Estimates are made of the contribution of dust devils to the aerosol mass burden over the U.S. These estimates have been derived from experimental data generalized by using climatic and vegetation maps of the U.S. The area of maximum calculated dust production by dust devils in the southwestern U.S. Comparison of our calculated fluxes with background aerosol data shows general agreement. Comparison of alkaline dust emissions from wind erosion shows that dust devils provide a similar magnitude input of atmospheric dust.

A new ice nucleant aerosol was produced by combustion of a 2% AgI-0.5 mole % Bil3-NH4I-acetone-water solution. The ice nucleating effectiveness of this aerosol is an order of magnitude greater than AgI alone at −10°C. An X-ray powder analysis identified the aerosol as the hexagonal crystal form of AgI having the closest match to ice ever reported for a nucleant of this type.

Several of the mesoscale dynamic and thermodynamic aspects of convective scale interaction are examined. An explanation of how sounding data can be coupled with satellite observed cumulus development in the warm sector and the arc cloud line's time evolution to develop a short range forecast of expected convective intensity along an arc cloud line. The formative, mature and dissipating stages of the arc cloud line life cycle are discussed. Specific properties of convective scale interaction are presented and the relationship between arc cloud lines and tornado producing thunderstorms is considered.

Regions of brightness variations are common in visible and near-infrared satellites images from clear coastal regions. The variations have been hypothesized to be caused by aerosol particles in the marine boundary layer. The hypothesis was tested using in situ particle measurements collected near the time of satellite overpasses. Boundary-layer particle concentrations related to the brightness variations: high concentrations existed in bright regions and vice versa. This result indicates that, in regions over the ocean free of clouds, sunglint and whitecaps, the visible and near-infrared sensors aboard certain orbiting meteorological satellites can detect variations in the concentrations of haze particles in the marine boundary layer.

Severe storm aircraft measurements are documented, as well as the instrumentation and operational features of aircraft mobility capabilities. The measurements and data analyses indicate that the concept of a highly mobile research aircraft capability for obtaining detailed measurements of wind, temperature, moisture, spherics, etc., near and within severe storm systems, forecast 48 hours in advance in a 1000 nm operating radius, is feasible, and was successfully demonstrated. The measurements and analyses reveal several severe storm features and insights with respect to storm air flow circulations and inflow-outflow orientation. Precipitation downdraft air is recirculated back into the updraft core below the scud cloud in both back and front feeder type storms. In a back feeder type storm, the downdraft outflow air ahead of the storm is also recirculated back into the updraft region near cloud base.

Using an airborne lidar, we have measured atmospheric aerosol backscatter coefficients (differential backscatter cross section per unit volume) for 10.6 μm wavelength laser radiation as a function of height to 5200 m for a number of meteorological conditions over the United States high plains. Airborne in situ samplers measured the particle size distribution at the same time and altitude as the lidar measured backscatter. One backscatter coefficient profile at 10.6 μm was compared with a 0.694 μm lidar backscatter profile as well as with the particle size distribution profile. The average infrared backscatter coefficient ranged from ∼8 × 10−9 m−1 sr−1 at the surface to 1 × 10−10 sr−1 at 5200 m altitude.

A Doppler lidar measures the line-of-sight velocity of cloud droplets in a waterspout much as a meteorological Doppler radar measures the velocity of larger hydrometeors. We discuss details of the application of an airborne Doppler lidar to waterspout velocity measurements, including intensity weighting and limitations of the technique. One type of result available from the lidar data is the velocity spectrum of the line-of-sight velocity component of scatterers in the flow, integrated along the lidar axis, as a function of distance from the vortex axis. From the velocity spectra, peak winds in the portion of the waterspout marked by cloud droplets, turbulence levels, and interaction with the ambient flow can be inferred. In one example the maximum velocity observed in the visible part of the waterspout is 10 m s−1. This double-walled waterspout showed a two-peaked velocity spectrum, which we interpret as a dynamic difference between the two coaxial components of the vortex.

During September 1974 in the Lower Florida Keys, the first successful penetrations of mature waterspouts were accomplished by a specially instrumented research aircraft. Throughout the course of each penetration, the measurement system recorded the temperature, the pressure and the three-dimensional velocity field near and within the visible funnel. Multiple penetrations of both cyclonic and anticyclonic waterspouts in various life-cycle stages were achieved. The results indicate that the waterspout funnel structure exhibits 1) a warm central core region, 2) positive vertical velocities of 5–10 m s−1 outside of the warm core, and 3) tangential velocities and horizontal pressure gradients with characteristics similar to but with magnitudes greater than those of the dust devil. A scale analysis of each term in the governing equations of motion suggests a simplified set of modeling equations. The simple Rankine-combined vortex model with cyclostrophic flow explains approximately 75% of the total measured pressure deficit. This compares favorably with Sinclair's (1966, 1973) earlier results for the dust devil vortex.

Measurements from an extensive field program in the desert near Tucson, Ariz., are used to formulate a quantitative dust devil model. The model is based on measurements of temperature, pressure and wind velocity taken from a mobile instrumented tower near and within dust devils at levels of 6, 17 and 30 ft. Three dust devil penetrations are analyzed with respect to the temperature, pressure and wind velocity profiles contained in a plane defined by the dust devil axis (or local vertical) and the direction of motion. These profiles corroborate earlier preliminary measurements by Sinclair and thus firmly establish the existence of the warm, low-pressure core of the dust devil. Maximum temperature and pressure variations within the dust devil varied from 4 to 8C and 2.5 to 4.5 mb. The most significant result of the measurement program was the discovery of a downcurrent or a region of markedly reduced vertical motion along and surrounding the sloping dust devil axis. Immediately outside the downcurrent, the vertical velocity reaches positive peak values and then diminishes rapidly with respect to increasing radius. Superimposed on this motion is a strong rotary and radial motion which results in a combined flow pattern similar to that of a helical vortex. The measurements are synthesized into a dust devil model. The observational model represents a two-cell structure which also appears to exist in other atmospheric vortices, such as the tornado and hurricane, and some laboratory and theoretical vortex models.