.e., the Lewis number, Le = α/Dυ ≈ 1]. The system starts to supersaturate from the tube walls and is extremely inhomogeneous. In this example, the maximum saturation reaches values larger than 1.10 close to the wall and is less than 1.05 near the centerline (r = 0).

Assuming a homogeneous distribution of seed particles at the tube inlet, the plotted saturation field yields a broad droplet number size distribution as the seed particles experience different saturations while traveling through the flow tube depending on their radial position at the inlet.

This problem can be avoided by injecting the seed particles only into a zone close to the flow tube centerline surrounded by particle-free sheath air. The radius of this injection zone should be chosen such that the radial inhomogeneities are as small as possible (ideally the radius should be 0, i.e., particles are only traveling along the centerline). As a consequence, in the following calculations an injection zone radius for the particle beam of 1 mm has been chosen. The corresponding particle/droplet number concentration profile is shown in Fig. 1c, which indicates a well-defined particle–droplet beam with negligible expansion. Figure 2 depicts the saturation profile along the flow tube centerline as a function of the axial position z and for different inlet relative humidities, that is, inlet saturations Se.

The saturation increases sharply about 0.1 m downstream of the injection point (z = 0), reaches its maximum between z = 0.2 and 0.4 m, and then decreases slowly, approaching its equilibrium value S = Sw = 1 near the flow tube outlet. The observed decline in saturation is mainly due to wall condensation as only a small fraction (less than 1% for all examples shown here) of the incoming vapor condenses on the injected particles. Again the particularity of the water–air system becomes obvious. As in certain parts of the tube Le < 1; that is, vapor diffuses faster than heat, the vapor becomes subsaturated although it is cooled, and a saturation of Sw = 1 is assumed at the tube wall. The strong dependence of the saturation profile on the inlet saturation Se is obvious. Here, maximum supersaturations of around 1% are achieved for the lowest and around 6% supersaturations for the highest inlet saturations. Maximum supersaturation can be adjusted to higher or lower values by changing the inlet saturation Se and/or the wall temperature Tw. The effect that the saturation decreases below 1 in certain parts of the tube is a somewhat unrealistic behavior compared to the conditions in real clouds. Future work will concentrate on avoiding this effect by means of (a) adjusting the system parameters and/or (b) measuring the droplet sizes at different axial positions inside the tube.

Figure 3 shows particle/droplet diameter dp along the flow tube centerline as a function of the axial position z and for different values of Se. The injected NaCl particles (initial dry diameter dp,e = 100 nm) almost instantaneously approach their equilibrium sizes corresponding to the relative humidity at the flow tube inlet. Then, the evolution of the particle/droplet diameter follows more or less the saturation profiles. About 0.1 m downstream of the injection point, at the location where the saturations increase (see Fig. 2), particles–droplets start to grow, reach their maximum size about halfway down the tube, and then reevaporate as the saturation drops below 1, that is, toward the end of the flow tube. For inlet Se values of 0.898, 0.927, 0.955, and 0.985, outlet droplet diameters dp (at z = 1.5 m) of 0.55, 2.4, 3.9, and 5.1 μm result, respectively. It should be noted that only for Se = 0.898 is the droplet in equilibrium with its surrounding at the tube outlet; that is, the droplet diameter reported above is the equilibrium droplet size for the thermodynamic conditions at z = 1.5 m.

Figure 4 presents the liquid water mass fraction Mp,υ at the flow tube outlet (z = 1.5 m) as a function of radial position r. Again, different inlet saturations Se are considered.

The liquid water mass fraction Mp,υ is nearly constant within the region determined by the injection zone and decreases sharply toward its boundary at r = 1.0 mm. This indicates that particles–droplets experience similar supersaturations while traveling through the flow tube. At the flow tube outlet, the relative differences comparing droplet sizes at the tube centerline (r = 0) to those found close to the boundary of the injection zone (r = 0.95 mm) are on the order of 3%. Again, the high sensitivity of the system with respect to the inlet saturation Se (more than three orders of magnitude difference in the liquid water mass fraction) becomes obvious.

4. Experimental realization and first results

The system design with the parameters given in Table 1 has been realized in a prototype laboratory setup. A schematic diagram of the setup is given in Fig. 5. The whole system is currently operated by pushing compressed air through it. All flow rates are adjusted using mass flow controllers (MKS 1179, MKS Instruments Deutschland GmbH, Munich, Germany). With an atomizer (item 1 in Fig. 5; TSI 3075), polydisperse aerosol particles are generated from an NaCl solution.

The generated aerosol particles pass through a drying unit (item 2 in Fig. 5) after which a flow rate reduction down to about 500 mL min−1 is achieved by removing an excess aerosol flow from the system. The remaining aerosol stream is passed through an adjustable dilution system (item 3), through a neutralizer (item 4), and then enters a differential mobility analyzer (DMA; item 5; Knutson and Whitby 1975), which extracts a quasi-monodisperse fraction of aerosol particles (here 100 ± 5 nm in diameter). Downstream of the DMA, the actual particle number concentration is determined by means of a condensation particle counter (item 6; TSI 3010), while the remaining aerosol flow is led through a saturator (item 7). There the aerosol is saturated with water vapor and the aerosol particles are hydrated. The saturator (MH-110-12S-4, Perma Pure, Toms River, New Jersey) is kept at the desired temperature by means of a high-precision thermostat (item 8; Haake C25P, Gebrüder HAAKE GmbH, Karlsruhe, Germany). Particle-free sheath air is passed through a second saturator (item 9; PH-30T-24KS, Perma Pure). The temperature of this saturation system is controlled by the same type of high-precision thermostat (item 10) as is used for the first saturator (item 7). Both saturators are operated in counterflow mode and saturator inlet temperatures are controlled and outlet temperatures are monitored.

At the inlet of LACIS (item 11), aerosol and sheath air are combined such that the aerosol is surrounded by laminarized particle-free sheath air (Fig. 6). The wall temperature of LACIS itself is adjusted by means of a counterflow water jacket, temperature controlled with a third precision thermostat (item 12 in Fig. 5; Haake F6, Gebrüder HAAKE GmbH). Again, the water jacket inlet and outlet temperatures are controlled and monitored, respectively. Currently, the LACIS wall material is quartz glass for both the inner and outer tubes. The outlet of LACIS is currently open, allowing the exiting droplets to pass a section where their number size distribution is measured by means of an optical particle counter, currently a fast forward scattering spectrometer probe (fast-FSSP; see section 4b; item 13).

Operating conditions (flow rates, temperatures, relative humidity, particle number concentration, and size) can be adjusted over a wide range. For the experiments described here, the operating parameters are summarized in Table 2. Based on the calculation it was decided that the flow tube length can be reduced to 1.0 m instead of 1.5 m as given in Table 1. The other parameters are the same or are in the range of those given in Table 1.

a. Characterization of the flow field

One of the important issues in designing LACIS is achieving a well-defined and stable particle–droplet beam surrounded by well-defined sheath air. To analyze the flow field, two different approaches were used: 1) The radial distribution of the axial flow velocity at the outlet of LACIS was measured using a hot-wire anemometer (Dantec Streamline, Dantec Dynamics GmbH, Erlangen, Germany) and 2) the particle beam was visualized by feeding droplets generated with a fog generator (Techno Fog, Martin Manufacturing, Lincolnshire, United Kingdom) into LACIS and illuminating the beam along its axis using a commercial red laser diode.

Both experiments were performed under dry (Se < 2%) and wet (Se > 90%) conditions and for wall temperatures Tw = 22.5°C (no vapor condensation) and 3.3°C (with vapor condensation).

Figure 7 shows measured axial velocities as a function of the radial position at the LACIS outlet. Plotted are normalized profiles, that is, actual velocity distributions u(r) divided by their corresponding maximum values, umax = u(r = 0). Here, umax was determined by fitting the ideal parabolic profile function
i1520-0426-21-6-876-e7
to the experimental data points; that is, umax was varied to minimize the least squares difference between the measured and fitted velocity values. Here, R = 7.5 mm corresponds to the actual LACIS geometry. In Fig. 7, the results for two different average flow velocities, u = 0.382 and 0.542 m s−1, for dry and wet conditions and wall temperatures Tw = 22.5° and 3.3°C are given. For comparison, the normalized ideal laminar flow velocity distribution (solid line) is given.

The measured and ideal velocity distributions in Fig. 7 are in good agreement, indicating a well-defined and fully developed laminar flow inside LACIS.

These findings are strongly supported by the beam visualization experiments (not shown here). The generated particle beam was checked visually with respect to its spatial position, shape, size, and stability. An undisturbed, well-defined, and stable particle beam was maintained for up to over 48 h under both wet and dry conditions and for the two considered wall temperatures, that is, with and without vapor condensation taking place.

b. Measurement of mean diameters of the droplet populations

To characterize the particle–droplet growth within LACIS, mean droplet diameters were measured at the LACIS outlet with a fast-FSSP. The fast-FSSP was placed at the outlet of the LACIS as shown in Fig. 8, so that the laser beam of the probe was entering the measuring area through the openings in a stream-shaping tube, which prevented backstreaming of room air into the flow tube.

The commercial FSSP is an optical, single-droplet counter originally designed for aircraft measurements of cloud droplet number size distributions. The fast-FSSP is a modified version of the commercial FSSP that has been developed at Météo-France (Brenguier et al. 1998). Detailed descriptions of the standard FSSP and its fast version, together with discussions of its characteristics, can be found in Dye and Baumgardner (1984), Kim and Boatman (1990), Wendisch et al. (1996), and Brenguier et al. (1998).

The principle of operation of the fast-FSSP is based on the measurement of light scattered by single droplets crossing the laser beam inside the scattering volume. The size and position of this volume are determined by geometry of beam-shaping and receiving optics of the probe. In the described experiment the scattering volume had a lateral cross section of approximately 200 μm and was located on the axis of the aerosol stream. The light scattered by a droplet into the solid angle in the near-forward direction (between about 3° and 12°) is collected by receiving optics and directed onto a photodiode. The response signal is proportional to the intensity of the light scattered by the droplet and hence is a function of droplet size. A data acquisition system classifies the signal pulse heights into 255 channels. According to the standard data procession algorithm, the actual size distribution of the droplets is then obtained from the pulse height distributions using the calibration curve. The calibration curve is a polynomial approximation of the Mie scattering function for water spheres. It was calculated from the measurements of calibration particles of known size and refraction index (borosilicate glass microspheres, Postnova analytics) and the geometry of the probe optics.

Figure 9 shows normalized channel distributions of pulses, that is, the number Ni of counted events in a given pulse height channel i divided by the total number of droplet counts Σ Ni for three different saturator temperatures Ts and a wall temperature of Tw = 3.3°C. In each case, a total number of several tens of thousands of counted events were collected over a time period of approximately 15 min for each saturator temperature. Measurements were carried out on several subsequent days for identical saturator and wall temperatures. Due to the variation of a seeding particle number concentration in each experiment and different time periods for each of the measurements, direct comparison of the pulse height distributions was not possible, and the reproducibility of the measurements has been evaluated for normalized pulse height distributions. The error bars in Fig. 9 illustrate the spread between the minimum and the maximum of the measured values in each channel, for the three independent measurements.

In the standard application of the fast-FSSP the implementation of the monotonic calibration curve is justified by the commonly broad size distribution of atmospheric cloud droplet populations and much lower requirements with regard to the size resolution. In this work, the standard size retrieval procedure implemented in the fast-FSSP was not applicable because of the narrow size distributions of the droplets produced in LACIS. To calculate the actual mean sizes of the droplets the actual response function of the probe was used. This curve is an ambiguous oscillating function of particle size, refractive index, and shape, as well as optical geometry and electronic settings of the probe. In the case of spherical nonabsorbing water droplets and for known optical and electronic settings of the probe, the exact response curve can be calculated via Mie theory (Bohren and Huffman 1983). An example of such a theoretical response function, together with the standard polynomial calibration curve, are given in Fig. 10.

It can be concluded from this figure that the calculation of droplet size distributions from the measured pulse height distributions using the theoretical response curve is not possible due to its ambiguous nature. However, it is possible to explain the observed behavior of the measured pulse height distributions taking into account the oscillations of the response curve.

To illustrate this possibility, a population of water droplets with a narrow size distribution centered around, for example, 3 μm is considered. Using the theoretical response curve as a transfer function, it is easy to show (see inset in Fig. 10) that the corresponding signal pulse heights will be distributed between the 10th and the 25th channels. Supposing then that the droplets are growing, that is, the size distribution is shifting toward larger diameters without a pronounced widening of the distribution, the resulting pulse height distribution will be moving to the region of larger channel numbers until the mean droplet size reaches the value corresponding to the first maximum of the response curve. After passing through this maximum, further growth of droplet sizes will result in a shift of the corresponding pulse height distribution toward smaller channel numbers, until the first minimum on the response curve is reached. Even further droplet growth will result in a shift of pulse height distributions toward larger channel numbers again. Similar behavior has been observed for the present set of individual measurements (see Fig. 9). Comparing the range of channel numbers over which the actual signal pulses are distributed with the corresponding part of the theoretical response curve, it is clear that the observed pulse height distributions can be explained either by assuming growth of the droplet population between the first maximum and the first minimum of the response curve (the first section of the curve with negative slope), or successive decreases of the droplet mean diameter from the second maximum toward the first minimum of the response curve. The second assumption would be in conflict with the general aspects of the condensation theory and for that reason seems to be highly improbable.

In this way, assuming general growth of the droplet sizes with growing saturator temperatures, the section of the theoretical response curve responsible for the observed pulse height distributions can be identified and used as transfer functions for recovering the mean sizes of the droplet populations. This section of the theoretical curve is also bivalent (regions before and after a local maxima/minima of the curve have to be taken into consideration) and direct inversion is still not possible. For this reason, mean sizes were determined using the following algorithm: first, normally distributed droplet populations with certain mean diameters and standard deviations were calculated. Then, using the theoretical response curve of the fast-FSSP as the transfer function, the corresponding pulse height distributions were determined. By systematic variation of the mean droplet diameter and the standard deviation of the input normal distribution, the values of these parameters were retrieved. Those producing the best fit (in terms of dispersion minimization) of the measured channel distribution of counts are then accepted as the mean size and standard deviation of the droplet populations at the LACIS outlet.

Using this procedure, mean diameters dp and standard deviations σp of the droplets at the LACIS outlet were determined as a function of the saturator temperature Ts, that is, inlet saturation, Se, and for different LACIS wall temperatures, Tw. The results are listed in Table 3. It should be noted that for the system parameters used here doubly charged particles (corresponding to dp,e = 152 nm) leaving the DMA grow up to sizes similar to those achieved for the singly charged particles (dp,e = 100 nm). The standard deviations of the droplet size distributions, for all cases except one, are at most 5% of the mean diameter, indicating a quasi-monodisperse droplet size distribution.

c. Comparison of theoretical and experimental results

LACIS was designed on the basis of extensive numerical model calculations. A first evaluation of the modeled LACIS characteristics was performed via comparison with the experimental results listed in Table 3. Figure 11 shows the comparison of theoretical and experimental mean droplet diameters dp as a function of the saturator temperature Ts, that is, inlet saturations Se for two different wall temperatures Tw.

Except for the measurements at Ts = 22°C and Tw = 2.3°C, calculated and measured droplet diameters agree within the range given by the measurement uncertainties (vertical bars). The measurement uncertainties (here the standard deviations as determined from at least three independently measured droplet sizes) were found to be approximately ±0.4 μm. Considering the more than four orders of magnitude particle–droplet volume change taking place inside LACIS, the agreement between theory and experiment is good. Nevertheless, for most of the considered cases the model seems to systematically overpredict the actual droplet diameter. This is most likely due to the influences of the latent heat released during vapor condensation on the tube walls. This effect was not accounted for in the numerical model.

5. Summary and outlook

In this paper, a new device to study the formation and growth of cloud droplets under realistic supersaturations has been introduced. It is called the Leipzig Aerosol Cloud Interaction Simulator (LACIS) and has been designed to reproduce the thermodynamic conditions of atmospheric clouds as realistically as possible and to control and adjust these conditions as accurately as needed. LACIS was constructed utilizing a moving monodisperse particle–droplet model, which was realized in the framework of the Fine Particle Model, a newly developed Eulerian particle dynamical add-on module to FLUENT 6.

Monodisperse seed particles of known size and composition were injected into LACIS and mean droplet diameters were measured at the outlet of the LACIS tube with a fast-FSSP for a variety of thermodynamic conditions. These measurements showed that

  • for constant thermodynamic conditions, measured droplet sizes were constant and stable on the time scale of hours;

  • measured and predicted droplet sizes were in good agreement; and

  • in accordance with the theoretical model, the droplet diameter could be varied by changing the critical thermodynamic conditions of the setup.

Thus, it can be concluded that the new device is capable of addressing a variety of crucial open questions concerning the dynamical physical and chemical processes affecting cloud formation under controlled laboratory conditions.

Future efforts will focus on the investigation of the influences of the latent heat released during vapor condensation on the tube walls, the development of a new optical particle counter that allows the determination of particle–droplet concentrations and droplet (equilibrium) sizes in the submicrometer size range, the evaluation of existing activity and growth laws, and the realization of an upscaled version of LACIS to match atmospheric time scales.

Acknowledgments

The authors wish to acknowledge, in alphabetical order, O. Böge, P. Glomb, A. Haudek, H. Haudek, H. Macholeth, and W. Sarwatka for their valuable contributions concerning the planning and realization of LACIS. The financial contribution of IfT (funded by the Saxony government and the federal German government) is highly acknowledged.

REFERENCES

  • Barrett, J. C., and Clement C. F. , 1988: Growth rates for liquid drops. J. Aerosol Sci, 19 , 223242.

  • Bilde, M., and Svenningsson B. , 2003: CCN activation of slightly soluble organics: Importance of small amounts of inorganic salt and particle phase. Tellus, 56B , 93182.

    • Search Google Scholar
    • Export Citation
  • Bird, R., Stewart W. , and Lightfoot E. , 1960: Transport Phenomena. John Wiley, 780 pp.

  • Bohren, C. F., and Huffman D. R. , 1983: Absorption and Scattering of Light by Small Particles. John Wiley, 530 pp.

  • Brenguier, J. L., Bourrianne T. , Coelho A. A. , Isbert J. , Peytavi R. , Trevarin D. , and Wechsler P. , 1998: Improvements of droplet size distribution measurements with the fast-FSSP. J. Atmos. Oceanic Technol, 15 , 10771090.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bunz, H., Möhler O. , Neumann K-H. , Saathoff H. , Schöck W. , and Schurath U. , 1996: The novel aerosol chamber facility AIDA: Status and first results. Proceedings of the Seventh Symposium on Physicochemical Behavior of Atmospheric Pollutants, B. Larsen, B. Versino, and G. Angeletti, Eds., EUR 17482, 673–677.

    • Search Google Scholar
    • Export Citation
  • Charlson, R. J., Seinfeld J. S. , Nenes A. , Kulmala M. , Laaksonen A. , and Facchini M. C. , 2001: Reshaping the theory of cloud formation. Science, 292 , 20252026.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chuang, P. Y., Charlson R. J. , and Seinfeld J. H. , 1997: Kinetic limitations on droplet formation in clouds. Nature, 390 , 594596.

  • Chuang, P. Y., Nenes A. , Smith J. N. , Flagan R. C. , and Seinfeld J. H. , 2000: Design of a CCN instrument for airborne measurements. J. Atmos. Oceanic Technol, 17 , 10051019.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cziczo, D. J., and Abbatt J. P. D. , 1999: Deliquescence, efflorescence, and supercooling of ammonium sulfate aerosols at low temperature: Implications for cirrus cloud formation and aerosol phase in the atmosphere. J. Geophys. Res, 104 , 1378113790.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cziczo, D. J., Nowak J. B. , Hu J. H. , and Abbatt J. P. D. , 1997: Infrared spectroscopy of model tropospheric aerosols as a function of relative humidity: Observation of deliqescence and crystallization. J. Geophys. Res, 102 , 1884318850.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DeMott, P. J., and Rogers D. C. , 1990: Freezing nucleation rates of dilute solution droplets measured between −30° and −40°C in laboratory simulations of natural clouds. J. Atmos. Sci, 47 , 10561064.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dye, J. E., and Baumgardner D. , 1984: Evaluation of the forward scattering spectrometer probe. Part I: Electronic and optical studies. J. Atmos. Oceanic Technol, 1 , 329344.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Facchini, M. C., Mircea M. , and Charlson R. J. , 1999: Cloud albedo enhancement by surface-active organic solutes in growing droplets. Nature, 401 , 257259.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fluent, 2001: FLUENT 6 user's guide. Fluent Inc.

  • Grant, L. O., and Steele R. L. , 1966: The calibration of silver iodide generators. Bull. Amer. Meteor. Soc, 47 , 713717.

  • Hagen, D. E., Schmitt J. , Trueblood M. , Carstens J. , White D. R. , and Alofs D. J. , 1989: Condensation coefficient measurement for water in the UMR cloud simulation chamber. J. Atmos. Sci, 46 , 803816.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hall, W. D., and Pruppacher H. R. , 1976: The survival of ice particles falling from cirrus clouds in subsaturated air. J. Atmos. Sci, 33 , 19952006.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hindman, E. E., 1989: Formation and optical properties of a warm cloud in a slow-expansion cloud chamber. J. Atmos. Sci, 46 , 36533663.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hudson, J. G., 1989: An instantaneous CCN spectrometer. J. Atmos. Oceanic Technol, 6 , 10551065.

  • Hudson, J. G., and Squires P. , 1973: Evaluation of a recording continuous cloud nucleus counter. J. Appl. Meteor, 12 , 175183.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kim, Y. J., and Boatman J. F. , 1990: Corrections for the effects of particle trajectory and beam intensity profile on the size spectra of atmospheric aerosols measured with a forward scattering spectrometer probe. J. Atmos. Oceanic Technol, 7 , 673680.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Knutson, E. O., and Whitby K. T. , 1975: Aerosol classification by electric mobility: Apparatus, theory, and applications. J. Aerosol Sci, 6 , 443451.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kulmala, M., Laaksonen A. , Korhonen P. , Vesala T. , Ahonen T. , and Barrett J. C. , 1993: The effect of atmospheric nitric acid vapor on cloud condensation nucleus activation. J. Geophys. Res, 98 , 2294922958.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Leaitch, R., and Megaw W. J. , 1982: The diffusion tube: A cloud condensation nucleus counter for use below 0.3% supersaturation. J. Aerosol Sci, 13 , 297319.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mahata, P. C., Alofs D. J. , and Sinnarwalla A. M. , 1973: Supersaturation development in a vertical flow thermal diffusion chamber. J. Appl. Meteor, 12 , 13791383.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Möhler, O., and Coauthors, 2002: Experimental investigation of homogeneous freezing of sulphuric acid particles in the aerosol chamber AIDA. Atmos. Chem. Phys. Discuss, 2 , 14291465.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pruppacher, H. R., and Klett J. D. , 1997: Microphysics of Clouds and Precipitation. Kluwer Academic, 945 pp.

  • Rogers, D. C., 1988: Development of a continuous flow thermal gradient diffusion chamber for ice nucleation studies. Atmos. Res, 22 , 149181.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Saxena, V. K., and Carstens J. C. , 1971: On the operation of cylindrical thermal diffusion cloud chambers. J. Rech. Atmos., 11– 23.

  • Saxena, V. K., Burford J. N. , and Kassner J. L. Jr., 1970: Operation of a thermal diffusion chamber for measurements on cloud condensation nuclei. J. Atmos. Sci, 27 , 7380.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shulman, M. L., Jacobson M. C. , Charlson R. J. , Synovec R. E. , and Young T. E. , 1996: Dissolution behavior and surface tension effects of organic compounds in nucleating cloud droplets. Geophys. Res. Lett, 23 , 277280.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Twomey, S., 1974: Pollution and the planetary albedo. Atmos. Environ, 8 , 12511256.

  • Vietti, M. A., and Fastook J. L. , 1975: Water droplet growth in an expansion cloud chamber operating at small supersaturations. J. Rech. Atmos, 4 , 181195.

    • Search Google Scholar
    • Export Citation
  • Wendisch, M., Keil A. , and Korolev A. V. , 1996: FSSP characterization with monodisperse water droplets. J. Atmos. Oceanic Technol, 13 , 11521165.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wilck, M., Stratmann F. , and Whitby E. R. , 2002: A fine particle model for FLUENT: Description and application. Proc. Sixth Int. Aerosol Conf., Taipei, Taiwan, Chinese Association for Aerosol Research in Taiwan/International Aerosol Research Assembly, 1269–1270.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

(a) Temperature, (b) saturation, and (c) particle–droplet number concentration profiles inside the flow tube (flow direction top to bottom)

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Fig. 2.
Fig. 2.

Saturation as a function of axial position z along the tube centerline (r = 0) for different inlet saturations Se

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Fig. 3.
Fig. 3.

Particle/droplet diameter as a function of axial position z along the tube centerline (r = 0) for different inlet saturations Se

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Fig. 4.
Fig. 4.

Liquid water mass fraction as a function of radial position r at the tube outlet for different inlet saturations Se

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Fig. 5.
Fig. 5.

Experimental setup of LACIS: CPC, condensation particle counter; DMA, differential mobility analyzer; and MFC, mass flow controller

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Fig. 6.
Fig. 6.

LACIS inlet

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Fig. 7.
Fig. 7.

Comparison of measured (symbols) and ideal (solid line) normalized axial velocities at the tube outlet as a function of the radial position

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Fig. 8.
Fig. 8.

Side view of the fast-FSSP location at the outlet of the LACIS tube: 1, LACIS tube with cooling jacket and extension pipe; 2, particle stream; 3, He–Ne laser beam; 4, light scattered into forward directions and redirected by the prism, 5, into the collecting optics; the laser beam is dumped by the black stop on the prism

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Fig. 9.
Fig. 9.

Normalized number of counts (number of counts in pulse height channel i divided by total number of counts in all channels) distributed over the first 45 pulse height channels. Error bars indicate the variability within three sets of independent measurements

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Fig. 10.
Fig. 10.

Actual response curve of the fast-FSSP calculated via Mie theory for spherical water droplets (solid line). The dashed line represents the smoothed curve used in the standard size retrieval procedure. Inset shows the size range relevant to the case discussed

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Fig. 11.
Fig. 11.

Calculated (lines) and measured (symbols) mean droplet diameters dp as a function of saturator temperature Ts and for two different wall temperatures Tw

Citation: Journal of Atmospheric and Oceanic Technology 21, 6; 10.1175/1520-0426(2004)021<0876:LSANSO>2.0.CO;2

Table 1.

Set of parameters for the new flow tube

Table 1.
Table 2.

LACIS operating parameters during the experiments

Table 2.
Table 3.

Mean droplet diameters p and standard deviations σp for different saturator temperatures Ts and LACIS wall temperatures Tw

Table 3.
Save