This thesis presents theoretical and numerical studies of the radial diffusion of relativistic radiation belt electrons. The research has been focused particularly on the radiation belt phase space density profile, and radial diffusion due to particle drift resonance with ULF waves. Observations have shown a strong connection between magnetospheric ULF oscillations and electron flux enhancements. I investigate radial diffusion coefficients based on theoretical analysis of particle diffusion in ULF perturbation electric and magnetic fields. The analytical diffusion coefficients consist of two terms: a symmetric term and an asymmetric term. The symmetric term agrees with earlier works, and the asymmetric terms are new. Both terms show good agreement with numerical test particle simulations. The asymmetric terms have higher L dependence, which indicates they might be more important at higher L-shells or at times when the magnetospheric field is highly asymmetric. A numerical radial diffusion model has been developed which can take into account: dynamic boundary locations and values, plus effects of losses and sources. Several test cases are considered to study the effects of different diffusion coefficients, internal sources, external sources, and loss. A method of converting observational particle flux to phase space density is also presented. Identifying the source and loss processes using observational data is currently one of the key issues for understanding and modeling radiation belt dynamics. We present a new measurement technique which utilizes two GOES satellites located at different local times to calculate the radial gradient of phase space density at geostationary locations. The result shows positive gradient at geomagnetic quiet periods. To further study the high energy electron transport, especially the ULF related acceleration during storm times, I use the numerical radial diffusion model for the September 24-26, 1998 storm and compare the results with an MHD test particle simulation. The diffusion result using ULF-wave diffusion coefficients and a time-dependent outer-boundary condition can reproduce the main features of the MHD-particle results quite well. Using wave driven diffusion coefficients gives better results than using power law or Kp-dependent diffusion coefficients.

Observations of solar wind temperatures near the Earth indicate that heating of the solar wind plasma exists. An alternate approach to finding explicit heating terms for the energy equation is to use a polytropic approximation. Using data from the Helios 1 spacecraft, an empirical value for the polytropic index of the solar wind is found to be independent of speed state, within statistical error, and has an average value of 1.47. Application of this empirically derived index to a solar wind computer model is examined by comparing the MHD energy equation and the polytropic relation. The result is obtained that the polytropic index can replace the adiabatic index in the MHD energy equation to simulate the effects of heat conduction if the assumptions are made that the heat conduction flux has a specific form and the particle pressure has no explicit time dependence. Justifications and limitations of this approach are discussed.

Can the average bulk flow of an ensemble of charged particles in Earth's plasma sheet still be described by adiabatic theory even if the ensemble contains a significant number of particles executing non-adiabatic motion? This is part of a broader spectrum of questions which ask if chaotic microscopic processes can be parametrized as macroscopic ones when ensemble averaged. Wolf and Pontius (1993) have shown that at least for a simple 2D, tail-like magnetic field configuration, the average particle drift speed of an appropriately chosen ensemble of particles, including those executing chaotic motion, is given correctly by the simple adiabatic guiding-center drift formula. Here, we extend the proof to 2${1\over2}$D magnetic fields (3 component, 2 spatial dependences) and include the effects of an electric field. The results of numerical test-particle simulations further show that the dispersion of particles about the mean drift speed tends to decrease due to the presence of chaotic particle scattering. Thus, we have shown that the standard way of representing particle transport in the inner magnetosphere, namely the isotropic pitch angle, bounce averaged drift formalism, is valid for the central plasma sheet despite the presence of non-adiabatic particle motion.

Plasmas from a pulsed jet discharge have been characterized with respect to gas species and nozzle design. Spectral lines from the gas used in the pulsed jet are apparent in the visible region. The vacuum ultraviolet spectrum, in particular for gases with high Z, is dominated by emission from species sputtered from the nozzle. Although an extensive search was made for excimer emission, no evidence of such emission was found. The production of highly ionized and excited states from materials created by the sputtering of the nozzle has possible applications in VUV spectroscopy. By operating the pulsed jet discharge at a 50 Hz repetition rate with NF$\sb3$ to produce excited fluorine ions, etch rates in excess of 10 $\mu$m/min. have been achieved in silicon, which may have application to micromechanics.

Kubo's generalized cumulant expansion theorem is used to derive a theoretical expression for the nuclear magnetic resonance (NMR) signal received from a fluid moving in a time-dependent magnetic field gradient. Described in general terms by time-dependent correlation functions, this expression is used to investigate a new statistical model of microcirculation that incorporates both coherent and incoherent flow effects at the microscopic level. Based on a simple picture of the intravoxel environment, this model is developed by considering an arbitrary distribution of tortuous capillary flows. A statistical analysis of the Langevin equation describing slow tortuous capillary flow as a stochastic process reveals precisely how both coherent and incoherent flow effects contribute to the overall attenuation of the NMR spin-echo. Velocity compensated and non-compensated diffusion matched spin-echo imaging sequences are utilized to separate and quantify these respective effects noninvasively on phantoms of stationary and flowing fluid.

This paper continues the work of Ansher et al. (1992) by identifying and examining density structures in Jupiter's magnetosphere. The 110 hours of data used are from a 4-second temporal resolution density data set derived from plasma wave instruments on board both Voyager 1 and Voyager 2. One hundred five structures are identified. They are believed to be the same type of structures as seen by Ansher et al. (1992) and are found to have sharp density gradients at the boundaries, average scale sizes of about one Jovian radius, and typical density variations between 50% and 200% of the background. Many structures show good correlation with the magnetic field data. In addition, the existence of the density structures has little if any dependence on radial distance from Jupiter, System III longitude, or magnetic latitude. Comparison with four plasma transport models indicates that the observed structures resemble flux tubes of varying plasma content compared to the background density. These findings are in agreement with those of Ansher et al. (1992).

The interaction between the solar wind and a rotating planet causes the field lines in the planetary magnetotail to twist into a helix. Using a simplified magnetotail model, we examine hydromagnetic waves propagating down the magnetopause for such a field configuration and derive the dispersion relation of the waves. It turns out that only under certain special circumstances can the hydromagnetic waves be stable. In a thin magnetopause boundary layer, the helical wave is found to be always stable and its wave frequency depends weakly on the plasma and the field within the layer. The current system of the boundary layer is found to be modulated by the wave and the modulation is proportional to the velocity perturbation of the plasma. The wave influence on the spiral angle is examined briefly for some special cases for which we find the variation of the angle increases monotonically with increasing radial distance.

A model is constructed to describe the magnetic field, the global current system, the electric field and the potential in the solar wind interaction with Mars assuming that Mars has no intrinsic magnetic field. It, therefore, incorporates the physics learned from the Pioneer Venus Orbiter, which observed the interaction of Venus, an unmagnetized planet, with the solar wind for 14 years. Integrating recent knowledge of the global current system at Venus (Law 1995) into the Flow/Field model of Cloutier et al (1987) and expanding the model to represent three dimensions, we adapt the Flow/Field model for application to Mars. We investigate the 3-D current system to learn the physics of the interaction. Then, the model is applied to test simple geometries in order to validate it. Future applications are discussed.

The Rice Convection Model has been modified and applied to the study of the Jovian magnetospheric system, which is interchange unstable. The basic interchange instability of the Io plasma torus is opposed by pressure gradients in the energetic particles outside the torus. Many simulations have been performed for cases where the overall system is inter-change unstable under the ideal-MHD assumption E + v $\times$ B = 0. For such cases, the torus breaks up predominantly into long fingers unless the initial condition strongly favors some other mode. The ends of the fingers tend to be rounded, and they are connected to the main torus by tails that thin rapidly with time if the torus runs out of plasma. Our calculations place an upper limit of $\sim$1R$\sb{\rm J}$ on the average distance between fingers. For an initially asymmetric large-scale torus, fingers generally form on a time scale shorter than the one on which the heavy side of the torus falls outwards. However, the fingers form predominantly on the heavy side. Galileo may observe such finger features outside the Io torus, at L $\approx$ 7 to 15. Additionally, in this thesis, drift-wave theory has been used to investigate the effect of energetic (KeV or MeV) particles on the Io torus plasma transport. It is shown that the MHD stability criterion, where the interchange motion would be completely stabilized if the energy density of the hot stabilizing plasma is greater than or dual to 3/4 of that of the cold unstable plasma, no longer holds owing to the gradient/curvature drift of the energetic particles. This differential-drift effect, which is a departure from the ideal-MHD and frozen-in flux, may play a significant role in plasma transport in the Jovian magnetosphere.

We have undertaken a study of the low and mid latitude ionospheric electric field pattern, during both magnetospherically quiet and active periods. Our analysis can be conveniently split into two parts. i.In an effort to study the penetration of magnetospheric electric fields to low latitudes, we have compared Jicamarca F-region vertical drifts for 10 radar-observation periods with the auroral boundary index (ABI). The ABI is the latitude of the equatorward edge of the diffuse aurora at local midnight, as estimated from precipitating-electron fluxes measured from DMSP spacecraft. The periods occurred in the interval January 1984 to June 1991 inclusive and each lasted between 2 and 5 days. We focus on periods that occurred in September 1986, March 1990, and June 1991. In the post-midnight sector, where we expect the penetration to be strongest, we found many examples of correlation; specifically, associated with an ionospheric updraft (implying an eastward electric field) is a strong poleward motion of the auroral boundary. However, we also found a significant number of cases where there was little or no correlation. We conclude that there is only mediocre agreement between the observed Sudden Postmidnight Ionospheric Events (SPIEs) and the ABI. These SPIEs have also been compared with other magnetospheric parameters, namely $D\sb{\rm st}$ IMF $B\sb{z}$ and the polar cap potential. $D\sb{\rm st}$ showed significantly better correlation with the SPIEs. We summarize the proposed models for SPIEs and compare their predictions with the data, concluding that no single model can account for all events. While it is clear that some of these SPIEs can be explained in terms of direct penetration of magnetospheric electric fields, we suggest that the remainder may be due to magnetospherically-generated neutral wind effects. ii. We have constructed a model of the low- and mid-latitude potential distribution, applicable for both quiet and active times. We use the Mass-Spectrometer-Incoherent-Scatter (MSIS) model to input the number densities and temperature of the neutral species, and the International reference Ionosphere (IRI) model to input the electron/ion densities and temperatures. As our wind input we use the Horizontal Wind Model (HWM). We find that our model can reproduce the all of the main features of the low latitude ionosphere during quiet times, and supports some of our ideas about magnetospheric penetration during active periods. We use the model to probe the dependency of the low latitude penetration pattern on solar conditions and season and found that the inferred equatorial drifts are relatively insensitive to either. Thus we conclude that ionospheric pre-conditioning is unlikely to play a significant role. On the other hand, the low latitude penetration pattern is strongly dependent on the assumed poleward boundary.