Previous observational, laboratory, and numerical model results have shown that collisional breakup should limit raindrop diameters to about 3 mm. However, observations during the Joint Hawaiian Warm Rain Project in 1985 revealed much larger raindrops, with equivalent spherical diameters of 4-8 mm.
As part of an on-going study of warm-rain processes and the development of very large raindrops, Marcin Szumowski and Robert Rauber (both of University of Illinois), Harry Ochs (Illinois State Water Survey), and L. Jay Miller completed a study of the microphysical structure and evolution of Hawaiian rainband clouds. For this they used available Doppler radar and aircraft data on three days from the Hawaiian Rainband Project (HaRP) conducted in 1990. They found that radar reflectivities can grow from -20 to 50-60 dBZ within 15 minutes in clouds whose peak updrafts were initially 5-9 m/s. Such updrafts are capable of supporting 2-5 mm diameter raindrops. The very highest reflectivities typically formed in the middle and upper parts of the clouds, within nearly erect updraft cores, then descended as the updrafts weakened. Strong updraft outflows near the trade-wind inversion transported smaller raindrops away from the cores, but the narrowness of the cores allowed larger drops to descend very near the updrafts. The absence of large numbers of smaller drops allowed the larger drops to fall without collisional breakup. Such size-sorting particle trajectories apparently allowed some drops to re-enter the updraft for continued rapid growth while falling through the updraft core of highest cloud water content. This cycle of recirculation and size-sorting of raindrops apparently allows growth to unusually large sizes. A precipitation growth model will be used in the context of Doppler-derived winds to address issues raised by this observational study.
The SCMS project was conducted during the summer of 1995 to investigate, among other things, the first formation of warm precipitation in Florida clouds. Charles Knight, in collaboration with Miller, undertook the analysis of two separate cases with these data. One case followed the complete evolution of a cumulus cloud from formation stages to about 50 dBZ, focusing on the interpretation of differential reflectivity (ZDR) return at 10-cm wavelength. The cloud already had a ZDR indicating drops at least 1 to 2 mm in diameter before reaching 0 dBZ echo intensity, which was consistent with early precipitation development from giant or ultra-giant nuclei. The other case was a continuation of interpreting the 10- and 3-cm returns from very early clouds started in the Convection and Precipitation/Electrification Experiment (CaPE) in 1991, also from Florida. The 3-cm radar data showed the existence of adiabatic cores in the early clouds up to about 1 km above cloud base, where the maximum reflectivity factor becomes too large for just cloud droplets, indicating coalescence growth. A problematical, systematic relation between the 10- and 3-cm returns in the range from about -10 to +10 dBZ was found and, so far, has resisted explanation.
The SCMS observations provided good documentation of the rates at which rain formed via coalescence in those clouds. Calculations by William Cooper (NCAR's Advanced Study Program; currently the Director), Knight, and Jean-Louis Brenguier [MMM/ATD (NCAR's Atmospheric Technology Division) affiliate scientist from Meteo France, Toulouse, France] of growth rates expected for the combined effects of condensation and coalescence are in reasonable agreement with the observations of droplet size distributions during the early stages of development in these clouds, in the ascent of unmixed parcels. These comparisons support the validity of standard representations of condensation and coalescence in these early stages. Continuing comparisons in the later stages of precipitation development show less satisfactory agreement, and appear to point to the importance of mixing and/or recirculation in the subsequent stages of warm-rain formation.
In a related study, Cooper and Roelof Bruintjes [joint MMM/RAP (NCAR's Research Applications Program) appointment] in collaboration with Graeme Mather (South African Weather Bureau, Nelspruit, South Africa) used calculations to investigate the possibility that hygroscopic seeding by relatively small particles (0.1-1.0 microns) could accelerate the coalescence process by broadening the initial droplet size distribution that forms near cloud base. These calculations were motivated by the apparently successful cloud seeding results in South Africa, accomplished through the use of hygroscopic material dispensed at cloud base. The calculations support the possibility that such seeding might accelerate the warm-rain process and thereby affect total rainfall.
The glaciation process in convective clouds encompasses many outstanding "cloud" problems being linked to cloud dynamics, cloud physics (by definition), and cloud electrification. Analysis by Daniel Breed (joint MMM/RAP appointment) of both in-situ aircraft and radar measurements made during CaPE clearly shows a sequence of events in the glaciation of Florida clouds. In these warm-based clouds coalescence occurs first, then rapid updraft development which lofts drops above the melting level, followed by freezing and riming of the supercooled drops and then production of many ice particles, via the Hallett-Mossop ice multiplication process. The speed of glaciation can be quite rapid, and there appears to be a close link between the rapid electrification of the Florida clouds and the glaciation process. When the glaciation process does not occur there is a lack of strong electrification. Two special case studies analyzed by Breed involving the NCAR sailplane and CP-2 data show details of this process, while measurements from the NCAR King Air analyzed by Breed and James Dye (joint MMM/RAP appointment) provide less detailed evidence from a wider variety of cases. The development of a ZDR "column" is a good indicator of supercooled drops, and the rapid decay of ZDR in a few minutes demonstrates how rapidly the cells can glaciate.
An examination of data from the Winter Icing and Storms Project (WISP) by Cooper, Roy Rasmussen (joint MMM/RAP appointment), and Bruintjes showed that ice nucleus concentrations, as detected in the Colorado State University (CSU) cloud chamber from processing of bag samples collected upwind of wave clouds, provided a reasonable match to the observed concentrations of ice that developed in those wave clouds. There was also reasonable agreement between observations and the concentrations detected when the same bag samples were processed through a continuous-flow ice nucleus (IN) detector at CSU. This work, collaborative with CSU's Paul DeMott and David Rogers, indicated that ice nuclei measured in these ways provide reasonable indicators of the initial ice formation, at least in wave clouds. Other studies of data from the wave clouds strengthened earlier indications that the most important heterogeneous process producing ice in wintertime clouds of the Colorado area is condensation-freezing. In particular, deposition nucleation was deduced to be insignificant, because conditions near water saturation were required for significant production of ice in wave clouds.
Knight continued to work with various collaborators on several ice crystal studies. He commenced a more systematic study of clathrate hydrate crystallization, with E. Sloan, Jr. (Colorado School of Mines) and graduate students from the School of Mines. Clathrate formation has considerable importance for applied science, because clathrates can cause blockage of oceanic or arctic petroleum pipelines. He also began a study of HNO3 adsorption on ice, using large single crystals grown in the laboratory in collaboration with Richard Sommerfeld (U.S. Forest Service, Fort Collins). Additionally, Knight and Rasmussen studied the feasibility of using an artificial snow-making machine to calibrate aircraft wing deicing fluids. Finally, with several outside collaborators Knight continued to work occasionally on fish antifreezes.
The SUCCESS project (Subsonic aircraft: Contrail & Cloud Effects Special Study), sponsored by NASA, was conducted in Salina, Kansas in April and May 1996. One objective of the program, which involved six aircraft and a host of ground-based remote sensors, was to determine the microphysical and radiative properties and formation mechanisms of contrails, and of cirrus and wave clouds. MMM scientists Andrew Heymsfield, Steve Aulenbach, Janine Goldstein, and Gregory McFarquhar participated in the field phase and provided and operated a number of instruments on a NASA DC-8 to measure the water vapor density and temperature, and properties of the ice crystal population from 5 microns to above 1 mm. Data from a wave cloud at temperatures below -60 C showed that nucleation of ice began at approximately 80% relative humidity with respect to water (~125-130 % saturation with respect to ice), consistent with earlier observations of Heymsfield and Larry Miloshevich in wave clouds at temperatures of -55 C. Surprisingly, peak relative humidities in some clouds reached nearly 100%, suggesting that droplets may form in clouds at very low temperatures under certain conditions.
A unique set of measurements was also obtained in precipitation which developed from a contrail produced by one of the research aircraft at a temperature of about -60 C. Infrared satellite images clearly showed the racetrack shape of the flight track for at least four hours following generation, at which time it covered hundreds of square kilometers. Ice crystals falling from the contrails reached sizes of 0.5 mm and the virga extended continuously for thousands of feet below the contrail generation height. Nevertheless, high concentrations of small ice crystals remained at the generation height, indicating that some mechanism existed for the generation of a wide spread in particle sizes from an initial instantaneous production of ice crystals.
Recent studies suggest that extensive shields of cirrus clouds over the equatorial Pacific "warm pool" may have a significant influence on the global climate. Heymsfield, McFarquhar, and Goldstein used microphysical measurements during the Central Equatorial Pacific Experiment (CEPEX) to understand details of the links between cloud microphysical properties and upper-tropospheric radiative heating rates. Specifically, they examined whether relatively reflective ice crystals with dimensions smaller than about 100 microns near the tops of tropical cirrus clouds are principally responsible for the high albedos observed in this region. The second and third moments of measured ice crystal size spectra (cross-sectional area and ice water content) showed that the different measures of particle size all decreased with decreasing temperature and increasing altitude. While small ice crystals located near the cloud tops could account for moderate albedos of approximately 0.3, the predominantly large particles in the lower, warmer parts of the cirrus contain at least an order of magnitude greater mass, and are dominant in producing the high observed albedos. These conclusions were further supported by radiative transfer calculations and the combined use of lidar and radiometer data acquired on-board a NASA ER-2 which overflew the in-situ aircraft.
Based on these observations, Heymsfield and collaborators produced a parameterization of the average ice crystal size distributions as a function of temperature and ice water content (IWC). This parameterization has potential applications for cloud resolving models and climate studies because it is mass conserving and easily integrable, and it provides accurate estimates of the radiative properties of both small and large ice crystals.
A balloon-borne cloud particle sampling instrument was described by Miloshevich and Heymsfield in the FY95 Annual Scientific Report. This "ice crystal replicator" measures vertical profiles of cloud microphysical properties, collects particles at least as small as 10 microns, and preserves the detailed characteristics of ice crystal shapes needed to better understand the solar and infrared radiative properties of ice clouds. The acquisition of detailed cloud microphysical measurements in the vertical is particularly important for validating techniques that retrieve cloud microphysical properties from ground-based or satellite remote sensor measurements, as well as for understanding the formation and development of cirrus clouds. This year the quantitative capabilities of the replicator for measuring particle size distributions were assessed using a combination of windtunnel measurements and numerical fluid dynamics modeling to determine the instrument's size-dependent collection efficiency function. The fraction of the ambient particle distribution collected was found to increase slowly from 3% of 5 micron particles to 11% of 20 micron particles, then increase rapidly to 59% for 40 micron particles, with a slow increase in collection efficiency for larger particle sizes.
A new detector for cloud condensation nuclei (CCN) and ice nuclei (IN) was constructed and flown by Cooper, and Larry Radke (ATD) in its first field experiment, the SUCCESS experiment in April-May 1996. The instrument uses a controlled expansion to produce supersaturated conditions, then counts droplets that form under these conditions as they pass through an optical counter. After air is humidified to near 100% relative humidity, small pressure changes are produced by flow through capillaries, and capillaries of different sizes are switched into the line to produce a sequence of supersaturations. The instrument can also be operated at low temperature, where it can detect nuclei active as condensation-freezing or deposition nuclei. A future application, not yet operational, is to provide low-temperature operation for detection of the homogeneous freezing of solution droplets, as is thought to occur in the formation of cirrus and perhaps of contrails.
To understand the global distribution of aerosol sulfate, sulfur dioxide, and dimethyl sulfide in the troposphere and to learn what impact aerosol sulfate has on the climate, a global sulfur model was developed by Mary Barth (joint appointment between MMM and NCAR's Atmospheric Chemistry Division and Climate and Global Dynamics Division), Philip Rasch (CGD), and Jeffrey Kiehl (CGD). Comparison of model results with observations showed fairly good agreement in polluted locations such as the Eastern United States and Western Europe. Comparison of model results to observations in remote locations was undertaken. Analysis of the results showed that conversion of sulfur dioxide to sulfate by aqueous chemistry is by far the dominant process for creating aerosol sulfate in both winter and summer. Preliminary analysis of the influence of source regions showed that Asian sources significantly contribute to the aerosol loading in the northern hemisphere. Work will continue to understand the processes that control the seasonality of the sulfate distribution. Furthermore, the sulfur model will soon work interactively with the Community Climate Model (CCM) to examine both radiative effects on sulfur chemistry and aerosol-sulfate-loading effects on precipitation.
Part A of STERAO (Stratospheric Tropospheric Experiment: Radiation, Aerosols and Ozone), Deep Convection and the Chemical Composition of the Upper Troposphere and Lower Stratosphere, was conducted in northeastern Colorado, June through August 1996. Led by Dye, Steven Rutledge (Colorado State University), and Adrian Tuck [Aeronomy Laboratory, National Oceanic and Atmospheric Administration (NOAA)], Part A was designed to investigate the transport of chemical constituents by deep convection, the production of NOx (comprised of NO and NO2) by lightning, and the effect of deep convection and overshooting tops on the water vapor and nitrogen budgets of the upper troposphere and lower stratosphere. It was unique in combining and coordinating extensive chemistry, air motion and electrical measurements in and around thunderstorms, and included participants from several universities and national laboratories. The major research facilities in the experiment included: (1) the NOAA P3 aircraft to extensively characterize the chemical composition of the boundary layer and middle troposphere and to determine air motions in storms using a tail Doppler radar; (2) the University of North Dakota Citation jet aircraft to make chemical, microphysical and air motion measurements in the upper troposphere in and around thunderstorm anvils; (3) the CSU/CHILL [Colorado State University/University of Chicago/Illinois State Water Survey] multiparameter Doppler radar to remotely investigate microphysical properties of storms and to coordinate with the P3 in Doppler radar studies; (4) the French Office National d'Etudes et de Recherches Aerospatiales (ONERA) lightning interferometer to provide 3D maps of lightning inside storms; (5) the National Lightning Detection Network for cloud-to-ground lightning; (6) two fixed and one mobile electric field charge meters to detect total lightning and distinguish between intra-cloud and cloud-to-ground lightning; and (7) the SSSF (Surface Sounding Systems Facility) Mobile CLASS (Cross-Chain Loran Atmospheric Sounding System) soundings to characterize the environment in the project area. Forecasting for the project was provided by the NOAA Forecast Systems Laboratory, with input from the MM5 (NCAR/Pennsylvania State University Mesoscale Model, Version 5) and Goddard Cumulus Ensemble models.
The project was blessed with frequent thunderstorms and obtained six excellent joint-aircraft cases and six single-aircraft cases, ranging from isolated multicells to a supercell. The CSU/CHILL radar and ONERA lightning interferometer operated throughout most of the summer, providing a wealth of information on the co-evolution of storm structure and electrical activity for a wide variety of storm types. The analysis is in a very early stage at this point, but preliminary results obtained by Brian Ridley (ACD) from the Citation show a clear signature of significant enhancements of NOx in the upper parts of the storm anvils, much like Ridley and Dye had shown in an exploratory experiment in New Mexico in 1989.
Dye, with Darrel Baumgardner and Bruce Gandrud (both of ATD), completed a study of an Antarctic polar stratospheric cloud (PSC) which was observed with the NASA ER-2 aircraft during the Airborne Southern Hemisphere Ozone Experiment (ASHOE) flown from Christchurch, New Zealand in July 1994. The analysis and comparison with the microphysical model of Katja Drdla (former ASP graduate fellow; currently with the University of California, Los Angeles), shows that much of the cloud was composed of ternary solution droplets of HNO3/H2SO4/H2O, but that in portions of the cloud solid particles--perhaps nitric acid trihydrate--were also present. These observations, coupled with previous observations by Dye and colleagues in Arctic PSCs, help to unify our understanding of some PSC formation processes and also show that liquid PSCs can occur in both hemispheres. Additionally, they provide evidence that mixed-phase PSCs containing both liquid and solid particles can co-exist. Although much progress has been made in understanding PSC formation processes in the last few years, there is still major uncertainty about the solidification mechanism of PSC particles, including supercooled sulfuric acid droplets.