This work presents research on the one-microsecond, L-band relativistic klystron amplifier (RKA) project conducted at Los Alamos National Laboratory. A collection of theoretical analyses is presented for rf cavities, intense electron beams, beam-cavity interactions, and small-signal klystron bunching. Electromagnetic field simulations were done for three dimensional cavity structures using HFSS with very accurate results. Particle-in-cell simulations of the complete RKA were done using the two dimensional code ISIS. Extraction efficiency for intense modulated beams is discussed and verified in simulations. Designs for input and idler cavities are reviewed. Extremely low-Q, single-gap, output cavities are investigated for coupling rf power from very low-impedance, modulated, electron beams. Output cavities with a Q less than 4 have been designed, measured, and tested. Methods were implemented for designing 2D equivalent output cavity structures to model 3D structures in 2D codes. A technique for ex-situ rf conditioning of the output cavity gap pieces is presented. A beam-pipe center conductor, intended to reduce the space-charge potential depression of the beam, is discussed. Diagnostics for intense-beam and high-power rf measurements are presented. A coaxial directional coupler and load, capable of handling more than 500 MW at 1300 MHz, were designed. Mode conversion from coax to waveguide is discussed for $>$100 MW power levels. Methods for determining the gap voltage in an operating cavity are presented. Pulse-shortening of the rf in the RKA is also discussed. A 650 kV, 5 kA, one-microsecond, annular beam has been produced from a stainless-steel, explosive-field-emission cathode. The beam current was modulated up to 70% $(I\sb1 /I\sb0$ = 70%) using a two-cavity bunching section operating at 1300 MHz. RKA structures simulated in ISIS have extracted up to 250 MW. This number was consistent with the extracted power actually measured in the equivalent experiment. Overall energy extraction was as high as 160 J per pulse. The average rf output power coupled into the 6-inch-diameter coax transmission line was approximately: 300 MW for 300 ns, 250 MW for 500 ns, and 100 MW for 1 $\mu$s. Peak power levels as high as 475 MW have also been produced.

The Voyager 2 encounter with Neptune and Triton in August of 1989 provided clues to an intriguing problem. Instruments onboard the spacecraft showed a large ionosphere at Triton. Subsequent studies have tried to explain the production of such high levels of ionization but have ignored the possible plasma dynamics originating from the interaction between Neptune's magnetosphere and Triton. This study applies knowledge gained from studying the solar wind-Venus interaction to this case. In doing so we find that observations made by Voyager 2 can be explained by downward convection of magnetospheric plasma into Triton's atmosphere. Furthermore, we find that the flowing momentum is transferred to the moon just below the exobase, calculated here to be approximately 750 km. From this point down the atmosphere is not in hydrostatic equilibrium, and cannot be until below the ionization peak. Finally we show that when the momentum gets transferred to the moon the flow must shut off. This is accomplished when both the convective velocity and magnetic field go to zero. By showing the magnetic field vanishes at an altitude of roughly 650 km, we conclude the accepted mechanism by which the ionosphere is produced to be invalid. This mechanism was identified early-on to be impact ionization from hot, or superthermal, electrons originating in Neptune's magnetosphere. These precipitating hot electrons are thus shown to operate independently of the magnetic field below the exobase. This is a result not previously discovered, and one which implies that the plasma interaction between Neptune's magnetosphere and Triton cannot be ignored.

A modified steady state flow/field model is applied to the direct interaction of the solar wind with the Martian ionosphere. The original flow/field model (Cloutier et al., 1987) is a one-dimensional, self-consistent derivation of differentials in vertical velocity, magnetic field, and ion densities from the coupled MHD equations. While successful in reproducing features of the ionosphere of Venus (Cloutier et al., 1987; McGary, 1987) and of Mars (Stewart, 1989), the flow/field model required an independently specified heating term (Q). The requirement of this term implies the presence of an energy source not accounted for in conventional calculations. This source was previously simulated with the inclusion of Q, but an unrecognized momentum or pressure term may also provide the coupling with the solar wind without the need of the free parameter Q. An in-depth analysis of Pioneer Venus data in relation to the total conservation of momentum of the system led to the discovery that the total momentum was in most cases not entirely accounted for, and that this "missing" term was correlated with solar wind dynamic pressure. By including this missing pressure, a new set of differential equations, which were also extended to include horizontal velocity terms, was derived. Extrapolation of the missing pressure to Mars gave results that faithfully reproduced the ionospheric features associated with previous flow/field models while maintaining agreement with Viking 1 and 2 observations. Finally, we suggest that the source of P$\sb{\rm missing}$ could be a population of suprathermal particles within the ionosphere. The missing pressures in the Viking simulations are consistent with measured suprathermal pressures at Mars (Hanson and Mantas, 1988).

The Pioneer Venus Orbiter Electric Field Detector (PVO OEFD) detected numerous impulsive electric field events in its 100 Hz channel when in nightside ionospheric troughs. What the source of the signals is has been under debate for over 10 years. Some researchers claim that lightning is generating whistler waves which are then being detected, while other researchers have supported the view that the signals are caused by a local generation of a plasma instability due to the conditions inside the troughs. We believe that the evidence collected by the PVO clearly disproves a lightning source and, in fact, points to an electrostatic wave as being the producer of the transients. We studied several wave modes in an effort to determine the source of the signals. We looked at the two-stream instability, the ion-acoustic instability, the gentle-bump instability, the lower-hybrid-drift instability and the Alfven wave mode. We found that not one of the wave modes studied could account for all of the signals. It would appear that there is not a large enough current within the troughs to support the two-stream instability. While the ion-acoustic instability may well be present within some troughs, the conditions needed to produce 100 Hz waves may not be universal trough conditions. Thus, this wave mode is not a likely generator of all the signals. The gentle-bump instability cannot produce 100 Hz waves and any Alfven waves will be damped. The lower-hybrid-drift instability seems to be the most likely candidate of the wave modes studied for producing the transient signals. The frequencies of the lower-hybrid-waves, however, appear to be too low to be detected by the OEFD. We believe that either the frequencies that we calculated are too small due to factors not taken into account, or there is a cascade of energy from the lower-hybrid waves to waves with frequencies that can be detected by the OEFD, or there is a combination of the two.

A semi-quantitative comparison has been made of the observed and calculated precipitating electron energy fluxes for the April 1988 magnetic storm. Electron energy fluxes were calculated by the Rice Magnetospheric Specification Model (MSM), a comprehensive model of the inner magnetospheric environment, and by the Hardy et al. model, a statistical model of electron precipitation in the auroral zone. The MSM correlates better with the observed fluxes than does the Hardy et al. model in terms of auroral boundaries, latitudinal profile and extent, and the actual magnitude of the energy flux. The sources of error in the MSM are probably: (1) Artificial flux dropouts created near the ionospheric projection of the model outer boundary, (2) an overestimate of the convection electric field, and (3) errors in locating the polar cap boundary.

A series of improvements to the Rice Field Model (RFM) are described, which both increase the accuracy and extend the capabilities of the model. A new ring current parameterization improves the determination of storm-time fields in the inner magnetosphere. Replacement of the tail current module with a more flexible representation also contributes to improved accuracy in the inner magnetosphere, and enables realistic variations in current strength and orientation over the entire magnetotail length. Revision of the tail shielding/interconnection field eliminates inconsistencies in the model magnetotail, and permits variation in the normal component distribution over the tail portion of the magnetopause. The enhanced flexibilities of the interconnection field and cross-tail current module make possible the modeling of variations in the interplanetary magnetic field (IMF) as it propagates downstream, thereby advancing the steady-state RFM an important step toward time-dependent modeling. The modified RFM is used to explore a number of time-dependent magnetospheric phenomena. In simulations of the March 1998 magnetic storm, the new model displays an improved representation of the inner magnetosphere, accurately predicting both storm-induced variations and day-night asymmetry in the field at geosynchronous orbit. The effects of time-dependent interplanetary fields on magnetospheric convection are examined, using a new method to compute ionospheric flow and electric fields in non-steady configurations. This method is applied to simulations of the growth and contraction of the polar cap in Southward and Northward turnings of the IMF. Model convection patterns for Southward turnings are shown to be consistent with theoretical expectations. The RFM is also used to simulate polar cap convection in the particular IMF conditions believed to trigger formation of the theta aurora. The results of the simulation prove to be consistent with several observed properties of the theta aurora, and shed light on the plasma sheet and magnetotail configurations associated with this phenomenon.

Before the IMAGE mission, electron densities in the high latitude, high altitude region of the magnetosphere were measured exclusively by in situ means. The Radio Plasma Imager instrument onboard IMAGE is capable of remotely observing electron densities between 0.01 and 100,000 e-/cm-3 from distances of several Earth radii or more. This allows a global view of the high latitude region that has a far greater accuracy than was previously possible. Soundings of the terrestrial magnetic cusp provide the first remote observations of the dynamics and poleward density profile of this feature continuously over a 60-minute interval. During steady quiet-time solar wind and interplanetary magnetic field conditions, the cusp is shown to be stable in both position and density structure with only slight variations in both. Peak electron densities within the cusp during this time are found to be somewhat higher than predicted. New procedures for deriving electron densities from radio sounding measurements are developed. The addition of curve fitting algorithms significantly increases the amount of useable data. Incorporating forward modeling techniques greatly reduces the computational time over traditional inversion methods. These methods are described in detail. A large number high latitude observations of ducted right-hand extraordinary mode waves made over the course of one year of the IMAGE mission are used to create a three dimensional model of the electron density profile of the terrestrial polar cap region. The dependence of electron density in the polar cap on average geocentric distance (d) is found to vary as d-6.6. This is a significantly steeper gradient than cited in earlier works such as Persoon et al., although the introduction of an asymptotic term provides for basic agreement in the limited region of their joint validity. Latitudinal and longitudinal variations are found to be insignificant. Both the mean profile power law index of the electron density profile and, to a stronger degree, its variance show dependence with the DST index.

This thesis is an observational and theoretical study of the optical emission from nonradiative shocks in three supernova remnants: the Cygnus Loop, RCW 86 and Tycho, which together span a wide range of shock velocities (300 ≲ vS ≲ 2000 km s-1). The spectra are dominated by Balmer lines of H which have both a broad component caused by proton-neutral charge exchange and a narrow component produced by collisional excitation close to the shock front. The broad to narrow flux ratios observed in all three remnants are systematically smaller in Halpha than in Hbeta, and the narrow Balmer decrement is larger than the broad Balmer decrement. The broad component Halpha profiles of RCW 86 and Tycho are Gaussian, indicating that the postshock protons follow a Maxwellian velocity distribution. To model the data, a new numerical shock code was developed which computes the broad and narrow Balmer line emission from a nonradiative shock in partially neutral gas. The models calculate the density, temperature and size of the postshock ionization layer for arbitrary electron-proton temperature equilibrations, and use a Monte Carlo simulation to compute narrow Balmer line enhancement from Lyman line trapping. The models constrain the shock velocity and equilibration of nonradiative shocks in each remnant using the observed Halpha and Hbeta broad to narrow ratios. The models show that differences between the observed broad and narrow Balmer decrements can be explained by Lyman line trapping. The models also show that variations in electron-proton equilibration can reproduce the observed range of broad to narrow ratios. The results give 50--100% equilibration in nonradiative portions of the NE Cygnus Loop (vS ∼ 300 km s-1) and 40--50% equilibration in nonradiative portions of RCW 86 (vS ∼ 600 km s-1 ). In Tycho there are major discrepancies between the predicted and observed broad to narrow ratios, with only the Hbeta ratio matching the observations. The discrepancies may be due to additional narrow component emission that arises from a cosmic ray/fast neutral precursor. The observed Hbeta broad to narrow ratio implies ≲ 20% equilibration in Tycho. Hence, there is an inverse correlation between Mach number and equilibration for the three observed remnants. This correlation suggests there may be significant differences between collisionless shocks at low and high Mach numbers. The spectroscopic observations also led to the serendipitous discovery of a photoionization precursor in Tycho. The precursor appears as a faint, diffuse region that extends several arcminutes ahead of the Balmer-dominated shocks in Tycho. A new photoionization model shows that the diffuse emission is mostly warm, neutral gas heated to ∼12,000 K by He II lambda304 photons from the nonradiative shock. High resolution spectra indicate that the upstream gas is further heated to ∼40,000 K just before entering the shock. This additional heating may arise from a second precursor produced either by cosmic rays or fast (broad component) neutrals from behind the hot postshock gas.

Venus' nightside ionosphere may be characterized by two general magnetic field geometries. Most of the ionosphere is characterized by largely horizontal fields carried over across the terminator from the dayside. However, regions of mostly vertical fields, associated with the nightside ionospheric 'holes' and connected to regions in the ionosheath where there are strong tailward fields, are also present. A magnetic potential model has been developed to describe the interface of the ionosphere, represented as an infinite slab bounded by two magnetically impermeable planes, with the tail lobes idealized as semi-infinite solenoids terminating at the ionopause. From this model, ionospheric boundary currents are generated to confine all field lines in either the flat ionosphere or the vertical tail lobe, and the field geometries are explored. Techniques for mapping this solution to a sphere and approximating it at long ranges from the transition region are discussed, and a preliminary global nightside model is offered.

We analyze the solar wind interaction with Mars through examination of Mars Global Surveyor Magnetometer/Electron Reflectometer (MGS MAG/ER) data. We focus on data in the postshock, dayside plasma flow, especially the Magnetic Pileup Boundary (MPB). First, we present the data to characterize the features of the MPB. Next, we argue the evidence that the MPB is formed through a series of processes, beginning with electron impact ionization of planetary neutrals in the exosphere of Mars. These new ions form an unstable population of energetic ions in the post-shock flow. Ion cyclotron waves are established to scatter the ion distribution into one more stable. This removes thermal pressure from the ions. In order to maintain a constant total pressure in the flow, the magnetic field intensifies, transferring the lost ion thermal pressure into magnetic energy. Also, we develop a model of the electron impact ionization process in the martian exosphere. This model calculates the evolution of the electron distribution function as the flow encounters exospheric planetary neutrals. It reproduces the electron spectrum observed by the ER in the MPB very well. Therefore, we conclude that electron impact ionization is the process responsible for the onset of magnetic pileup in the postshock flow at Mars.