The NCAR Geophysical Turbulence Program (GTP) within MMM is comprised of Jackson Herring, Robert Kerr, Donald Lenschow (joint appointment with ATD), Chin-Hoh Moeng, and Peter Sullivan. Of these, the research of Herring and Kerr is primarily devoted to basic research in turbulence theory and direct numerical simulation (DNS). DNS is distinguished from Large Eddy Simulation (LES) in that the former should produce accurate numerical solutions to the Navier-Stokes equations, whereas the latter is a parameterization procedure designed to produce practical results for geophysical scale flows. The disadvantage of DNS is, of course, that the Reynolds number of the flow is modest. DNS and the statistical theory of turbulence should serve as verification of LES, pointing the way towards improvement.
In addition to permanent staff members, visitors in FY 95 included Joseph Werne (jointly supported by GTP, SCD, and MMM) and Yoshi Kimura (Nagoya University). Other collaborators not resident at NCAR were Axel Brandenburg (ASP and HAO postdoctoral fellows) and Julian Domaradzki (University of Southern California).
Research in this group during the past year focused on the following topics: (1) eddy diffusion in stratified and rotating turbulence, (2) improvement of LES through a critical examination of detailed turbulence transfer dynamics as revealed through DNS, (3) a continuation of the DNS convection studies, designed to understand the large-scale organization of convective systems, and (4) the use of DNS and statistical theory to examine the limits of accuracy of LES and possible methods of improving its traditional analytic form.
This project is a continuation of the research by Kimura and Herring reported in last year's ASR. The purely stratified case has by now been submitted for publication in the Journal of Fluid Mechanics. Although the effects of stratification can be simply parameterized by Langevin models of the Csanady type, for decaying turbulence, stratification limits eddy diffusion far more than earlier theories would suggest. For purely stable stratification, the vorticity field is organized into scattered pancake structures. As the rotation increases, the orientation of the vortex patches changes from horizontal oblate to prolate vertical patches.
Work by Kerr, Domaradzki, and Gilles Barbier (Ecole Polytechnique, France), to appear in the January issue of the Physics of Fluids, uses ideas from vortex dynamics to analyze the subgrid terms in detail and proposes some new dynamic models based on this analysis. The analysis supports the stochastic backscatter concept of dividing the subgrid terms into a nearly Smagorinsky type of eddy viscosity and a term that has equal parts of forwards and backscatter, except now this term has structure rather than being stochastic.
The goals of these studies are to understand the scaling regimes and how they relate to flow structures, and their dependence on Prandtl number (viscosity/diffusivity). Also of interest is the low Prandtl number regime for applications to astrophysical convection. To these ends simulations have been done for a range of Rayleigh numbers at Prandtl number 0.07, 0.7 and 7.0. The simulations have no-slip tops and bottoms and free-slip sidewalls to more closely represent convection in an enclosed container in the laboratory. Some intriguing similarities between measurements of atmospheric convection and simulated structures have been identified (work in progress by Eleanor Praskovskaya (visitor, unaffiliated), Herring, Robert Grossman (visitor, University of Colorado, CU), and Kerr).
Kerr, Herring, and Brandenburg (Chaos, Solitons and Fractals, 1995) have successfully reproduced the dependence of the Nusselt number or normalized heat flux upon Prandtl number identified by the experiments of Ciliberto (such results are at variance with the recent theories of Shraiman and Siggia). There is circumstantial evidence for a change in flow structure, as in water experiments of Krishnamurti (1994), between Rayleigh numbers one and ten million. The data to make this more quantitative has been stored for future analysis. Their greatest analysis effort so far has been to attempt to reproduce the recent experimental paper of Belmonte, Tilgner and Libchaber (1994). In the experimental paper, an innovative way of determining a new boundary layer is found by analyzing spectral cutoffs. The contention is that this new boundary layer has the theoretically-predicted scaling of the velocity boundary layer, which has never been directly measured before. Simulations and analysis by Kerr, Brandenburg, and Herring have reproduced the experimental result, validating the experimental technique. But they show that the interpretation in terms of the velocity boundary layer is unwarranted.
Considered from the statistical point of view, LES may be viewed as a problem in estimating the errors caused by using a system with reduced degrees of freedom, instead of the full dynamics. Thus, an eddy viscosity is required for the reduced system in order to make its spectra identical to those of the full system. Using statistical theory, Herring has estimated the correlation between the reduced system and the full system. Such correlation is less than fifty percent.
James McWilliams (University of California, Los Angeles, UCLA), Sullivan, and Moeng analyzed large-eddy simulations of the phase-averaged equations for oceanic currents in the surface Planetary Boundary Layer (PBL), where the averaging is over high-frequency surface gravity waves. These equations have additional terms proportional to the Lagrangian Stokes drift of the waves, including vortex and Coriolis forces and tracer advection. For the wind-driven PBL, the turbulent Langmuir number, La (the square root of the ratio of friction velocity, U, to the Stokes drift, Us), measures the relative influences of directly wind-driven shear and Stokes drift. This study focuses on equilibrium solutions with steady, aligned wind and waves and a realistic La (= 0.3). The mean current has an Eulerian mass transport to the right of the wind and against the Stokes drift. The vertical turbulent fluxes of momentum and tracers are enhanced by the presence of the Stokes drift, as are the turbulent kinetic energy and its dissipation and the skewness of vertical velocity. The dominant coherent structure in the turbulence is a Langmuir Cell, which has its strongest vorticity aligned longitudinally (with the wind and waves) and trapped near the surface on the scale of the Stokes drift profile. Associated with this are downwind surface convergence lines connected to interior circulations whose horizontal divergence axis is rotated about 45 degrees to the right of the wind. The horizontal scale of the Langmuir Cells expands with depth, and there are also intense motions on a scale finer than that of the dominant cells very near the surface. In a turbulent PBL, Langmuir Cells have irregular patterns with finite correlation scales in space and time, and they undergo occasional mergers.
Ching-Long Lin (visitor, Stanford University), collaborating with Moeng, Sullivan, and McWilliams, studied the dynamics of a neutrally stratified PBL flow through a large-eddy simulation, which includes surface roughness, Coriolis force, and a capping inversion. The budget of fluctuating pressure is investigated to provide information on the development of horseshoe vortices and to evaluate the applicability of pressure minima in identifying vortical structures. Quadrant analysis and flow visualization show that a low-speed momentum flux (or ejection event) is the dominant feature in the boundary layer except very close to the surface where sweep events dominate. The formation of vortical structures is observed to be associated with the collision of low-speed and high-speed momentum fluxes. A four-dimensional (space and time) quadrant 2 conditional sampling technique was developed to determine the statistical behavior of these dominant ejection events, focusing on their shape, strength, lifetime, and origin.
Lenschow, Steven Oncley (ATD), Kenneth Davis (visitor, University of Colorado, CU), Qing Wang (ASP postdoctoral visitor), Jakob Mann (Riso National Laboratory, Denmark), and Teresa Campos (ATD) have used data from NCAR Electra flights over the Canadian boreal forest during the Boreal Ecosystem-Atmosphere Experiment (BOREAS) in the summer of 1994 to study air/surface interactions and boundary layer structure. They examined the effects of lakes on both the mean and turbulent structure of the atmospheric boundary layer. This region is dotted with lakes of various sizes. They found that in light wind cases even a 10 km wide lake with a surface temperature within a couple of degrees of the mean surface temperature of the surrounding land can generate drastically different fluxes and variances than the surrounding land. They also studied differences in surface characteristics across the boreal forest and their influences on fluxes of latent and sensible heat, ozone, and carbon dioxide. They found that the southern areas, which contained more deciduous trees, were photosynthetically more active than the predominately coniferous forests of the northern area. Photosynthesis over the tundra region was less than half that over the forests during the midsummer period of maximum activity.
A water vapor differential absorption lidar (DIAL), developed and operated by Gerhard Ehret, Andreas Giez, Christoph Kiemle and Hans-Georg Schreiber (German Aerospace Research Agency, DLR), was also flown on the Electra. Useful data were obtained from the lidar on all BOREAS flights. The aerosol backscatter data are useful for estimating the top of the boundary layer and delineating structures within the boundary layer. Data from pre-BOREAS test flights over the Imperial Valley and Salton Sea in California were analyzed, showing that the lidar water vapor measurements compared very well with in situ data, and that what appeared in the lidar data to be a sea breeze was revealed by in situ data to be due to a larger-scale flow pattern.
K. Davis and Gordon Bonan (CGD) began a comparison of BOREAS tower fluxes with Bonan's Land-Surface Model (LSM), a surface-atmosphere model used in the CCM2.
Measurements of fluxes from an intermittent-sampling system, operated during BOREAS on the NCAR Electra, were analyzed by K. Davis and William Cooper (joint appointment with ATD). These measurements, described further under Focused Programs, The NCAR Global Tropospheric Chemistry Program (GTCP), were used to estimate fluxes of CO2, water vapor, and isoprene and to evaluate the intermittent-sampling technique.
K. Davis collaborated with colleagues from ACD on three other field projects aimed at characterizing forest emissions of isoprene and terpenes. K. Davis derived flux estimates of these species from tethered balloon mixing ratio measurements collected over Georgia, Alabama, and Tennessee. These flux estimates were the largest scale of a suite of flux measurements, ranging down in scale to tower, branch, and leaf-level estimates. The suite of flux estimates showed decisively that the current algorithms used in EPA air quality models underestimate isoprene emissions by as much as a factor of five.
The NCAR PBL Model Evaluation Project has finished its first phase of evaluating the system performance of six existing PBL parameterization schemes against a LES database, which consists of ten simulated PBL flows over a range of shear and buoyancy forces. This work was summarized by Keith Ayotte (visitor, Commonwealth and Industrial Research Organization, CSIRO, Australia), with co-authors Moeng and Sullivan; Joseph Tribbia, Scott Doney, and William Large (all of CGD); John Wyngaard and Martin Otte (Pennsylvania State University); McWilliams; Anders Andren (Uppsala University, Sweden); and Bert Holtslag (Royal Netherlands Meteorological Institute, KNMI). This paper describes the methodological framework, the LES solutions, and the evaluation results. The main result is the recognition that all of the examined PBL parameterizations have difficulty reproducing the entrainment fluxes at the top of the PBL, as given by the LES, in most PBL regimes. The sensitivity of the PBL models to vertical resolution was also assessed.
In supporting the Global Energy and Water-Cycle Experiment (GEWEX) Cloud System Study (GCSS), Moeng is working with an international group of scientists to examine the accuracy of large-eddy simulation of the stratocumulus-topped PBL, which involves many more physical processes than the clear PBL, and hence is more difficult to simulate. In a journal paper, Moeng, along with 12 international researchers, intercompared simulations of a stratocumulus-topped PBL from ten three-dimensional LES codes and four two-dimensional cloud resolving models. This study identified the major differences among the simulations, and pointed out possible sources for such differences. It is shown that the entrainment rate varies significantly among the ten LESs, hence a follow-up smoke-cloud experiment to specifically study the entrainment-rate process was proposed. This follow-up experiment has been simulated and discussed in a Netherlands workshop, and the results are now being summarized.
In collaboration with Moeng and Sullivan, Jeffrey Weil (visitor, University of Colorado) used velocity fields from LESs to track passive "particles" in a Lagrangian dispersion model. Calculations of the mean concentration field and displacement statistics have been made for strongly and moderately convective boundary layers (CBLs) in which the stability index h/L= 110 and 14, respectively; h is the CBL height and L is the Monin-Obukhov length scale. Results for the strong CBL showed good agreement with the convection-tank experiments of Willis and Deardorff, but those for moderate convection showed appreciable differences from the experiments. The latter results (h/L = 14) were attributed to the greater surface shear and turbulence dissipation rate; they imply that shear effects on dispersion extend to h/L values much larger than suggested by Willis and Deardorff.
Margaret LeMone, Moeng, and Lenschow, along with Mingyu Zhou (visitor, National Research Center for Marine Environmental Forecasts, China), Grossman, and Robert Zamora (NOAA) continued their study of the baroclinic boundary layer using data from the Stormscale Operational and Research Meteorology-Fronts Experiment Systems Test (STORM-FEST) boundary layer array. They are concentrating on five cases in which the NCAR King Air sampled the boundary layer for 3-4 hours around noon, when the boundary layer depth varies little. The objective is to document and study the sources of the wind profile under these conditions. Comparisons among aircraft, radiosonde, and number of good data points at each time from five 915 MHz profilers (a proxy for reflectivity) provided consistent measures for the depth of the convective mixed layer. Profiler winds were different from aircraft winds due to an unknown mix of local-topography and ground-clutter effects. L. Jay Miller is investigating radar data as an independent estimate of the wind profile; surface, radiosonde, aircraft, and model data will provide estimates of the thermal wind. The days examined are sufficiently convective that the turbulence kinetic energy budget is affected little by the baroclinity.
LeMone, William Blumen (CU), Grossman, and John Pflaum (NOAA) continued developing the Cooperative Atmosphere Surface Exchange Site (CASES), a mixture of long-term (at least three years) measurements to document the interaction between the upper few meters of the earth's surface and the lower atmosphere; and shorter-term measurements more in the mold of the traditional field program, but supporting multidisciplinary as well as disciplinary goals. At a workshop in Wichita, Kansas, in February, around 100 scientists and federal agency respresentatives divided themselves into four groups (weather and climate, hydrology, ecology, and chemistry) to discuss the potential use for such a facility in their discipline and jointly with other disciplines. These needs were distilled into an array design and data management strategy. The results of the Wichita meeting formed the skeleton of the prospectus document, which was mailed to several federal agencies in June, and discussed with the agencies in Washington, D.C. in July. Under the direction of Marvin Wesely, Argonne National Laboratory is now setting up a long-term boundary-layer array with a subset of the desired instrumentation in the southern portion of the CASES site, which coincides with the Walnut River Basin east of Wichita, Kansas; NOAA will support John Pflaum to continue CASES development, which will include coordinating principal investigator responses to anticipated announcements of funding availability.
Research in physical meteorology in FY 95 concentrated on precipitation formation processes (especially coalescence), ice origins, the roles of aerosols in tropospheric and stratospheric clouds, and cloud electrification. Studies of aqueous chemistry in clouds and studies seeking to deduce cloud properties from satellite observations are also underway, and the research program includes development of new instrumentation to support field observations. Close ties have developed between these studies and related ones in the NCAR Clouds in Climate Program (CCP).
Several MMM scientists (Charles Knight, Cooper, Miller, and Ilga Paluch), Jean Louis Brenguier (affiliate scientist visitor, Centre National de Recherches Meteorologiques, CNRM, Toulouse, France), Darrel Baumgardner (ATD), and scientists from the Universities of Illinois, Wyoming, and Chicago, and New Mexico Institute of Mining and Technology, and other institutions, participated in the Small Cumulus Microphysics Study (SCMS), June-August 1995, at Cape Canaveral, Florida. The main goal was to study the development of the cloud droplet size distribution and the onset of coalescence growth in small cumulus clouds that were entirely warmer than freezing. The NCAR C-130, the Wyoming King-Air, and the French Merlin were flown in missions coordinated with each other and with the NCAR dual-wavelength CP-2 radar scanning the small clouds in detail. Many good cases with clouds located close to the radar were obtained.
The dual-wavelength capability and high sensitivity of the radar in SCMS made it possible to distinguish the Rayleigh and Bragg scattering contributions to the radar reflectivity as precipitation developed, and so provided unique documentation of the time required for precipitation development. When used in combination with the observations of droplet size distributions and aerosol concentrations measured by the aircraft, these observations will support a detailed comparison between predicted and observed rates of coalescence, and so will provide a critical test of current understanding of the warm-rain process.
In a related study, Cooper and Roelof Bruintjes (joint appointment with RAP), in collaboration with Graeme Mather (CloudQuest, South Africa), used calculations of coalescence rates to investigate possible effects of large Cloud Condensation Nuclei (CCN) on the coalescence process. The calculations support broadening of the droplet spectrum and the acceleration of coalescence by the addition of hygroscopic particles with diameters of about 1 micron. To investigate the role of such particles further, aerosol measurements in this size range were emphasized in the SCMS field program.
Remote sensing of cloud liquid water would be valuable for studies of entrainment and vertical flux of cloud water. If a consistent relationship between radar reflectivity factor and cloud liquid water content could be established, radar might be used for this purpose. Earlier efforts to relate cloud liquid water to radar reflectivity factor (derived from droplet size measurements) have revealed much scatter in the data. For a constant reflectivity factor, the cloud liquid water typically varied by at least a factor of two, and in some cases by an order of magnitude. Paluch, Knight, and Miller now report that the scatter can be much less, corresponding to an uncertainty in liquid water of only about thirteen percent, if observations are limited to small cumulus clouds in which coalescence growth of water drops is not significant, provided that cloud base conditions and altitude above cloud base are known. The close correlation between radar reflectivity factor and cloud liquid water suggests that radar will be useful for mapping cloud water in space and time in nonprecipitating cumuli. Furthermore, the observed relation between cloud liquid water and reflectivity factor has significance for understanding the development of the cloud droplet size spectrum in cumulus clouds under the influence of entrainment and mixing. While the observed relation cannot be explained in terms of perfectly homogeneous or inhomogeneous mixing, it can be explained in terms of a simple model of a mixed region that is locally homogeneous, but highly nonuniform over 100-m aircraft sampling distances. This study was based on cloud droplet size spectra recorded by a Forward Scattering Spectrometer Probe (FSSP) in developing cumulus over the Florida coast during the Convection and Precipitation/Electrification Experiment (CaPE). To test the liquid water-reflectivity factor relationship derived from the FSSP, coordinated radar and aircraft observations were made in Florida this summer during SCMS.
In June of 1995, a major upgrade to the NCAR sailplane's data acquisition system was completed by Reif Heck (ATD) and other personnel at the Research Aviation Facility. Daniel Breed and Wolf-Dietrich Herold (visitor, Paul Scherrer Institute, Switzerland) then conducted a small coordinated field project called Sailplane Studies of Early Electrification (S2E2) that involved the sailplane and Colorado State University's CHILL (University of CHicago/ILLinois State Water Survey) radar. The goals of S2E2, which occurred in northeast Colorado from late June through early August, were to test and exercise the capabilities of the new data system and to collect data relating the microphysical and electrical characteristics in developing High Plains thunderstorms. The specific scientific objectives were to relate in situ electrical measurements with reflectivity structure and evolution, to compare in situ microphysical measurements with theoretical determinations of hydrometeor types from polarimetric radar data, and to assess the potential for polarimetric radar measurements to detect the cloud glaciation process and its relation to initial cloud electrification.
Sailplane data were telemetered to a ground station located in the operations trailer at the CSU-CHILL facility. Using the ATD/RAF Windows Display System (WINDS) software, the data were displayed for real-time monitoring and decision-making, and Global Positioning System (GPS) position was passed to the radar displays, allowing for close coordination of scanning strategies. Flights were made on ten days, seven of which resulted in potentially useful microphysical/radar studies. The convective conditions encountered during S2E2 were generally weak, limiting most of the in situ measurements to the lower parts of clouds. Extended sailplane climbs (6.3 to 7.5 km MSL) in moderately strong updrafts occurred on three days, each with contrasting electrical conditions. Initial analysis will concentrate on comparing the reflectivity development with the electric field evolution for these cases. Analyses of particle charge measurements (at small scales) and polarimetric parameters (at larger scales) should further characterize the processes of precipitation formation and electrification.
James Dye and Brian Ridley (ACD) are nearing completion of the analysis of the data from the Electrified Clouds as Sources of NOx and O3 (ELCHEM) project conducted in New Mexico in 1989, an exploratory study to investigate the production of NOx by lightning. The NCAR Sabreliner was used to make measurements of NO, NOy, NO2 and O3 in the vicinity of and in penetrations of the anvils of small thunderstorms forming in the vicinity of the Langmuir Laboratory of the New Mexico Institute of Mining and Technology (NMIMT). Dye, working with William Rison (NMIMT), showed that for the two days with electric field change measurements the ratio of intra-cloud to cloud-to-ground lightning was roughly 1.5:1. This ratio is less than the frequently quoted ratio of 4 or 5 to 1 for midlatitude continental thunderstorms.
The results of the study show that for two classical, isolated airmass thunderstorms, about 200 to 250 kg(N) was deposited in the upper altitude above 8 km. Because of uncertainties in global lightning flash rates, distribution with latitude and region, and the amount of NOx produced by different types of lightning strokes, extrapolation to the global scale is little more than an arithmetic exercise. But if this is done and the frequently quoted global flash rate of 100 per second is assumed, the data yield an estimate of about 2 to 5 Tg(N)/yr only in the upper parts of the storm.
Andrew Heymsfield, Gregory McFarquhar, Larry Miloshevich, Janine Goldstein, and Steven Aulenbach continued analyses of the data collected in the First International Satellite Cloud Climatology Project (ISCCP) Regional Experiment (FIRE)-II and Central Equatorial Pacific Experiment (CEPEX). These studies emphasized the documentation of ice crystal size distributions in cirrus clouds and the relationship of those distributions to the processes responsible for ice formation in cirrus.
In FIRE-II, research focused on two main areas: continued analysis of the in situ data collected in cirrus, and acquisition of new cirrus data using balloon-borne ice crystal replicators. A three-layer conceptual model describing the structure of cirrus clouds in the vertical was developed using the observations of relative humidity and the size distributions from in situ aircraft data and the replicator and concurrent cryogenic hygrometer data collected in cirrostratus. The uppermost layer is a thin zone of ice crystal nucleation, the middle layer is a deep zone of ice crystal growth, and the lowest layer is a zone of ice crystal sublimation.
Balloon-borne ice crystal replicators were used to determine the ice crystal size distributions during ascents through two cirrocumulus/wave clouds and two deep thunderstorm anvils in Colorado, and three cirrostratus clouds in Arizona. A newer and lighter version of the replicator, better suited also to future aircraft operations, performed satisfactorily during a launch into a cumulus cloud. Other replicator-related activities include improvements in our data processing and image analysis techniques and studies of the collection efficiency.
In CEPEX, the analyses have emphasized the data from the Video Ice Particle Sampler (VIPS) and the PMS two-dimensional spectrometer probes. The VIPS data can be used to determine the size distribution of ice crystals with sizes smaller than 100 microns, where measurements from the two-dimensional spectrometer are not reliable. The VIPS data processing technique has been improved dramatically over the past year, and the throughput has been increased by an order of magnitude. With merged VIPS and two-dimensional data, it has been possible to document the important influence of crystals smaller than 100 microns on the radiative properties of cirrus clouds. In these analyses, the CEPEX data have been augmented with largely unanalyzed data collected in tropical cirrus anvils in the vicinity of Kwajalein, Marshall Islands, in the mid-1970s. Two articles describing the results of the analysis, submitted to the Journal of Atmospheric Sciences, indicate that the observed high albedos of tropical cirrus clouds result primarily from the larger ice crystals lower in the thick cloud, rather than smaller crystals near cloud tops.
In preparation for future continuation of the studies of cirrus properties, preliminary plans have been assembled for an experiment to study tropical cirrus clouds. This project, being developed in collaboration with investigators from NOAA and several universities, will take advantage of the new NCAR WB-57F high-altitude research aircraft, which should make it possible to penetrate the upper regions of tropical cirrus clouds. The experiment is a component of the NCAR-wide Clouds in Climate Program.
Cooper and Lawrence Radke (ATD) began construction of a new instrument for the detection of the soluble particles thought responsible for the formation of ice in cirrus clouds. This instrument will detect the freezing of solution droplets and the associated formation of ice, and will also provide measurements of ice nuclei and/or cloud condensation nuclei. The instrument is scheduled for operation during the next year.
Knight has completed a critical study of a group of theoretical treatments of ice surfaces that describe very water-like layers on the surfaces at equilibrium, at temperatures below freezing. These ideas have been applied to geophysical phenomena including charge separation in storms and snow crystal growth. He concludes that these treatments contain a fundamental error, and argues that ice surfaces probably do not have thick, disordered layers.
Cooper and Roy Rasmussen (joint appointment with RAP) continued their analyses of the extensive set of observations collected during the 1994 Winter Icing and Storms Project (WISP94). This experiment concentrated on the formation of ice in wintertime clouds, especially upslope and wave clouds because of the relative simplicity of the airflow patterns in those clouds. In collaboration with Paul DeMott and David Rogers (Colorado State University, CSU), they compared observed ice concentrations to those deduced from measurements of aerosols and ice nuclei. The ice concentrations observed in clouds were similar to those that developed in one of the CSU ice nucleus detectors, which suggests that such measurements may be a reliable indicator of ice nucleating potential in these cases.
During October and early November of 1994 Dye, Baumgardner, Bruce Gandrud (RAF), and Keith Barr used the Multiangle Aerosol Spectrometer Probe (MASP) to make aerosol measurements during the fourth and last deployment of the NASA ER-2 in Airborne Southern Hemisphere Ozone Experiment/Measurements for Assessing the Effects of Stratospheric Aircraft (ASHOE/MAESA). ASHOE/MAESA used the NASA ER-2 flying south from New Zealand to examine the causes of ozone loss in the southern hemisphere and to investigate polar, midlatitude, and tropical processes responsible for the ozone loss and to study how stratospheric aircraft might affect these processes. This last deployment found very large losses of ozone in the southern polar vortex and also regions with reduced aerosol loading in regions even somewhat outside of the vortex. Measurements were made on the return ferry flights from Christchurch to Tahiti, Tahiti to Hawaii, several flights south from Hawaii to the equator, the return from Hawaii to Moffett Field, California, and one flight north from Moffett Field to about 60 N. Thus, a latitude survey of the lower stratosphere was made from about 68 S to about 60 N in a time period of about two weeks. This survey showed that the effects of the eruption of Mt. Pinatubo in June 1991 on the particles in stratospheric sulfate layer had greatly diminished. The measurements also showed that there was not a substantial difference in stratospheric aerosol loading between the southern and northern hemispheres.
Continued analysis by Dye of the measurements obtained in a polar stratospheric cloud (PSC) on the edge of the southern polar vortex show that this cloud was primarily formed by the uptake at cold temperatures of HNO3 and H2O onto the supercooled stratospheric H2SO4/H2O solution droplets. This result is similar to what had been found by Dye and coworkers in some Arctic polar stratospheric clouds, and contrary to some earlier measurements in the Antarctic. The similarity in formation temperature and behavior as temperature decreases shows that the formation process of some PSCs can be the same in both regions. In this Antarctic case, however, comparison of the results with a model of equilibrium ternary droplet growth suggests that the formation of some solid phase may be starting at the time of the measurements. The modeling is being done by Katja Drdla (ASP graduate fellow) who is working on her dissertation jointly between NCAR and UCLA.
Bruintjes, William Hall, and Terry Clark have been actively involved in the Arizona Program over the past year. Two main objectives of this program are to improve our understanding of the processes that determine the spatial and temporal distribution of precipitation and to assess the potential for artificially enhancing winter precipitation over the Mogollon Rim.
A multi-institutional field program was conducted from 15 January to 15 March 1995 in Arizona, making detailed measurements of the precipitation evolution in both natural and seeded winter storms as they pass through the area. The data collected during this field program will be used to gain additional physical insights into precipitation evolution in winter storms and to further validate modeling results. In particular, the physical parameterizations of the model and the treatments of both the boundary layer and microphysical processes will be assessed. This work includes seeding simulations and comparisons with observations. Once the model performance is proven acceptable, simulations will be performed to determine the precipitation enhancement potential via cloud seeding for a typical winter season.