Observations of the surface of Mars suggest a high probability of surface water activity in that planet's past. Consequently, many studies of Mars' early atmosphere have attempted to estimate the carbon dioxide level by requiring that surface temperatures be high enough to support surface liquid water. In the main, these studies have employed one-dimensional, radiative-convective climate models capable of considering only a single solar zenith angle, typically chosen to represent a global and annual average. Such models are hence not well suited for considering meridional variations in the temperature profile, which are affected by variations in the orbital obliquity and the meridional redistribution of heat by dynamic processes. I describe modifications to a more complex model, the multi-level energy balance model designed at NASA's Goddard Laboratory for Atmospheric Sciences, which make it suitable for study of an atmosphere with varying carbon dioxide levels. Vertically and meridionally defined, the model includes heating and cooling by radiation, mean meridional circulation, large-scale (baroclinic) and small-scale (convective) eddies, and surface turbulent flux. I present annually-averaged results for an examination of potential atmospheres of early Mars, given that its carbon dioxide level may range from 0 to 500 Pa and the orbital obliquity from 0$\sp\circ$ to 50$\sp\circ.$ These results are compared with those obtained from a radiative-convective model.

Globally representative atmospheric electric field and current data have been obtained from two sites at Amudsen-Scott South Pole Station at ground level (3 m) with 1 Hz time resolution. The average diurnal variation has been calculated for individual months and seasons, and well known features of the Carnegie curve observed. Amplitude ratios for seasonal curves range between 30-43%, in agreement with other studies. The Northern hemisphere winter is found to be a minimum for global convective-electrical activity and summer a maximum, in contradiction to the original 1929 Carnegie results. Seasonal phase shifts observed in previous studies of the diurnal variation of deep convective activity (DCA) in the tropics also appear in the results of this study. Good correlations between low amplitude peaks suggest our data are highly accurate and could be useful in detecting changes in weather patterns as might occur with global warming.

We develop a stochastic spatial-temporal model of the wind field in Houston, Texas, using wind data collected at pollution monitoring stations scattered about the area. This model is intended to be used as an input to an air pollution model, since wind is a very important part of modeling pollution transport. Since wind is naturally bivariate, in our case east and north wind components, we develop a method that assumes the wind is the gradient of some smooth, latent, univariate potential field. The mean and covariance functions of this potential field then induce the mean and covariance functions for the wind field. The advantage to this approach is that developing a scalar-valued covariance model for the potential field is much easier than developing a general, matrix-valued covariance model for the wind directly. We develop the mathematics for this gradient process model, and in particular explore the structure that arises when the underlying potential has an isotropic covariance. This structure allows us to develop variogram-like functions that can be estimated from the data in a way very similar to standard variogram estimation. For the data analysis, we propose a model of the wind following Breckling's 1989 time-series study: the wind is decomposed into a geostrophic term, a diurnal term, and an error term. The geostrophic term is constant over space, since the spatial scale we are working on is small compared to pressure systems, and captures the long-range time dependence caused by these pressure systems. The diurnal term is allowed to vary over space and time, is periodic with a period of 24 hours, and captures the sea-breeze oscillations. We adjust for errors in the data that are caused by local features, such as buildings and trees, that prevent the "true" wind from being measured. We explore the nature of the sea-breeze, and find that rotating the components of the wind allows the sea-breeze to be captured in a single component that is perpendicular to the sea coast. We fit the elements of the model using the variogram and maximum-likelihood estimation.

A model describing a thunderstorm's interaction with the global electric circuit is presented. The model includes a thunderstorm, the surrounding atmospheric and ionospheric region, and the magnetically conjugate atmospheric and ionospheric region. The model is time-dependent, and includes lightning and thundercloud evolution. A method of using experimental data to more accurately simulate observed thunderstorms has been developed. Of the upward current generated by a thunderstorm, about 50% flows through the Earth's magnetosphere to the conjugate hemisphere. This percentage is fairly constant over the storm's active life, and varies little with storm size or structure. Infrequent lightning activity (less than one flash/minute) within a thunderstorm does not appear to greatly affect the thunderstorm's efficiency in transferring separated charge to the global circuit. Lightning does limit the magnitude of the electric fields and resulting currents within and below the storm.

The existence of infrasonic pulses induced by lightning within thunderclouds is now firmly established from experimental observation. The observed waveform has a compressive pulse followed by a rarefactive pulse. This is of opposite polarity to that predicted by pre-existing theories. This thesis presents a version of a previous theory modified to explain the difference in polarity. It is shown that these infrasonic pulses are wholly electrostatic in origin and arise out of the interaction of the large space charges inside thunderclouds. The positive part of the pulse is due to the relaxation of the mutual attraction of the negative charge layer and the positive shielding layer beneath it during a lightning event. The negative portion of the pulse is due to the reduction in the mutual repulsion of the charged cloud particles in the negative charge storage layer. A series of calculations, done with spherical charges, is presented that describes the changes in the electrostatic potential energy in a simple thundercloud as the result of the partial neutralization of the stored charge by a lightning flash. These calculations are used to demonstrate that two competing theories, ohmic heating of the air within the charge storage region and rapid charge separation, cannot explain the amplitude of the initial positive excursion in the infrasonic waveform.

A series of 48-hour experiments performed quarterly in 1977 to study radar returns from the lower stratosphere is described. Current theories of stratospheric scatter of radio waves are presented and the radar equation pertinent to the large array at Jicamarca, Peru is derived. Observational and data reduction techniques are discussed. Experimental results concerning scattered power, mean wind velocities and velocity fluctuations are presented. It is found that isotropic turbulent scatter is not the only scattering mechanism operating in the stratosphere, and modifications to the radar equation which arise from turbulent rolls or thin layers of refractive index variation are needed. Mean winds measured by radar are comparable to those determined by rawinsonde ascent. Velocity oscillations of approximately 5-minute periods are interpreted as buoyancy oscillations with 2 km horizontal wavelength and phase velocity of 6 m/s. Tidal oscillations with periods of 24 and 12 hours are also observed. Suggestions for future measurements at Jicamarca and other institutions are offered.

A Terrier-Malemute sounding rocket was launched into a stable, premidnight auroral arc. Among the instruments carried by the payload were a cesium vapor magnetometer, electrostatic curved-plate electron energy analyzers, and two sets of Langmuir double-probes. Simultaneous line-of-sight velocity and electron density measurements were made during the flight by Chatanika radar operating in an elevation scan mode in the magnetic meridian. Data from the magnetometer indicated that a broad region of eastward current flowed within the diffuse aurora equatorward of the arc, while westward current flowed within the arc itself. The field aligned current was downward in the diffuse aurora and uniformly upward in a 100 km wide region which contained the arc. This current pattern suggests that the payload traversed the entire Region 1 field aligned current system. The narrowness of the current was probably due to the expansion of the auroral oval associated with the growth phase of a substorm which occurred one hour after the flight. The double probe measured a 40 mV/m northward electric field just equatorward of the arc. Within the arc the northward electric field was zero. The tangential component of the electric field was 7 mV/m and was constant across the arc boundaries. Since the electric field in the arc was not strong enough to account for the westward current measured by the magnetometer, the presence of a westward neutral wind was inferred. This westward wind is consistent with the difference between the meridional electric field measured by the double probe and the radar. Differences in the east-west electric field measured by the two instruments could be accounted for by an upward drift of ions at 100 m/s within the arc. This ion current constituted half of the measured upward field aligned current within the arc, while the other half was carried by precipitating electrons. The simultaneous measurements enabled the association of the aurora with large scale regions in the magnetosphere. They also allowed the construction of a self consistent perpendicular current system for the arc involving a time varying neutral wind. There was some discrepancy in the field aligned currents derived from the data. This was attributed to the arc's association with the Harang Discontinuity.

A Balloon Electric Field Sensor, released into a thunderstorm at Langmuir Laboratory on 12 August 1976, measured the electric vector and wind profile along its track. During a major portion of its ascent, the instrument was in an intense updraft. We infer a maximum vertical wind component of about 15m/s. Horizontal flow, at 10-15m/s towards the northeast, was encountered above 7000m (msl). The behavior of the electric vector indicated that the balloon rose through an 800m thick region of positive charge in the lower portion of the cloud. Above this, negative charge extended from 5200m to 8200m (msl). We imputed intense horizontal electric field components to a volume of negative charge, situated near 6700m (msl) in a region where the rainfall rate was only about 3mm/hr. Lightning subsequently discharged this volume. The cloud contained positive charge above 8200m; however, an increase in the electric field's magnitude after a lightning flash probably resulted from a nearby negative distribution, close to the radar echo top. Electrical corona, induced in pointed rods affixed to a radiosonde, was measured as the sonde rose through a thundercloud on 21 July 1978. The storm's most vigorous activity was well to the south of Langmuir Lab. Although our instrument was carried 6000m in this direction, it did not reach the region of heaviest precipitation. The corona record indicates that strong fields and charge of both polarities were encountered during the ascent. An abrupt decline in upwards directed corona current as the instrument passed through the cloud's upper surface may have resulted from a negative screening layer.

The rotational motion of atmospheric ice particles is analyzed to determine the conditions under which the particles align with atmospheric electric fields. The particle shapes are approximated by prolate and oblate spheroids and may be treated as conductors or dielectrics, depending on the speed of the rotation in relation to the electric response times. At low Reynolds numbers, the electrical torque is opposed by a Stokes drag torque. The resulting equation of motion is that of a damped pendulum oscillator. At higher Reynolds numbers, the electrical torque competes with an aerodynamic torque which itself attempts to produce an alignment of the ice crystals not necessarily compatible with that for the electrical torque. This aerodynamic torque is investigated in some detail in Chapter 1. In Chapter 2, the dynamics of a spectrum of ice particle shapes and sizes in a wide range of electric fields is examined. Numerical values are obtained for the relevant characteristic times and torque ratios, and the different dynamical regimes are delineated. It is found that electric fields of at least 1 kV/m, or perhaps 10 kV/m, are required to produce significant alignment. The significance of these results and the possibilities for further research are briefly discussed.

The electron temperature changes due to 3, 5 and 430 MHz radio wave heating in the lower ionosphere are measured using incoherent scatter diagnostic techniques and are also calculated from heating/cooling theory. The experiments were performed at Arecibo Observatory using the new HF heating facility and the Arecibo Observatory 430 MHz incoherent backscatter system. In order to interpret the incoherent scatter results a spectral parameter library is developed which gives the spectral width, the spectral maximum and the bandlimited power of the incoherent scatter spectrum for a wide range of ionospheric parameters. There are two collisional formulations which have been widely used to reduce incoherent scatter data in the D and E regions: Dougherty and Farley (1963) and Waldteufel (1970). To determine which collisional formulation should be used, we examine the results of recent Arecibo experiments performed in an unheated ionosphere. A comparison of the measured electron-neutral collision frequency values derived from the two different collisional formulations to the predicted model values show excellent agreement for the Dougherty and Farley formulation but less than satisfactory agreement for the Waldteufel formulation. Using the Dougherty and Farley formulation we determine electron temperature changes from the measured heated-to-ambient spectral parameter ratios. In comparing the measured electron temperature changes to the predicted changes for 430 MHz heating we find a large discrepancy throughout the D and E regions: the measured electron temperature changes are much less than the predicted. The discrepancy in the 75-100 km region can be removed by increasing the model O(,2) rotational cooling rate by a factor of 10, while the discrepancy below 75 km can be removed by a factor of 4 increase. The cooling rate increases, however, are not the only possible explanation for the discrepancies. Two other effects, the non-Maxwellian electron velocity distribution and heat conduction, could remove the discrepancies if the magnitude of their effects were significantly increased in the model. The discrepancies could also be removed by using a f('2.18) frequency scaling law for the predicted heating rather than the currently accepted f('2) law, but there are no physical explanations to support this modification. The 3 and 5 MHz heating results are in satisfactory agreement with the model if D region absorption is taken into account and, thus, do not support the increased cooling rates suggested by the 430 MHz results. The agreement of these results, however, would not be significantly affected by the other suggested modifications. The heating due to the 52 (mu)sec diagnostic pulse is also measured. The diagnostic pulse heating at 70 km is found to increase the electron temperature by a factor of 2.85 (+OR-) 1.35 above ambient. Although the error estimates are large, this increase is in agreement with the predictions of the model.