Search Results

You are looking at 1 - 10 of 131 items for

  • Author or Editor: Andrew J. Heymsfield x
  • All content x
Clear All Modify Search
Andrew J. Heymsfield

Abstract

The growth of ice particles through aggregation is investigated for seeded clouds using currently available field data and a numerical particle-growth model. Observations indicate that the aggregation process is fairly common, even when moderate liquid water contents, ~0.5 g m–3, are available for particle growth through accretion. The modeling study suggests that certain temperature ranges are especially conducive to aggregate formation.

Full access
Andrew J. Heymsfield

Abstract

The physical characteristics of graupel particles were investigated from in in situ collections of ice particles sampled in first-echo summertime cumulus congestus clouds in northeastern Colorado with the NCAR/NOAA sailplane operating during the National Hail Research Experiment. Ice particles were collected in vials containing silicone oil. Each particle was photographed at several different orientations and then melted to determine its equivalent diameter. This permitted the mass and axial dimensions to be determined and the density, axial ratios and terminal velocity to be estimated for each of 125 particles.

The mass and terminal velocity of graupel particles were found to be considerably lower than those of equivalent diameter ice spheres. Best-fit equations to the mass-diameter and terminal velocity-diameter data were computed. Graupel densities were typically lower than 0.5 g cm−3 systematic differences were noted between the densities of lump and conical graupel. Calculations indicated that the results reported in this paper may not he applicable to thunderstorms with warm cloud-base temperatures.

Measurements of the microstructure of the penetrated cumulus congestus clouds permitted several generalizations to he made on the growth of conical graupel and enabled previous theories of conical graupel formation to be examined. Conical graupel formation through riming of small single crystals appears to be the dominant mechanism operating in the clouds investigated.

Full access
Andrew J. Heymsfield

Abstract

A laboratory experiment was initiated to grow the banded columnar crystals found in cirrus clouds and to determine if they nucleated through a freezing nucleus. Banded columnar crystals were collected in cirrus clouds and also grown in a laboratory cold box over a temperature range of −20 to −46C. They were observed to grow from a frozen droplet through several distinct crystalline transitions. The first distinct crystalline form observed following the growth of a frozen droplet was a polyhedral crystal. With further growth, the polyhedral crystal developed end plates, and it appeared to be a “double plate.” Then the separation between the plates nearly filled in to leave the characteristic band. Certain columns (−6 to −10C) and plates (−10 to −20C) were also observed to grow from a frozen droplet and showed similar crystalline transitions.

Full access
Andrew J. Heymsfield

Abstract

Mesurements obtained in a thin, cirriform cloud that formed in the temperature range of −83° to −84°C and an altitude range at 16.2 to 16.7 km are presented. Implications of the results for particle generation in polar stratospheric clouds developing with similar conditions are discussed.

Full access
Andrew J. Heymsfield

Abstract

This is Part I of a study that characterizes several bulk properties of ice particle populations sampled in midlatitude and tropical cirrus and deep stratiform ice clouds, for the purpose of developing an understanding of how particles evolve in ice clouds and to derive empirical and analytical relationships that describe microphysical properties for use in cloud and climate models. The effort focuses on describing the microphysical properties of ice cloud layers in the vertical. The size distribution data and particle imagery were obtained during Lagrangian spiral descents and balloon-borne ascents through cloud layers that formed in association with synoptic-scale lifting (midlatitude) and deep convection (Tropics). Temperatures ranged between −20° and −63°C for the midlatitude clouds and between 0° and −50°C for the tropical clouds. Optical depths spanned the range 0.4–7 for the midlatitude clouds and 20–30 for the tropical clouds.

This part of the study characterizes median mass diameter (D m) and median fall velocities (V m) for the more than 2000 ensembles or particle size distributions (PSDs) examined. The D m and V m increase downward from cloud top to base, with the smallest D m and V m values found in the coldest (midlatitude) clouds and the largest values found in the warmest (tropical) clouds. The range of sizes that dominate the ice water content, and the associated range of particle fall speeds, are characterized in terms of D m and V m.

The V m are represented in terms of D m and the slopes (λ) of gamma distributions fitted to the particle size distributions. The V m values increase with D m and decrease with λ in a predictable manner. The magnitudes of the changes in V m that result from differing ambient pressures between 250 and 1000 hPa are quantified. The observations are generalized so that the results can be extended to different pressure levels and other particle size distributions.

The coefficients γ and β in the power-law relationship V t = γD β fitted to the individual spectra are found to be inversely related to D m. Many earlier studies have derived these coefficients from measurements at the surface. The wide variability noted in these coefficients may partially be attributed to variations in the D m values of the populations considered. The relationship of the γ and β coefficients found for particle ensembles at the surface to those at the pressure levels of ice clouds are derived.

Full access
Andrew J. Heymsfield

Abstract

This is the second part of a study that characterizes several bulk properties of ice particle populations sampled in synoptically generated midlatitude and convectively generated tropical ice clouds, for the purpose of developing empirical and analytical relationships that describe microphysical properties for use in mesoscale and climate models. The purpose of this paper is to examine the interrelationships between the mass, area, and fall velocity properties of the particle size distributions, and the dependence of these properties on temperature.

Gamma distributions of the form N = N 0 D μ e λD are fitted to the measured particle size distributions (PSDs) over sizes (D) from as small as 10 μm to as large as 1.5 cm. Exponential distributions (μ = 0) are also fitted to the PSD. The intercept parameter N 0 and the slope λ are directly related, and decrease monotonically with increasing temperature. The μ values for the gamma fits tend from positive values at large λ to negative values at small λ. The maximum measured diameter D max increases with decreasing λ. The N 0 values from the midlatitude clouds are about an order of magnitude lower than those for the tropical PSDs at the same temperatures.

Bulk properties are derived from the fitted PSDs. The ice water contents (IWC) are about an order of magnitude higher for the tropical than for the midlatitude clouds. The median mass diameter (D m) and the effective diameter (D e) each increase with temperature, and are found to be related to each other.

Several aspects related to the modeling of ice particle sedimentation in general circulation models (GCMs), and the relationship of these velocities to other bulk properties, are investigated. On average, the median mass-weighted terminal velocity (V m) increases weakly with temperature. Correlations between V m and IWC are also weak. It is found that for a given particle ensemble, most of the ice mass is contained within a relatively narrow range of fall velocities, although the values of V m can be appreciable. Calculations reveal that the fallout of particles that dominate the extinction cannot be ignored, except at temperatures below −50°C. Also, the effective diameter is found to be strongly related to the ensemble mean V m, perhaps allowing the two variables to be linked in GCMs.

Full access
Andrew J. Heymsfield

Abstract

The rates of development of graupel and hail in High Plains storms are calculated based on mechanisms for the growth of particles of various types. In the first part of this study, planar crystals, aggregates, graupel particles and frozen drops grow in a region of updraft specified by a one-dimensional parcel model. Their rates of development are calculated from a model which considers their growth through accretion and diffusion, “dry” or “wet” growth, and melting. At most initial embryo diameters and for liquid water content of 1 g m−3, frozen drops grow most rapidly, followed in order by graupel, aggregates and planar crystals. Particle growth rates over the range of liquid water contents from 0.5 to 2.0 g m−3 can be estimated from those given for 1 g m−3 through the use of a simple parameter.

In the second part of the study, the processes of hail production are deduced by calculating the times required for each of the above embryo types to develop into hail from the time of nucleation over the range of typically observed liquid water contents. These times are then compared to the typically observed cell lifetimes to infer the predominant hail growth mechanisms. The calculations suggest that hail is not likely to be produced when particle growth is confined to one region of updraft. A particle which grows in the peripheral regions of a storm through diffusion and then aggregation, and is then introduced into a region of updraft is the most likely candidate to become a hailstone in a High Plains storm. An unrimed aggregate which melts completely to form a drop and then subsequently freezes and develops into hail is the most likely growth sequence that produces hail from frozen drops.

Full access
Andrew J. Heymsfield

Abstract

Aircraft microphysical measurements in ice clouds associated with warm frontal overrunning systems, warm frontal occlusions, closed lows aloft and the jet stream were combined with Doppler radar measurements in four case studies. Good agreement was obtained between aircraft calculations of the radar reflectivity factor and air velocity, and radar measurements and calculations of these parameters. Vertical velocities typically ranged from 10 cm s−1 in warm frontal overrunning systems to in excess of 50 cm s−1 in clouds associated with a closed low aloft, longitudinal rolls and isolated convective cells. Ice crystal seeding in trails emanating from longitudinal rolls were measured to extend over 7 km in the vertical and over 100 km in horizontal distances. Several general results were deduced from the aircraft measurements. Vertical velocities generally in excess of 50 cm s−1 at temperatures lower than –5°C were found to be necessary for liquid water occurrence in deep stratiform ice clouds. Water saturation was not necessary for nucleation to occur. The ice water content and ice crystal concentration, parameterized in terms of the vertical air velocity and temperature, were found to be directly dependent on the vertical velocity. Ice crystal concentrations were found to he 2–4 orders of magnitude higher than ice nuclei concentrations at temperatures warmer than –15°C.

Full access
Andrew J. Heymsfield

Abstract

A technique is described for simulating the development of particles in a storm when data on the internal composition and wind-field structures are available. Simulations such as these can be used to investigate particle growth mechanisms.

The method is used to simulate particle development in a hailstorm that occurred in northeastern Colorado on 22 July 1976. Wind fields derived from triple-Doppler radar scans during six periods are used for the calculations and the temporal evolution of the winds during a measurement period is considered. Particles of different types and sizes are initialized at positions throughout the storm at the beginning of the analysis period from radar data using correlations from the in-situ measurements. Particles are “nucleated” within updrafts at later times. Particle growth and sedimentation are calculated according to the habits of particles as they are carried through the storm until they fall to the ground. The liquid water content and drop-size distribution at positions along particle trajectories are calculated from the vertical air velocities. Input parameters of the calculations are varied for the purpose of sensitivity analyses. A data set was compiled of information on positions, sizes, terminal velocities, and other parameters during the development of each of more than 130 000 initialized particles. Information from this data set was compared with available radar, surface and in-situ observations to verify the model inputs and evaluate the simulations.

The calculated manner of particle development in the storm compares favorably with the observed radar, surface and in-situ observations. Several discrepancies between the calculations and the observations are attributed primarily to inadequacies of the wind-field data. Underestimates in the liquid water content could also have accounted for some discrepancies.

Full access
Andrew J. Heymsfield

Abstract

The processes of development of graupel and had which fell to the ground from a storm in northeastern Colorado on 22 July 1976 are investigated over a one-hour period. The growth and trajectories of 130 000 particles of different types and sizes are calculated in the measured three-dimensional wind fields. The growth conditions that these graupel and hail (spherical ice particles smaller than and larger than 1 cm, respectively) experience are presented and their trajectories are described.

Particles which become hail are those whose terminal velocities are nearly equal to the vertical velocities of the air parcels in which they develop. This enables them to fall into regions of relatively high liquid water content in the main updraft cores. The terming velocities of the particles which become graupel are not as well-matched to the parcel velocities; particles grow with lower liquid water contents.

Embryos of the graupel and hail are found to be aggregates (snowflakes) of 0.5–1.5 cm in diameter. Embryo formation takes place in a 1 km wide region of divergence located along the forward portion of the storm where the mid-level air is flowing around the updraft cores Aggregates becoming hail are ingested by feeder cells located adjacent to the main updraft cores and are then carried into newly forming cells which become the main updraft region, wherein they complete their development. Aggregates becoming graupel are ingested directly by the main updrafts.

Mechanisms by which particles are selected to become graupel and hail embryos are related to the position at which they develop in the storm, the processes by which they initially develop, and the stage of development of the main updraft core at the time at which they begin to grow in that region.

Full access